journal of public health in the deep south · survey designed by mathematica policy research was...

VOLUME 1, ISSUE 1

MARCH 2019

JOURNAL OF PUBLIC HEALTH IN THE DEEP SOUTH

Journal of Public Health in the Deep South | Volume 1, Issue 1 | March 2019

Contents

Using Medicaid Data to Identify Factors that Predict Reinstitutionalization of Mississippians with Disabilities and Elderly People Hwanseok Choi, Mina Li, Selena Frederick, and Charkarra Anderson-Lewis The Relationship Between Obesity and Depression Among Federally Qualified Health Center Patients Hwanseok Choi, Joohee Lee, Stephanie T. McLeod, Rambod A. Rouhbakhsh, Michelle Brazeal, Tim Rehner, and David M. Cochran Remaining in the Workforce After Motherhood: Does the Family Medical Leave Act Play a Role in the Decisions of Mississippi Mothers? Jennifer Balcazar, Danielle Fastring, and Avery Hilbert Speech-Language Pathologists and Respiratory Therapists: Team Approach to Caring for Patients with Long-Term Tracheotomy Javis M. Knott and Celeste R. Parker

Maternal, Child, and Parenting Factors Associated with Obesity Among Pre-Kindergarten Children in Mississippi Jerome R. Kolbo, Angel Herring, Hwanseok Choi, Bonnie L. Harbaugh, Elaine Fontenot Molaison, Olivia Ismail, Lindsey Hardin, and Nichole Werle Training the Next Generation of Primary-Care Physicians: Are Student-Run Free Clinics (SRFCs) the Way to Go? Tobe Momah, Rita Momah, William Replogle, Elizabeth McClain, and Makayla Merritt Rural Medical Scholars Program: Filling the Gap for Healthcare and Public Health Leaders in Mississippi Ann Sansing, David R. Buys, Marion W. Evans, Laura Downey, and Jasmine Harris- Speight

Editorial Board Editor David Buys, PhD Mississippi State University [email protected] Executive Director Buddy Daughdrill Mississippi Public Health Association [email protected]

Co-Editor Danielle Fastring, PhD University of Southern Mississippi [email protected] Managing Editor Mary Nelson Robertson Mississippi State University [email protected]

Identifying Factors that Predict Reinstitutionalization of Mississippians


Using Medicaid Data to Identify Factors that Predict Reinstitutionalization of Mississippians with Disabilities and Elderly People

Hwanseok Choi University of Southern Mississippi

Mina Li

University of Southern Mississippi

Selena Frederick Denver Health

Charkarra Anderson-Lewis


Background: Mississippi Bridge to Independence (B2I) was Mississippi’s Money Follows the Person (MFP) program seeking to rebalance the state’s long-term care system by transitioning Medicaid beneficiaries from institutional living to home- and community-based settings (HCBS). Success of initial transitions has been documented in state cost-savings and participants’ quality of life increases. However, reinstitutionalization poses a challenge to sustaining a positive outcome for the initiative. Purpose: Therefore, the purpose of this research is to identify the underlying causes of participants’ reinstitutionalization. Methods: The Quality of Life (QoL) survey designed by Mathematica Policy Research was used to measure variables in 7 categories. Data was collected from 399 participants in face-to-face interviews over a 4-year period (2012–2016). Results: Among participants, 71.9% (n = 287) completed the B2I program successfully, whereas 8.27% (n = 33) were reinstitutionalized. Utilizing the logistic regression model, results determined elderly people were 15 times and those with physical disabilities were 5 times more likely to be reinstitutionalized than those with intellectual disabilities. Among 7 QoL variables, 2 were found to be significant: “Happiness” and “Choice and Control.” Conclusion: Implications from this study can be important to sustaining the project, developing new policies, and advancing community-supportive infrastructure in Mississippi. Keywords: disability study, reinstitution, quality of life, independence, logistic regression



Introduction

In the last three decades, the United States has had an increasing trend of long-term care in facilities such as nursing homes as the aged population has increased. This trend is likely to continue due to the increasing number of people surviving to 65-plus years of age. In 2016, around 49.2 million Americans were ages 65 or older, and that number is expected to rise to 98 million by 2060 (U.S. Department of Health and Human Services, 2017). According to the Centers for Disease Control and Prevention (CDC), around 5.8 million people ages 18 or older had limitation in activities of daily living, and more than 106 million had limitation in instrumental activities of daily living in 2016 (National Center for Health Statistics, 2016). To reduce the burden of states’ long-term care cost and to increase beneficiaries’ quality of life, various federally funded programs have provided home-and community-based services (HCBS) over the last decade. Some studies reported that these programs can reduce their expenditure if the participants stay at HCBS for a 12-month period (Bohl, Schurrer, Miller, Lim, & Irvin, 2015). Programs like Money Follows the Person (MFP) can be successful if the participants experience better quality of life and higher quality of care with no additional costs than they would have if they remained in institutional care. However, many of the participants return to the institutions after transitioning to HCBS for various reasons, such as low quality of care, poor medication system, language barriers, and so forth (Hostetter & Klein, 2012). In Mississippi, MFP was implemented from 2012 to 2016 under the name Bridge to Independence (B2I). Prior to the mid-20th century, it was common in the U.S. to institutionalize people with disabilities who required long-term care, excluding those who were labeled “abnormal” from normal social interaction (Burrell & Trip, 2011). In 1977, it was estimated that 83.7% of people with intellectual disabilities/developmental disabilities (ID/DD) using residential services lived in institutional settings with 16 or more residents, representing more than 200,000 people with ID/DD (Lakin & Stancliffe, 2007). More and more research studies found that alternative, community-based care was more effective than hospital care for patients with mental disabilities in terms of treatment results, living arrangements, and expenses (Kiesler, 1982). This coincided with a shift in the philosophical approach used for those who required long-term care (Burrell & Trip, 2011). Across the world, societal treatment of people with disabilities began to be seen as a human rights issue, focusing more on quality of life and encouraging individualization of needs-response and treatment programs (Mental Welfare Commission for Scotland, 2003). As research continued and community-based residential programs began to develop, those patients who did not require acute treatment began to be deinstitutionalized, shifting the care of people with disabilities from a hospital setting to smaller, community-based settings (Dorwart & Hoover, 1994). This process of deinstitutionalization began to significantly reduce the populations in state-run hospitals (Fisher et al., 2001). By 2005, a mere three decades later, only



16.3% of people receiving residential services lived in large-scale institutional settings (Lakin & Stancliffe, 2007). This applies not only to those with ID/DD but also those with physical disabilities and elderly people who require long-term care. As civilization moved from the large, extended-family model to the nuclear model, it became atypical for the elderly to be cared for at home and led to higher rates of institutionalization. But studies have shown that receiving even small amounts of help in the home setting may prevent the elderly from needing institutionalization (Yamada, Siersma, Avlund, & Vass, 2012). Those who utilized available home- and community-based services, especially respite care and rental services for assistive devices, were less likely than non-users to be institutionalized (Tomita, Yoshimura, & Ikegami, 2010). Studies also found that supplementing family or informal care with home-based formal care might reduce the risk for hospitalization in community-living older adults who were eligible for long-term care (Li, Chi, & Xu, 2010). In addition, hospitalization rates significantly decreased after non-institutionalized, disabled elderly people were provided with help for activities of daily life (ADL) disabilities, through community-based, long-term care services (Sands et al., 2006). Community-based services could have a very positive effect on the lives of those who might have otherwise been hospitalized or institutionalized. These changes in both philosophy and execution of approach represent an advantage in cost and treatment effectiveness, as well as in the quality of life experienced by those in long-term care. In 2001, O’Brien, Thesing, Tuck, and Capie found that, after deinstitutionalization,

the positive changes and advantages discerned for [those receiving long-term care] were further encapsulated in how the person’s current quality of life was perceived. Both staff and family indicated there had been significant increases in satisfaction levels related to the extent and type of material possessions, the person’s health, meaningful daily activities, safety, the person’s place in the community and emotional well-being (p. 78).

The same study also reported an increase in social and adaptive skills among those with long histories of institutionalization. Community-based care provided a sense of community that usually was not possible in hospital and institutional settings. However, it was not a simple matter of moving patients from one building to another. In order for deinstitutionalization to be effective, it had to be a proactive and self-aware process. Programs had to be in place to ensure that patients continued to receive appropriate levels of health care and that they were given the tools needed to succeed in a community-based setting. Research showed that inadequate preparation before discharge left a significant number of patients with mental disabilities homeless or in prison (Eikelmann, 2000). In some cases, community services provided insufficient levels of treatment for patients with disabilities (Lamb,



2001). Even after deinstitutionalization, many patients with severe illnesses were still without work, had few social contacts, and lived in sheltered environments (Fakhoury & Priebe, 2007). This lack of social integration for patients following their discharge from institutional settings was still far from the desired full integration (Eikelmann, Reker, & Richter, 2005). The challenge ahead was to develop an adequate supply of high-quality, community-based service systems to successfully sustain deinstitutionalization (Vladeck, 2003). The Mississippi Bridge to Independence (B2I) program was focused on the transitional period after deinstitutionalization. It helped to remove barriers to home- or community-based living for people with physical, intellectual, or developmental disabilities, as well as older adults (65 years and older). It was an MFP initiative funded by the Centers for Medicare and Medicaid Services (CMS). B2I provided transition services for qualified individuals after their discharge from an institutional setting. These services included transition care management, housing start-up costs, transportation, extended pharmacy benefits, durable medical equipment, caregiver and peer support, life skills training, and safety planning. Schurrer and Wenzlow (2011) studied the post-transition outcomes (from institutionalized to home- or community-based settings) of MFP participants in 25 states and found that 9% of the participants were reinstitutionalized during the first year of transition. They concluded that those participants who re-entered the institution did so mostly within the first 3 to 6 months after transition. This was the period when the responsibility of the MFP services was moved from transition experts to care coordinators. Researchers concluded that it was essential to have a smooth transfer process to ensure the continuity of services during the first, most vulnerable months of participants’ integration into the community (Schurrer & Wenzlow, 2011). A decline in the individual’s mental or physical health was the main factor for reinstitutionalization. Other reasons included short-term hospitalization, lack of family or other support systems within the community, and loss of housing (Denny-Brown, Lipson, Kehn, Orshan, & Stone, 2011). Though researchers have suggested a few factors that might lead to reinstitutionalization, there was no definitive explanation in the literature. But this was a situation that could not be ignored: with the maturation of the so-called baby boomer generation, the need for adequate and effective long-term care was becoming more urgent with each passing year. By 2050, the U.S. population aged 65 and over is predicted to reach 83.7 million, nearly twice its estimated population of 43.1 million in 2012; similar increases are expected in all developed countries (Ortman, Velkoff, & Hogan, 2014). A large portion of the global population would need long-term health-care services, and it would be important to ensure that we have programs in place to help patients transition from an institutional setting to a community-based service model. Therefore, the purpose of this research was to determine the factors affecting reinstitutionalization among the Mississippi B2I program participants at 12 months post-transition from institutions to HCBS in an effort to lessen the likelihood of reinstitutionalization. Findings could lead to new policy



development, long-term care cost-effect rebalancing, program and practice improvement, and enhancement of beneficiaries’ quality of life.

Methods

Participants and Process According to the eligibility criteria provided by the Centers for Medicare and Medicaid Service’s Money Follows the Person (Mathematica Policy Research Report, 2017a), Mississippi Medicaid beneficiaries who were willing to transition from institutionalized care to home- and community-based settings were referred to the B2I program from local nursing homes or Intermediate Care Facilities for Individuals with Intellectual Disabilities (ICF/IID) across the state. Qualified B2I participants were those who had lived in an institution for at least 90 days and were either 18 years of age or older with a disability or 65 years of age or older. Based on the program procedure, each B2I participant was given a personal interview using a survey instrument to determine quality of life pre- and post-transition. From January 2012 to December 2016, a total of 399 beneficiaries had transitioned to the community or enrolled in B2I. Data collected by Medicaid program staff from these individuals were used as a cohort for the study. Of the 399 participants, 232 completed a 12-month follow-up survey using an identical survey instrument (Mathematica Policy Research Report, 2017a). According to U.S. Department of Health and Human Services (HHS) regulations for the protection of human subjects in research (Office for Human Research Protections, 2018), this study did not require approval by the Institutional Review Board. Instrument The Quality of Life (QoL) survey was developed by Mathematica Policy Research to measure MFP implementation. Using the same instrument, B2I was evaluated and this study conducted. The QoL contains 41 multiple-choice and short, open-ended questions in seven modules, plus two follow-up questions. The seven modules used to measure the quality of life of participants include Living Situation, Choice and Control, Access to Personal Care, Respect and Dignity, Community Integration and Inclusion, Overall Life Satisfaction, and Health Status. The instrument was administered to the program participants at three points in time: just prior to transition, 12 months after transition, and about 24 months after transition. Trained Medicaid personnel administered the questionnaire. To ensure a consistent data-collection procedure and high-quality data, the process followed the protocol and instructions provided by Mathematica Policy Research. Socio-demographic variables. The following socio-demographic variables were included: gender, age, kinds of disability (physical, intellectual, and elderly), race/ethnicity (Caucasian, African American, and other), living with family (yes/no), and residence type



(family/participant-owned home, leased apartment, and group-home living with four or more people). Quality of life variables. The items in the seven modules for the quality of life measurements were summed up as total score. The seven modules were Living Situation (0–4), Choice and Control (0–12), Access to Personal Care (0–4), Respect and Dignity (0–5), Community Integration and Inclusion (0–13), Overall Life Satisfaction (0–10), and Health Status (0–6). Aggregated higher scores indicated more positive or optimistic outcomes of B2I participants’ quality of life. The detailed scores for the seven modules are in the Appendix. Outcome variable. Researchers recorded and followed the B2I participants who remained in the community-living setting and those who were reinstitutionalized. Reinstitutionalization is defined as any admission to a hospital, nursing home, intermediate-care facility, or institution for the same diseases after transitioning to a community-living setting. Reinstitutionalization is one of the key success indicators of B2I transitions. A lower rate is a better indicator of transition success.

Statistical Analysis After the descriptive analyses of the socio-demographic and quality of life variables, we evaluated the relationships between each variable and the outcome variable (reinstitutionalization) using chi-square independence tests, Cochran-Mantel-Haenszel chi-square tests, and Fisher’s exact tests at 5% significance level. Continuous variables were checked for a normality assumption and assessed using independent two sample t-tests for continuous variables and chi-square independent tests for categorical variables. We performed multivariate logistic regression models to attempt to identify which socio-demographic and disaster-related factors were associated with an increased likelihood of being reinstitutionalized.

Results

Sample Characteristics This study included 399 B2I program beneficiaries in Mississippi from 2012 to 2016. Among them, 232 participants responded to the 12-month follow up. Since all reinstitutionalization happened within a year of the transition, we used the 12-month follow-up data as the study sample. During the year, nine participants were deceased, nine moved to other regions, 11 no longer needed services, and three were suspended from eligibility. Two participants were removed from the sample pool for health reasons. After the transition to home- or community-based settings, 33 were readmitted to institutional facilities (8.3%), while 287 completed the program (71.9%) and remained in the communities.



Among 399 participants, about 57% were people with intellectual/developmental disabilities, 28% had physical disabilities, and 14% were elderly. The participants were 59% male and 53% Caucasian. Participants’ ages ranged from 4 to 91, with a median of 43 years old. The type of residences participants were transferred to included 23% homes owned by participants or families, 34% leased apartments, and 43% group homes. Only about 26% of participants were living with their families. Table 1 summarizes the characteristics of the participants. The descriptive analysis performed on outcome variables indicated the reinstitution rate was 8.3% (n = 33).



Table 1 Participants’ Demographic Characteristics (N = 399)

Variables

N % Mean (SD)

Benchmark Elderly Physically disabled Intellectually or developmentally disabled

Gender

57 114 228

14.29 28.57 57.14

Male 237 59.40 Female 162 40.60

Age Age group

Less than 25 years old 25 to 34 years old 35 to 44 years old 45 to 54 years old 55 to 64 years old More than 65 years old

74 77 42 80 63 63

18.55 19.30 10.53 20.05 15.79 15.79

44.36 (18.85)

Race/Ethnicity

African American 184 46.12 Caucasian 212 53.13 Other 3 0.75

Residence

Family/participant-owned home 90 22.56 Group home, four people or fewer 136 34.09 Participant-leased apartment 173 43.36

Living with family

Yes 102 25.69 No 295 74.31

Program completion Reinstitutionalized Completed B2I program Dropped

33 287 79

8.27 71.93 19.80



Bivariate Analyses To explore which variables were significantly associated with reinstitutionalization, bivariate analyses were performed (Table 2). The chi-square test of independence was conducted with significance level at 0.05 for the nominal-scale variables, and an independent two-sample t-test was conducted for the continuous variable. For the ordinal-scale variables, the Cochran-Mantel-Haenszel test for linear association was performed. Results indicated that the control variables, such as disability type, living status after the transition, and age, were significantly associated with reinstitutionalization. Among the QoL variables, the total score of living situation, happiness, community integration and inclusion, and choice and control were statistically significant at α = 0.05. These variables were included in the final model to explain reinstitutionalization by multivariate analyses. Table 2 Bivariate Analyses Between Reinstitutionalization and Other Variables

Test Test

Statistic P Socio-demographic variables

Benchmark χ2 22.82 < 0.001 Residence χ2 17.42 < 0.001 Gender χ2 0.02 0.883 Age t 4.76 < 0.001 Race/Ethnicity χ2 1.76 0.415 Living with family χ2 1.10 0.294

Quality of life variables (1-year follow-up)

Respect and Dignity Score CMH 1.64 0.200 Living Status Situation Score CMH 7.967 0.005 Health Status Score CMH 1.615 0.203 Happiness Life Satisfaction (Happiness) Score

CMH 13.714 < 0.001

Community Integration and Inclusion Score

CMH 3.930 0.047

Choice and Control Score CMH 9.758 0.002 Access to Personal Care Score CMH 0.131 0.717



Multivariate Analyses Prior to applying the logistic model, multicollinearity was checked among independent variables. Only two variables were strongly correlated (disability type and age), which had more than two variance influence factors (VIF). Therefore, one of these variables was dropped from the model to avoid redundant information and to allow confounding factors among the remaining variables. The multivariate logistic regression was used to determine which socio-demographic and quality of life factors were associated with an increased likelihood of reinstitutionalization (Table 3). The first model included only socio-demographic variables such as gender, race, disability type, age group, post-transition residence type, and whether or not participants were living with family. The global model likelihood ratio test indicated that the overall model was statistically significant (χ2 = 30.42, df = 3, p < .001). Among the socio-demographic variables, age and residence were significantly associated with reinstitutionalization. If a participant’s age increased by 1 year, the odds of reinstitutionalization increased by 3.4% (OR = 1.03, 95% CI: 1.01–1.06, p = .003). If a participant lived in a leased apartment or a home owned by the participant or the participant’s family, the odds of reinstitutionalization were five times more than those who lived in a group home (OR = 5.14, 95% CI: 1.40–18.79, p = .013, OR = 5.06, 95% CI: 1.28–19.94, p = .021, respectively). Secondly, seven quality of life variables were added to the model to detect the relationship between reinstitutionalization and those variables, considering socio-demographic variables simultaneously. The global model likelihood ratio test indicated that the overall model was statistically significant at α = 0.05 (χ2 = 34.92, df = 4, p < .001). Compared to the group of people with intellectual or developmental disabilities, the elderly group had more than 15 times the chance of reinstitutionalization (OR = 15.26, 95% CI: 4.55–51.13, p < .001); the group with physical disabilities had about 5 times the chance of being readmitted into institutions (OR = 5.22, 95% CI: 1.56–17.47, p = .007). Among the QoL variables, the total score of happiness showed the statistical significance at α = 0.05. Those who reported being happy after the transition were about 25% less likely to be reinstitutionalized (OR = 0.75, 95% CI: 0.59–0.96, p = .023). Furthermore, participants who had more chance to interact in the community or to have job/work opportunities in their communities were 24% less likely to be reinstitutionalized (OR = 0.76, 95% CI: 0.57–0.99, p = 0.046).



Table 3 Multivariate Logistic Regression Model

Discussion

The Mississippi Division of Medicaid B2I project, Mississippi’s Money Follows the Person initiative, was implemented (2012–2016) through grant support from the Centers for Medicare and Medicaid Services (CMS). The project has made positive impacts on rebalancing the state’s long-term care system, strategizing Medicaid money-saving practices, and changing clients’ quality of life through transitioning Medicaid beneficiaries from institutional facilities to home- and community-based settings (HCBS). The success of the project implementation has been well documented in program reports and evaluations throughout the awarded states, including Mississippi. To sustain the impact of the initiative, one of the challenges is to reduce the number of program participants being re-admitted to institutional facilities after transitioning to HCBS. In this study, we used seven variables from an established quality of life (QoL) survey instrument to identify the factors that may contribute to B2I participants’ reinstitutionalization using descriptive analyses, bivariate analyses, and multiple logistic regression. Among 399 B2I participants, the reinstitutionalization rate was 8.3% at 1-year post-transition sample follow-up, which is higher than the national rate. For all states participating in the program, about 5% of participants were reinstitutionalized for more than 30 days in 2016

Socio-demographic variable model B p OR 95% CI

Socio-demographic variables-only model Age Residence

0.033 0.003 1.034 1.011–1.057

Participant-leased apartment Family/participant-owned home

1.636 1.621

0.013 0.021

5.135 5.059

1.404–18.785 1.284–19.940

Participant-leased apartment

Benchmark Elderly Physically disabled

2.725 1.652

< 0.001 0.007

15.259 5.216

4.554–51.129 1.557–17.472

Happy score -0.287 0.023 0.751 0.586–0.961 Community score -0.270 0.046 0.764 0.586–0.995



(Mathematica Policy Research Report, 2017a). Significant indicators for reinstitutionalization were identified as follows:

• Age of the participants: It appears advanced age of the participants is associated with likelihood of reinstitutionalization, which is concordant with the results of the national report.

• Type of disability: People with physical disabilities are more likely to be readmitted to institutions compared to the group of people with intellectual or developmental disabilities, which is also concordant with the results of the national report.

• Type of residence (after transition): Participants living in a group home (four or fewer people) seem to have a smaller risk of reinstitutionalization than those who transition to leased apartments or homes owned by participants or families.

• Satisfaction and happiness (after transition): Participants who have a higher level of satisfaction (happiness) with their life after transition show a reduced risk for reinstitutionalization.

• Community integration and inclusion: Participants who lack family and community support after transition or who live in social isolation have higher rates of reinstitutionalization.

Factors impacting B2I participants’ reinstitutionalization rates were identified from this study as age, type of disability, level of community integration and inclusion, and type of housing. These findings agree with the literature for increasing the risk of reinstitutionalization. Community integration and inclusion was identified as one of the key factors of success for the program by many states, including Mississippi, through separate quantitative and qualitative studies (Mathematica Policy Research Report, 2017b). In this study, we confirmed that increased involvement and support from the community, such as job or volunteer opportunities, community organized activities, and so on, reduced Medicaid beneficiaries’ risk of return to institutions. In addition, our results highlight the importance of community involvement and its subsequent effects on people’s happiness or life satisfaction. Housing is also a crucial factor for the success of the program, according to the national MFP evaluator (Mathematica Policy Research Report, 2017b). We found type of housing could affect sustainability of transition. Specifically, we found better outcomes for participants who transitioned to group homes (defined as “community homes where a small number of unrelated people in need of care, support, or supervision can live together with supports and services”) than for those who took part in supported living (defined as “services designed to help persons with disabilities live in their own home or live in a home that they share with roommates of their choosing”). Further investigation is suggested to determine if group homes affect the reinstitutionalization rate for the elderly and people with physical disabilities whose higher institution returning rate has been detected. In our investigation, we are unable to identify the relationship between three of the seven variables and reinstitutionalization (p > 0.05). They are: Respect and Dignity, Health Status, and



Access to Personal Care. It is our belief that some of the questions from the survey are not accurate, precise, or specific enough to reflect the true perception. Better instrument development to fulfill the study purpose is recommended for a future study. In addition, confounding factors are suspected in this study. Interactive effects could be considered for further research using factorial analysis. Structural equation modeling may be used to explore more intricate relationships among the factors. We realize identified factors in this study are limited by the variables derived from the QoL survey instrument. Factors associated with B2I participants’ reinstitutionalization rate should be multi-faceted and include participants’ socio-economic status; education level; level of support received from health-care providers, family, and community; and ability to live independently. Further investigation may result in a better understanding of reinstitutionalization risk indicators if redesigned variables with clear-cut specificities are applied to a survey instrument and subsequent test methods.

References

Bohl, A., Schurrer, J., Miller, D., Lim, W., & Irvin, C. V. (2015). The changing medical and long-term care expenditures of people who transition from institutional care to home- and community-based services. Mathematic Policy Research, 14. doi: 10.13140/RG.2.1.2406.6324

Burrell, B., & Trip, H. (2011). Reform and community care: Has deinstitutionalization delivered for people with intellectual disability? Nursing Inquiry, 18, 174–183. doi:10.1111/j.1440-1800.2011.00522.x

Denny-Brown, N., Lipson, D., Kehn, M., Orshan, B., & Stone, C. (2011). Money follows the person demonstration: Overview of state grantee progress, January to June 2011. Final report submitted to the Centers for Medicare & Medicaid Services. Cambridge, MA: Mathematica Policy Research.

Dorwart, R. A., & Hoover, C. W. (1994). A national study of transitional hospital services in mental health. American Journal of Public Health, 84(8), 1229–1234. doi:10.2105/AJPH.84.8.1229

Eikelmann, B. (2000). Limits of deinstitutionalization? —Perspective of the specialty clinic. Psychiatrische Praxis, 27, 53–58.

Eikelmann, B., Reker, T., & Richter, D. (2005). Social exclusion of the mentally ill—a critical review and outlook of community psychiatry at the beginning of the 21st century. Fortschritte der Neurologie-Psychiatrie, 73(11), 664–673.

Fakhoury, W., & Priebe, S. (2007). Deinstitutionalization and reinstitutionalization: Major changes in the provision of mental healthcare. Psychiatry, 6(8), 313–316.



Fisher, W. H., Barreira, P. J., Geller, J. L., White, A. W., Lincoln, A. K., & Sudders, M. (2001). Long-stay patients in state psychiatric hospitals at the end of the 20th century. Psychiatric Services, 52(8), 1051–1056.

Hostetter, M., & Klein, S. (2012). Avoiding preventable hospital readmissions by filling in gaps in care: The community-based care transitions program. The Commonwealth Fund. Retrieved from https://www.commonwealthfund.org/publications/newsletter/avoiding-preventable-hospital-readmissions-filling-gaps-care-community

Kiesler, C. A. (1982). Mental hospitals and alternative care: Noninstitutionalization as potential public policy for mental patients. American Psychologist, 37(4), 349–360. http://dx.doi.org/10.1037/0003-066X.37.4.349

Lakin, K. C., & Stancliffe, R. J. (2007). Residential supports for persons with intellectual and developmental disabilities. Mental Retardation and Developmental Disabilities Research Review, 13, 151–159. doi:10.1002/mrdd.20148

Lamb, H. R. (2001). Deinstitutionalization at the beginning of the new millennium. New Directions for Mental Health Services, 2001(90), 3–20. https://doi.org/10.1002/yd.23320019003

Li, H., Chi, I., & Xu, L. (2010). Factors associated with volunteerism among community-living older adults in urban China. Social Development Issues, 32(1), 62–75.

Mathematica Policy Research Report. (2017a). Money Follows the Person 2015 annual evaluation report. Retrieved from https://www.mathematica-mpr.com/our-publications-and-findings/publications/money-follows-the-person-2015-annual-evaluation-report

Mathematica Policy Research Report. (2017b). Money Follows the Person demonstration: Overview of state grantee progress, January to December 2016. Retrieved from https://www.medicaid.gov/medicaid/ltss/downloads/money-follows-the-person/2016-cross-state-report.pdf

Mental Welfare Commission for Scotland. (2003). Mental Health Act 2003. Retrieved from https://www.mwcscot.org.uk/the-law/mental-health-act/

National Center for Health Statistics. (2014). Disability and functioning (noninstitutionalized adults 18 years and over). Centers for Disease Control and Prevention. Retrieved from http://www.cdc.gov/nchs/fastats/disability.htm.

O’Brien, P., Thesing, A., Tuck, B., & Capie, A. (2001). Perceptions of change, advantage, and quality of life for people with intellectual disability who left a long stay institution to live in the community. Journal of Intellectual and Developmental Disability, 26(1), 67–82.

Office for Human Research Protections. U.S. Department of Health and Human Services. (2018). Code of federal regulations, Title 45, public welfare, Part 46, protection of human subjects. Retrieved from https://www.hhs.gov/ohrp/regulations-and-policy/regulations/45-cfr-46/index.html

Ortman, J., Velkoff, V., & Hogan, H. (2014). An aging nation: The older population in the United States (Current Population Reports, P25-1140). Washington, DC: U.S. Census Bureau.



Sands, L. P., Wang, Y., McCabe, G. P., Jennings, K., Eng, C., & Covinsky, K. E. (2006). Rates of acute care admissions for frail older people living with met versus unmet activity of daily living needs. Journal of the American Geriatrics Society, 54(2), 339–344.

Schurrer, J., & Wenzlow, A. (2011). A first look at how MFP participants fare after returning to the community. Mathematica Policy Research. National Evaluation of the Money Follows the Person (MFP) Demonstration Grant Program, Report from the Field, 7. Retrieved from https://pdfs.semanticscholar.org/eecd/7a03beb00da1d2c58a5c66f44a8bb6d38430.pdf

Tomita, N., Yoshimura, K., & Ikegami, N. (2010). Impact of home and community-based services on hospitalization and institutionalization among individuals eligible for long-term care insurance in Japan. BMC Health Services Research 2010, 10, 345. https://doi.org/10.1186/1472-6963-10-345

U.S. Department of Health and Human Services, Administration for Community Living, Administration on Aging. (2017). A profile of older Americans: 2017. Retrieved from https://acl.gov/sites/default/files/Aging%20and%20Disability%20in%20America/2017OlderAmericansProfile.pdf

Vladeck, B. C. (2003). Where the action really is: Medicaid and the disabled. Health Affairs, 22(1), 90–100.

Yamada, Y., Siersma, V., Avlund, K., & Vass, M. (2012). Formal home help services and institutionalization. Archives of Gerontology and Geriatrics, 54(2), e52–e56. https://doi.org/10.1016/j.archger.2011.05.023

Appendix Quality of Life Variable Score Range

Modules Score Range Module 1: Living Situation 0–4 Module 2: Choice and Control 0–12 Module 3: Access to Personal Care 0–4 Module 4: Respect and Dignity 0–5 Module 5: Community Integration and Inclusion 0–13 Module 6: Overall Life Satisfaction 0–10 Module 7: Health Status 0–6

Aggregated higher scores indicated more positive or optimistic outcomes of B2I participants’ quality of life.



Author Note

Hwanseok Choi, Department of Public Health, University of Southern Mississippi; Mina Li, Institute for Disability Studies, University of Southern Mississippi; Selena Frederick, Denver Health; and Charkarra Anderson-Lewis, Department of Public Health, University of Southern Mississippi. Correspondence concerning this article should be addressed to Hwanseok Choi, Department of Public Health, University of Southern Mississippi, 118 College Drive #5122, Hattiesburg, MS 39406-0001. E-mail: [email protected]

Relationship Between Obesity and Depression


The Relationship Between Obesity and Depression Among Federally Qualified Health Center Patients

Hwanseok Choi


Joohee Lee University of Southern Mississippi

Stephanie T. McLeod


Rambod A. Rouhbakhsh Forrest General Hospital

Michelle Brazeal


Tim Rehner University of Southern Mississippi

David M. Cochran


Background: Obesity has reached epidemic levels in Mississippi. In the shadow of these skyrocketing obesity levels, there are comorbid high levels of depression. Both obesity and depression complicate and, in many cases, compromise critical health outcomes. A significant association between obesity and depression has been suspected for decades. Purpose: The purpose of this study is to examine the relationship between obesity and depression among patients receiving medical care from a Federally Qualified Health Center (FQHC) in a southern state. Methods: The sample was comprised of 3,272 subjects. The Patient Health Questionnaire (PHQ-9) was used to measure the severity of depression, and the Body Mass Index (BMI) was used to measure obesity. Results: Multiple logistic regression analysis revealed that the likelihood of depression decreased as the level of BMI increased, which is the opposite of the results in most previous research. Good physical health lessened the likelihood of depression. Less stress and fewer traumatic life events and greater self-esteem lowered the chance of depression. Conclusion: The findings indicated a need for health education and interventions to influence changes within communities and to address the medical and emotional needs of individuals with obesity and depression. Keywords: obesity, depression, Duke health profile, mental health, traumatic life events



Introduction

Obesity is a complex and growing public-health issue at both the national and state levels. The prevalence of obesity among adults in the United States from 2013 to 2014 was 37.7% (Flegal, Kruszon-Moran, Carroll, Fryar, & Ogden, 2016). Mississippi was ranked second highest in the U.S. for its prevalence of obesity in 2016 (Segal, Rayburn, & Beck, 2017). Mississippi’s prevalence rate of obesity has risen considerably in the last decade; more than one-third of Mississippi adults were obese in 2016 (Mendy, Vargas, Cannon-Smith, & Payton, 2017; Segal et al., 2017). This trend can be expected to continue, especially considering the high rate of childhood obesity in Mississippi (Grant et al., 2016; Segal et al., 2017; UMCC, n.d.). The Child and Youth Prevalence of Obesity Survey reported that more than 40% of school-aged children in the state are classified as overweight or obese (Kolbo et al., 2016). The Centers for Disease Control and Prevention (CDC) lists obesity as a risk factor associated with depression (2015). Multiple studies have reported positive associations between obesity and depression, especially among women (Dearborn, Robbins, & Elias, 2018; de Wit et al., 2010; Hung et al., 2014; Luppino et al., 2010; Mannan, Mamun, Doi, & Clavarino, 2016; Pereira-Miranda, Costa, Queiroz, Pereira-Santos, & Santana, 2017; Xiang & An, 2015). Like obesity, depression is a growing health concern as its prevalence rate has increased among adults. In the United States, the prevalence of depression was 8.1% from 2013 to 2016, and an estimated 46 million people in the nation will be diagnosed with depression by 2050 (Brody, Pratt, & Hughes, 2018; Heo, Murphy, Fontaine, Bruce, & Alexopoulos, 2008). In 2008, Mississippi was ranked first in the U.S. for its prevalence of depression (Gonzalez et al., 2010). Both obesity and depression have been linked to decreased physical health and quality of life, and both co-occur with similar health problems, such as type 2 diabetes, cardiovascular disease, cancer, and mortality associated with these and other health issues (CDC, 2015; Lépine & Briley, 2011; Nigatu, Reijneveld, de Jonge, van Rossum, & Bültmann, 2016; Pan et al., 2010; Segal et al., 2017; U.S. Preventive Services Task Force [USPSTF], 2016; World Health Organization [WHO], 2018a; WHO, 2018b). Unsurprisingly, health outcomes are often worsened when obesity and depression co-occur (Nigatu et al., 2016). In addition to their association with myriad negative health outcomes, obesity and depression are costly illnesses. In the U.S, obesity-related health-care costs in 2014 were approximately $150 billion, and the cost of depressive disorders in 2013 was roughly $71 billion (Dieleman et al., 2016; Kim & Basu, 2016). Health-care costs associated with obesity in Mississippi are expected to approach $3.9 billion in 2018 (Grant et al., 2016; Mendy et al., 2017; UMMC, n.d.). The relationship between obesity and depression has been disputed for decades. Several studies have identified a positive relationship between obesity and depression. Specifically, research has



shown that, as the level of depression increases, so does the level of obesity (Dearborn et al., 2018; de Wit et al., 2010; Hung et al., 2014; Luppino et al., 2010; Mannan et al., 2016; Pereira-Miranda et al., 2017; Xiang & An, 2015). Yet some studies have found that there is an inverse relationship between depression and obesity. Assari (2014) found that the direction of association between depression and obesity differed based on demographic factors like ethnicity and gender. Some found a negative association between the two where depression decreased as obesity increased (Chang & Yen, 2012). Others have found a curvilinear relationship between the two or no association (de Wit, van Straten, van Herten, Penninx, & Cuijpers, 2009; John, Meyer, Rumpf, & Hapke, 2005). Specific demographic characteristics have been found to predict greater risk for obesity and depression. Middle-aged female and Black or Latino individuals have higher rates of obesity (Flegal et al., 2016; National Center for Health Statistics, 2017). A study of Mississippians found similar results; being female, Black, aged between 25 and 44, and having less education and lower income were associated with higher rates of obesity (Qobadi & Payton, 2017). Several national studies have found that females, younger people, multiracial people, and individuals with lower income levels were more at risk for higher rates of depression (Brody et al., 2018; Gonzalez et al., 2010; Heo et al., 2008; National Institute of Mental Health [NIMH], 2017; USPSTF, 2016). The diagnosis and treatment of depression and obesity in underserved areas is difficult. Mississippi was ranked 50th in access to health care and 51st in access to mental health care in 2018, which suggests that many people in the state do not have access to health services (Mental Health America, 2018; U.S. News & World Report, 2018). In addition, obesity and depression are both associated with negative stigma, so some individuals may avoid seeking professional help even if it is available (Harvard Law School Mississippi Delta Project, 2014; WHO, 2018a; Williams, Mesidor, Winters, Dubbert, & Wyatt, 2015). If the barrier to access is resolved, a potential solution to these challenges would be to integrate mental and behavioral health-care services into primary-care clinics (Harvard Law School Mississippi Delta Project, 2014). When mental and physical health are considered together, it is possible to use the presence of one condition as an indicator to screen for the other condition (Luppino et al., 2010; Pereira-Miranda et al., 2017; Xiang & An, 2015). The purpose of this study was to examine the association between obesity and depression among patients at a Federally Qualified Health Center (FQHC) in South Mississippi, while controlling for various covariates including socio-demographic factors, physical health, and psychosocial factors. While a considerable body of research has focused on the association between obesity and depression, relatively limited studies have included people in the southern part of the U.S. and those of low socioeconomic status. Therefore, this study will expand current knowledge of



the relationship between obesity and depression as well as factors that are significantly associated with depression among disadvantaged patients in South Mississippi who received care through an FQHC.

Methods

Sample and Sampling Procedures The sample used in this study was comprised of patients who received medical and behavioral health services at any of several FQHC clinic sites in South Mississippi. Inclusion criteria for this study were as follows: (a) patient at a participating FQHC clinic between May 2014 and October 2017, (b) received services from a social worker at the FQHC, and (c) at least 18 years of age or older at the time services were provided. A total of 3,272 patients were included in this study. During their first encounter with the social worker, patients were assessed using various standardized and established instruments to determine their baseline medical and mental-health status. Patients’ responses to the assessments were recorded into a HIPPA-compliant, web-based database, which was designed to allow the social work clinician to minimize scoring and input errors and monitor patient outcomes. All patients provided consent for their respective data to be used for research purposes. All data received by researchers were de-identified. The project was approved by the University of Southern Mississippi Institutional Review Board in accordance with Federal Drug Administration regulations (21 CFR 26, 111), Department of Health and Human Services regulations (45 CFR Part 46), and university guidelines (date of approval: February 26, 2016). Measures Depression. The nine-item Patient Health Questionnaire (PHQ-9) was used to measure patients’ severity of depression (Kroenke, Spitzer, & Williams, 2001). Each patient was asked to describe the frequency with which each symptom had been present over the past 2 weeks using a 4-point scale that ranged from 0 (“not at all”) to 3 (“nearly every day”). The nine items were summed to form a total score ranging from 0 to 27, with higher scores representing greater depressive symptoms. Scores of 10 or higher indicated the possible presence of clinical depression (Kroenke et al., 2001). Past studies have shown the PHQ-9 to have good internal consistency and criterion and construct validity (Arroll et al., 2010; Kroenke et al., 2001). Cronbach’s alpha calculated in the current study was 0.86. Physical health. The five-item subscale of the Duke Health Profile (The DUKE) – Physical Health was used to measure the level of physical health (Parkerson, Broadhead, & Tse, 1990). Items assessed somatic symptoms (2 items), pain (1 item), and ambulation (2 items). For each of the five statements, respondents were asked to choose the frequency with which that symptom had affected them using a scale ranging from 0 (“a lot”) to 2 (“none”). The raw scores were then multiplied by 10 to become 0, 10, and 20, respectively (Parkerson et al., 1990). The five values



were then summed to find the overall physical health score, which ranged from 0 to 100 with higher scores indicating better physical health. The DUKE – Physical Health scale has been shown to have acceptable internal consistency and construct validity (Parkerson et al., 1990). Cronbach’s alpha calculated in the current study was 0.68. Traumatic life events. The frequency of exposure to different traumatic events was measured by a nine-item lifetime traumatic events checklist: natural disaster, life disrupted by Deepwater Horizon oil spill, emotional abuse, sexual abuse, physical abuse, domestic violence, illness/medical trauma, war, and serious injuries. Patients were prompted to identify any and all of the negative life events that they had experienced throughout their lives. Events that had not happened to that particular subject were scored as 0. Thus, each event was given an equal weight. The number of positive responses was summed, and scores ranged from 0 to 9; higher scores indicated more traumatic life events. Current stressors. The number of current stressors was measured by a stressor checklist. Patients were prompted to identify stressors that were currently causing stress, such as strained relationships, legal problems, or health issues. Currently experienced stressors were scored as 1, while those that were not being experienced were scored as 0. Each stressor was given equal weight. The number of positive responses was summed, and scores ranged from 0 to 23; higher scores indicated more current stressors. Self-esteem. Patient self-esteem was assessed using the Duke Health Profile (The DUKE) – Self-Esteem scale, which was comprised of five items that measured patients’ perception of themselves (e.g., I like who I am; I am comfortable being around people; Parkerson et al., 1990; Parkerson, Broadhead, & Tse, 1991). Patients were prompted to select the most accurate response from two response options that included “Yes that describes me exactly” or “No that doesn’t describe me at all.” The raw scores were multiplied by 10 to become 0, 10, and 20 (Parkerson et al., 1990). The five values were then summed to find the overall self-esteem score, which ranged from 0 to 100 with higher scores indicating better self-esteem. The DUKE – Self-Esteem has been shown to have acceptable internal consistency and construct validity (Parkerson et al., 1990). Cronbach’s alpha calculated in the current study was 0.64. BMI group. Obesity and BMI groups were measured using the Body Mass Index (BMI), which is calculated by dividing an individual’s weight in kilograms by the square of height in meters (i.e., kg/m2; CDC, 2016). A BMI of 30.0 or higher falls within the obese category. For this study, the entire sample was divided into five categories based on BMI scores: underweight, <18.5; normal weight, 18.5–24.9; overweight, 25–29.9; obese, 30–34.9; extremely obese, > 35.0.



Statistical Analysis Plan Descriptive analyses of the socio-demographic variables as well as potential risk factors of depression and obesity were conducted to explore the status of the sample. Additional descriptive analyses for PHQ-9 scores and BMI variables were performed to examine rates of depression (PHQ-9 ≥ 10) and BMI groups. Secondly, the relationship between each independent variable and two dependent variables, depression and BMI group, was examined at the bivariate level with a significance level at 0.05. For the BMI group variable, the chi-square test of independence was performed for categorical variables, while the ANOVA test was used for continuous variables. To explore which variables were significantly associated with depression at the bivariate level, the chi-square independent test and t-test were conducted with a significance level at 0.05. The Mantel Haenszel chi-square test was performed when both variables were ordinal-scale variables. If the equal variance assumption was violated, the Satterthwaite t-test was performed. Lastly, a multiple logistic regression analysis was performed to identify the variables that can be used to predict depression as well as how those predictors contribute to the presence of depression. The likelihood ratio test (LR test) and the Hosmer-Lemeshow test (HL test) were performed to determine the degree of model fit and predictability. SAS version 9.3 was used for all statistical analyses.

Results

Sample Characteristics A total of 3,272 patients visited the FQHC in South Mississippi from May 1, 2014, to October 31, 2017. These patients were 18 years of age or older at the time of their first visit. Among them, almost 60% were Caucasian, and 32% were African American. About one-third of participants were male (n = 1,080, 34.21%). The average age of the participants was 47.45 years (SD = 12.72). Almost 70% of the cohort had a high school diploma or less, while only 6% of the total sample had a college degree or higher. More than 24% of the cohort lived alone, whereas around 70% of them lived with family (59.8%) or with friends (9.7%). Among the cohort, 31.6% were married, 40.4% were single, and more than 37% were divorced or widowed. Table 1 presents descriptive statistics of sample characteristics.



Table 1 Descriptive Statistics of the Study Sample (N = 3,272)

Variable N (%) Mean (SD) Race

Caucasian African American

Other

1,863 (59.52%) 1,009 (32.24%)

400 (12.22%)

Gender Female

Male

2,077 (65.79%) 1,080 (34.21%)

Age 47.45 (12.72) Education

Some high school High school diploma

2-year college 4-year college

More than 4-year college

521 (25.60%) 885 (43.49%) 503 (24.72%)

97 (4.77%) 29 (1.43%)

Living Alone

With family With friends

Foster care/shelter Homeless

Other

685 (24.10%)

1,686 (59.82%) 275 (9.68%) 29 (1.02%) 49 (1.72%)

118 (4.15%)

Marital status Single

Married Divorced Widowed

1,049 (40.39%)

821 (31.61%) 574 (22.10%) 153 (5.89%)

Table 2 presents descriptive statistics of study variables. More than 56% of the participants were either obese (22.8%) or extremely obese (33.8%), while 20% of participants were either underweight (1.7%) or normal weight (18.6%). Almost 38% of participants showed signs of depression (PHQ-9 ≥ 10). Of those who fell within the obese category, 36.4% showed signs of depression. Interestingly, higher prevalence of depression was found in participants who were underweight (51.5%) or normal weight (46.2%). Fifty-two percent of those who were underweight and 46% of those who were normal weight showed signs of depression, while 38% of those who were overweight, 34% of those who were obese, and 38% of those who were extremely obese showed signs of depression. Figure 1 shows the prevalence of depression by BMI group. The underweight groups had more frequency of depression. And, as severity of obesity increased from normal to overweight to obese, the percentages of depression decreased.



Table 2 Descriptive Statistics of Study Variables (N = 3,272)

Variable N (%) Mean (SD) Obesity group

Underweight (BMI < 18.5) Normal (18.5 ≤ BMI < 25.0)

Overweight (25.0 ≤ BMI < 30.0)

Obese (30.0 ≤ BMI < 35.0)

Extremely obese (35.0 ≤ BMI)

33 (1.73%)

355 (18.62%) 431 (22.60%) 439 (23.02%) 649 (34.03%)

Depression Yes No

1,243 (37.99%) 2,029 (62.01%)

Diabetes Yes

Not known

523 (15.98%)

2,749 (84.02%)

Smoking Yes No

177 (5.41%)

3,095 (94.59%)

Self-esteem 67.98 (25.06) Physical health 39.36 (26.53) Life stressors 3.24 (2.95) Traumatic events 2.50 (1.99)



Figure1. Prevalence of depression by BMI group. The underweight groups had more frequency of depression. As severity of obesity increased from normal to overweight to obese, the percentages of depression decreased. There were few patients in the smoking group (5.41%), and around 16% of patients reported having a diabetes diagnosis (15.98%). The average physical health score was 39.36 (SD = 26.5), and the average self-esteem score was 67.98 (SD = 25.1). The average number of life stressors was above 3, whereas the average number of traumatic life events was 2.5. Bivariate Analyses Based on the results, depression and BMI groups were statistically significantly related to each other (χ2 = 15.97, p = .03). The BMI group was significantly related to the following variables: race, gender, age, self-esteem, perceived health, number of life stressors, and physical health scores at the 0.05 level of significance. Table 3 shows the bivariate analyses results. Another set of bivariate analyses was performed between depression and other independent variables to find confounding factors that would be effective for both main variables (depression and obesity) at the multivariate analyses. Results revealed that depression was significantly related to the following variables: race, gender, age, smoking, number of traumatic life events, number of life stressors, perceived health, self-esteem, and physical health scores.

0

10

20

30

40

50

60

70

Underweight Normal Overweight Obese ExtremeObese

Not depressed Depressed



Table 3 Bivariate Analyses Between Obesity and Potential Risk Factors (N = 3,272)

Variable 1 Variable 2 Test Test Statistic

p

BMI group (5 groups) Depression χ2 15.97 0.003 BMI group Race χ2 40.34 < .0001

Gender χ2 19.01 0.001 Age F 5.45 < .0001 Living status χ2 28.91 0.090 Education MH* 1.96 0.162 Marital status χ2 16.43 0.173 Diabetes χ2 2.28 0.685 Smoking χ2 2.77 0.596 Self-esteem F 3.22 0.012 Perceived health F 8.45 < .0001 Physical health score F 6.93 < .0001 Life stressors F 5.22 < .0001 Traumatic events F 0.56 0.693

Depression Race χ2 104.53 < .0001 Gender χ2 55.79 < .0001 Age t-test 8.28 < .0001 Living status χ2 0.18 0.674 Education χ2 6.82 0.146 Marital status χ2 5.97 0.113 Diabetes χ2 2.87 0.090 Smoking χ2 8.00 0.005 Self-esteem t-test** 23.28 < .0001 Perceived health t-test** 6.31 < .0001 Physical health score t-test** 20.03 < .0001 Life stressors t-test** 14.42 < .0001 Traumatic events t-test** 11.99 < .0001

*Mantel Haenszel chi-square test was performed because both variables are ordinal scale variables. **Due to the unequal variance problem, the Satterthwaite t-test was performed. Multivariate Analyses Multiple logistic regression analyses were conducted to examine the relationship between BMI group (i.e., underweight, normal, overweight, obese, and extremely obese) and depression, while controlling for confounding factors. Based on the results of bivariate analyses, the following were included in the multivariate analyses as control variables: demographic variables (i.e., age, gender, and race/ethnicity), health-related variables (i.e., smoking, physical health, perceived health), life events variables (i.e., traumatic events, life stressors), and self-esteem. The global



model likelihood ratio test indicated that the overall model was statistically significant (χ2 = 479.99, df = 11, p < .001). The Hosmer and Lemeshow goodness-of-fit test results showed that this model has a good fit for prediction (χ2 = 4.13, df = 8, p = 0.845). Race, smoking, and perceived health were not statistically significant at α = 0.05. Considering all of the potential confounding factors, the BMI group was statistically significant with depression. As BMI changed from underweight to normal to overweight, the chance of depression decreased 30% (OR = 0.70, 95% CI: 0.59 – 0.83, p < .001). Age had a negative relationship with depression, which means that a participant in this population will have lower odds of depression when he or she is older (OR = 0.95, 95% CI: 0.94–0.97, p < .001). If a participant was male, he had almost 50% less odds of being depressed (OR = 0.51, 95% CI: 0.34–0.75, p < .001). For the life events variables, with each additional traumatic event, the odds of depression increased more than 14% (OR = 1.14, 95% CI: 1.04–1.26, p = .008). Also, with one additional life stressor, the odds of depression increased by more than 13% (OR = 1.13, 95% CI: 1.05–1.23, p = .002). The physical health score was statistically significant with depression. If the physical health score went up one unit, the odds of depression went down 4.3% (OR = 0.96, 95% CI: 0.95–0.97, p < .001). Self-esteem score had a negative relationship with depression. If the self-esteem score went up one unit, then the odds of being depressed went down 3.6% (OR = 0.96, 95% CI: 0.96–0.97, p < .001). Table 4 presents results of the multiple logistic regression analyses.

Table 4 Multiple Logistic Regression Analyses Results for the Dependent Variable, Depression

Variable Β Odds (95% CL) p Age -0.0500 0.943 (0.920 – 0.967) < 0.001 Gender -0.6797 0.507 (0.341 – 0.754) 0.001 Race

Caucasian vs. other African American vs. other

0.2749

-0.0965

1.574 (0.828 – 2.989) 1.085 (0.549 – 2.144)

0.049 0.533

Smoking status -0.1527 0.858 (0.482 – 1.530) 0.605 Self-esteem -0.0367 0.964 (0.955 – 0.973) < 0.001 Perceived health 0.0014 1.001 (0.997 – 1.006) 0.532 Physical health score -0.0438 0.957 (0.948 – 0.966) < 0.001 Life stressors 0.1250 1.133 (1.046 – 1.227) 0.002 Traumatic events 0.1338 1.143 (1.035 – 1.262) 0.008 BMI group -0.3591 0.698 (0.591 – 0.825) < 0.001



Discussion

This study examined the association between obesity and depression among patients at multiple sites of an FQHC in South Mississippi. Fifty-seven percent of the respondents fell within the obese or extremely obese categories. This is considerably higher than Mississippi’s overall adult obesity rate (37.3%; The State of Obesity, 2016). Of those who fell within the obese category, 36.4% showed depression. Interestingly, although this figure is quite high, even greater prevalence rates of depression were found among participants who were underweight (51.5%) or normal weight (46.2%). The association between the severity of obesity (i.e., underweight, normal, overweight, obese, extremely obese) and depression was further examined at the multivariate level, while adjusting for potential confounders including socio-demographic variables, physical health, self-esteem, life stressors, and traumatic events. Results revealed that there was a significant and negative relationship between BMI group and depression, which means that higher BMI was associated with a lower risk of depression. This finding is inconsistent with previous studies (Dearborn et al., 2018; de Wit et al., 2010; de Wit et al., 2009; Hung et al., 2014; Luppino et al., 2010; John et al., 2005; Mannan et al., 2016; Pereira-Miranda et al., 2017; Xiang & An, 2015). For example, a meta-analysis study regarding the association between obesity and depression reported that those considered obese were more likely to show symptoms of depression (de Wit et al., 2010). The population included in that study, however, was the general population; our study included only individuals who received services from an FQHC. Individuals receiving services from FQHCs have lower socioeconomic status and are more likely to be uninsured and have Medicaid than the general public (Nath, Costigan, & Hsia, 2016). Moreover, our study population had a uniquely high prevalence of obesity, even by Mississippi standards. Our findings suggest social normalization of obesity may influence the prevalence of depression. The association between obesity and depression may differ by various factors such as gender, race, ethnicity, and region (Assari, 2014). Future studies should examine how these factors play a role in the relationship between obesity and depression. Of the covariates examined in this study, being female, being younger, having more frequent exposure to traumatic events and stressors, having lower levels of physical health, and having lower self-esteem were associated with increased likelihood of depression. These findings are supported by existing literature. Younger age and being female have been found to increase the chances of depression (Brody et al., 2018; Gonzalez et al., 2010; Heo et al., 2008; NIMH, 2017; USPSTF, 2016). Stressful life events have been linked to the development of depression, and physical health scores have been found to be worse among depressed individuals compared to non-depressed individuals (Phillips, Carroll, & Der, 2015; Verma et al., 2010). Self-esteem has been found to have a negative association with depression (Orth, Robins, Meier, & Conger, 2016).



Implications for Practice and Policy The findings from this study are at odds with previous investigations that identified a correlation between obesity and depression (Luppino et al., 2010). As previously mentioned, samples used in these other studies were drawn from the general population. Our study only included patients receiving services from an FQHC located in South Mississippi. This population is comprised of lower socioeconomic status individuals, many of whom are either uninsured or underinsured. In addition, Mississippi has the second highest obesity rate in the nation at 37.3% (The State of Obesity, 2016). The difference in results between our sample and others may be found in the way the individuals and their communities view obesity. Studies have revealed that body image is negatively impacted as weight increases (Schwartz & Brownell, 2004). However, as the rate of obesity in our study sample is higher than the general population, our sample may not have experienced the discrimination or stigma reported in previous investigations that identified a link between obesity and psychological distress (Carr & Friedman, 2005). Past research has also shown that race/ethnicity and culture have an effect on body satisfaction and body size, particularly among African American people, who have reported greater acceptance of larger body sizes compared to other racial/ethnic groups (Caprio et al., 2008; Powell & Kahn, 1995). Our results underscore the need to explore the role of cultural norms in body image. It is possible that our sample views obesity in a more favorable light than the general population. It has been argued that depression can lead to obesity. This idea developed based on studies connecting physical activity and unhealthy eating habits to depression (Adamson, Yang, & Motl, 2016; Paans et al., 2018). Such factors clearly influence weight. It is possible, however, that the rate of obesity in our study population is related to a lack of food choices in low-income communities and not a result of depressive symptoms. For example, a 2009 study revealed a link between food choice and income level, as women purchasing food for their families believed that healthy food was unaffordable (Dammann & Smith, 2009). This would suggest that other factors might have more important roles in determining individual weight. Our results indicate that depression is not as prevalent among obese patients in low-income communities. This information highlights the need for health-education efforts and interventions to focus on influencing changes within communities as well as addressing the medical and emotional needs of individuals. This finding highlights the need to fund obesity prevention and intervention efforts at the community level, especially in low-income communities. Limitation This is cross-sectional research, so we cannot infer a causal relationship between the severity of obesity and depression symptoms. We have a participation bias since the study sample voluntarily visited the clinics for health problems. To determine obesity more clearly, it would be



better to obtain more information such as waist circumference or body fat percent. The general population of South Mississippi might show a relationship between obesity and depression that is more in line with the results of previous research.

References

Adamson, B. C., Yang, Y., & Motl, R. W. (2016). Association between compliance with

physical activity guidelines, sedentary behavior and depressive symptoms. Preventive Medicine, 91, 152–157. doi:10.1016/j.ypmed.2016.08.020

Arroll, B., Goodyear-Smith, F., Crengle, S., Gunn, J., Kerse, N., Fishman, T., & Hatcher, S. (2010). Validation of PHQ-2 and PHQ-9 to screen for major depression in the primary care population. The Annals of Family Medicine, 8(4), 348–353. doi:10.1370/afm.1139

Assari, S. (2014). Association between obesity and depression among American Blacks: Role of ethnicity and gender. Journal of Racial and Ethnic Disparities, 1, 36–44. doi:10.1007/s40615-014-0007-5

Brody, D. J., Pratt, L. A., & Hughes, J. P. (2018). Prevalence of depression among adults aged 20 and over: United States, 2013–2016. NCHS Data Brief No. 303. Retrieved from https://www.cdc.gov/nchs/data/databriefs/db303.pdf

Caprio, S., Daniels, S. R., Drewnowski, A., Kaufman, F. R., Palinkas, L. A., Rosenbloom, A. L., & Schwimmer, J. B. (2008). Influence of race, ethnicity, and culture on childhood obesity: Implications for prevention and treatment. Diabetes Care, 31(11), 2211–2221. doi:10.2337/dc08-9024

Carr, D., & Friedman, M. A. (2005). Is obesity stigmatizing? Body weight, perceived discrimination, and psychological well-being in the United States. Journal of Health and Social Behavior, 46(3), 244–25.

Centers for Disease Control and Prevention [CDC]. (2015). The health effects of overweight and obesity. Retrieved from https://www.cdc.gov/healthyweight/effects/index.html

Centers for Disease Control and Prevention [CDC]. (2016). Defining adult overweight and obesity. Retrieved from https://www.cdc.gov/obesity/adult/defining.html

Chang, H. H., & Yen, S. T. (2012). Association between obesity and depression: Evidence from a longitudinal sample of the elderly in Taiwan. Aging & Mental Health, 16(2), 173–180. http://dx.doi.org/10.1080/13607863.2011.605053

Dammann, K. W., & Smith, C. (2009). Factors affecting low-income women’s food choices and the perceived impact of dietary intake and socioeconomic status on their health and weight. Journal of Nutrition Education and Behavior, 41(4), 242–253.

Dearborn, P. J., Robbins, M. A., & Elias, M. F. (2018). Challenging the “jolly fat” hypothesis among older adults: High body mass index predicts increases in depressive symptoms over a 5-year period. Journal of Health Psychology, 23(1), 48–58. https://doi.org/10.1177/1359105316675209



de Wit, L., Luppino, F., van Straten, A., Penninx, B., Zitman, F., & Cuijpers, P. (2010). Depression and obesity: A meta-analysis of community-based studies. Psychiatry Research, 178(2), 230–235. doi:10.1016/j.psychres.2009.04.015

de Wit, L. M., van Straten, A., van Herten, M., Penninx, B. W., & Cuijpers, P. (2009). Depression and body mass index, a U-shaped association. BMC Public Health, 9(14). doi:10.1186/1471-2458-9-14

Dieleman, J. L., Baral, R., Birger, M., Bui, A. L., Bulchis, A., Chapin, A., … Murray, C. J. (2016). U.S. spending on personal health care and public health, 1996–2013. Journal of the American Medical Association, 316(24), 2627–2646. doi:10.1001/jama.2016.16885

Flegal, K. M., Kruszon-Moran, D., Carroll, M. D., Fryar, C. D., & Ogden, C. L. (2016). Trends in obesity among adults in the United States, 2005 to 2014. Journal of the American Medical Association, 315(21), 2284–2291. doi:10.1001/jama.2016.6458

Gonzalez, O., Berry, J. T., McKnight-Eily, L. R., Strine, T., Edwards, V. J., Lu, H., & Croft, J. B. (2010). Current depression among adults – United States, 2006 and 2008. Morbidity and Mortality Weekly Report. Retrieved from https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5938a2.htm?s_cid=mm5938a2_w

Grant, T., Lott, L., Courtney, C., II., Johnson, X., Sutton, V., & Zhang, L. (2016). Mississippi obesity action plan 2016. Mississippi State Department of Health.

Harvard Law School Mississippi Delta Project. (2014). Mental health in Mississippi: Analysis and recommendations. Retrieved from http://www.deltadirections.com/wp-content/uploads/2013/05/Mental-Health-in-Mississippi-.pdf

Heo, M., Murphy, C. F., Fontaine, K. R., Bruce, M. L., & Alexopoulos, G. S. (2008). Population projection of U.S. adults with lifetime experience of depressive disorder by age and sex from year 2005 to 2050. International Journal of Geriatric Psychiatry, 23(12), 1266–1270. doi:10.1002/gps.2061

Hung, C.-F., Rivera, M., Craddock, N., Owen, M. J., Gill, M., Korszun, A., … McGuffin, P. (2014). Relationship between obesity and the risk of clinically significant depression: Mendelian randomization study. The British Journal of Psychiatry, 205(1), 24–28. https://dx.doi.org/10.1192%2Fbjp.bp.113.130419

John, U., Meyer, C., Rumpf, H.-J., & Hapke, U. (2005). Relationships of psychiatric disorders with overweight and obesity in an adult general population. Obesity Research, 13(1), 101–109.

Kim, D. D., & Basu, A. (2016). Estimating the medical care costs of obesity in the United States: Systematic review, meta-analysis, and empirical analysis. Value in Health, 19, 602–613. http://dx.doi.org/10.1016/j.jval.2016.02.008

Kolbo, J. R., Zhang. L, Werle, N., Molaison, E. F., Harbaugh, B., Kirkup, M., & Walker, E. (2016). Overweight and obesity prevalence and trends among Mississippi public school students: A decade of data between 2005 and 2015. Journal of the Mississippi State Medical Association, 57(10), 1–9.



Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613.

Lépine, J.-P., & Briley, M. (2011). The increasing burden of depression. Neuropsychiatric Disease and Treatment, 7(Suppl 1), 3–7. doi: 10.2147/NDT.S19617

Luppino, F. S., de Wit, L. M., Bouvy, P. F., Stijnen, T., Cuijpers, P., Penninx, B. W., & Zitman, F. G. (2010). Overweight, obesity, and depression: A systematic review and meta-analysis of longitudinal studies. Archives of General Psychiatry, 67(3), 220–229. doi:10.1001/archgenpsychiatry.2010.2

Mannan, M., Mamun, A., Doi, S., & Clavarino, A. (2016). Prospective associations between depression and obesity for adolescent males and females: A systematic review and meta-analysis of longitudinal studies. PLoS ONE, 11(16). doi:10.1371/journal.pone.0157240

Mendy, V. L., Vargas, R., Cannon-Smith, G., & Payton, M. (2017). Overweight, obesity, and extreme obesity among Mississippi adults, 2001–2010 and 2011–2015. Preventing Chronic Disease, 14. https://doi.org/10.5888/pcd14.160554

Mental Health America. (2018). Mental health in America – Access to care data. Retrieved from http://www.mentalhealthamerica.net/issues/mental-health-america-access-care-data

Nath, J. B., Costigan, S., & Hsia, R. Y. (2016). Changes in demographics of patients seen at Federally Qualified Health Centers, 2005–2014. JAMA Internal Medicine, 176(5), 712–714.

National Center for Health Statistics. (2017). NCHS fact sheet: National Health and Nutrition Examination Survey. Retrieved from https://www.cdc.gov/nchs/data/factsheets/factsheet_nhanes.pdf

National Institute of Mental Health [NIMH]. (2017). Major depression. Retrieved from https://www.nimh.nih.gov/health/statistics/major-depression.shtml

Nigatu, Y. T., Reijneveld, S. A., de Jonge, P., van Rossum, E., & Bültmann, U. (2016). The combined effects of obesity, abdominal obesity, and major depression/anxiety on health-related quality of life: The LifeLines cohort study. PLoS ONE, 11(2). doi:10.1371/journal.pone.0148871

Orth, U., Robins, R. W., Meier, L. L., & Conger, R. D. (2016). Refining the vulnerability model of low self-esteem and depression: Disentangling the effects of genuine self-esteem and narcissism. Journal of Personality and Social Psychology, 110(1), 133–149. http://dx.doi.org/10.1037/pspp0000038

Paans, N. P., Bot, M., Brouwer, I. A., Visser, M., Roca, M., Kohls, E., & Penninx, B. W. (2018). The association between depression and eating styles in four European countries: The MooDFOOD prevention study. Journal of Psychosomatic Research, 108, 85–92. doi:10.1016/j.jpsychores.2018.03.003

Pan, A., Lucas, M., Sun, Q., van Dam, R. M., Franco, O. H., Manson, J. E., … Hu, F. B. (2010). Bidirectional association between depression and type 2 diabetes in women. Archives of Internal Medicine, 170(21), 1884–1891. doi:10.1001/archinternmed.2010.356



Parkerson, G. R., Jr., Broadhead, W. E., & Tse, C. K. (1990). The Duke Health Profile: A 17-item measure of health and dysfunction. Medical Care, 28(11), 1056–1072.

Parkerson, G. R., Jr., Broadhead, W. E., & Tse, C. K. (1991). Development of the 17-item Duke Health Profile. Family Practice, 8(4), 396–401.

Pereira-Miranda, E., Costa, P. R., Queiroz, V. A., Pereira-Santos, M., & Santana, M. L. (2017). Overweight and obesity associated with higher depression prevalence in adults: A systematic review and meta-analysis. Journal of the American College of Nutrition, 36(3), 223–233. doi:10.1080/07315724.2016.1261053

Phillips, A. C., Carroll, D., & Der, G. (2015). Negative life events and symptoms of depression and anxiety: Stress causation and/or stress generation. Anxiety Stress Coping, 28(4), 357–371. doi:10.1080/10615806.2015.1005078

Powell, A. D., & Kahn, A. S. (1995). Racial differences in women’s desires to be thin. International Journal of Eating Disorders, 17(2), 191–195.

Qobadi, M., & Payton, M. (2017). Racial disparities in obesity prevalence in Mississippi: Role of socio-demographic characteristics and physical activity. International Journal of Environmental Research and Public Health, 14(258). doi:10.3390/ijerph14030258

Segal, L. M., Rayburn, J., & Beck, S. E. (2017). The state of obesity: Better policies for a healthier America 2017. Retrieved from https://stateofobesity.org/wp-content/uploads/2018/08/stateofobesity2017.pdf

Schwartz, M. B., & Brownell, K. D. (2004). Obesity and body image. Body image, 1(1), 43–56. The State of Obesity. (2016). Adult obesity in the United States. Retrieved from

https://stateofobesity.org/adult-obesity/ University of Mississippi Medical Center [UMMC]. (n.d.). Obesity in Mississippi. Retrieved

from https://www2.umc.edu/Administration/Centers_and_Institutes/Mississippi_Center_for_Obesity_Research/Obesity_in_Mississippi.aspx

U.S. News & World Report. (2018). Best states 2018: Ranking performance throughout all 50 states. Retrieved from https://media.beam.usnews.com/72/92/aae484e24a139c33fd2658193683/171206-best-states-overall-rankings-2018.pdf

U.S. Preventive Services Task Force [USPSTF]. (2016). Final recommendation statement: Depression in adults: Screening. Retrieved from https://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/depression-in-adults-screening1

Verma, S. K., Luo, N., Subramaniam, M., Sum, C. F., Stahl, D., Liow, P. H., & Chong, S. A. (2010). Impact of depression on health related quality of life in patients with diabetes. Annals of the Academy of Medicine, Singapore, 39(12), 913–919.

Williams, E. P., Mesidor, M., Winters, K., Dubbert, P. M., & Wyatt, S. B. (2015). Overweight and obesity: Prevalence, consequences, and causes of a growing public health problem. Current Obesity Reports, 4, 363–370. doi:10.1007/s13679-015-0169-4



World Health Organization [WHO]. (2018a). Depression. Retrieved from http://www.who.int/mediacentre/factsheets/fs369/en/

World Health Organization [WHO]. (2018b). Obesity and overweight. Retrieved from http://www.who.int/mediacentre/factsheets/fs311/en/

Xiang, X., & An, R. (2015). Obesity and onset of depression among U.S. middle-aged and older adults. Journal of Psychosomatic Research, 78, 242–248. http://dx.doi.org/10.1016/j.jpsychores.2014.12.008

Author Note

Hwanseok Choi, Department of Public Health, University of Southern Mississippi; Joohee Lee, School of Social Work, University of Southern Mississippi; Stephanie T. McLeod, School of Social Work, University of Southern Mississippi; Rambod A. Rouhbakhsh, Family Medicine Resident, Forrest General Hospital; Michelle Brazeal, School of Social Work, University of Southern Mississippi; Tim Rehner, School of Social Work, University of Southern Mississippi; and David M. Cochran, Department of Geography and Geology, University of Southern Mississippi. Correspondence concerning this article should be addressed to Hwanseok Choi, Department of Public Health, University of Southern Mississippi, 118 College Drive #5122, Hattiesburg, MS 39406-0001. E-mail: [email protected]

FMLA Role in Mothers’ Remaining in the Workforce


Remaining in the Workforce After Motherhood: Does the Family Medical Leave Act Play a Role in the Decisions of Mississippi Mothers?

Jennifer Balcazar, Danielle Fastring, and Avery Hilbert University of Southern Mississippi

Background: Increased maternity leave has been shown to have a positive impact on maternal and child health, and to increase the length of time mothers breastfeed their infants. After childbirth, working women must decide if and when they will return to the workforce. Purpose: To determine the impact of current U.S. family leave policies on Mississippi mothers’ decisions to return to work after the birth of their first child. Methods: A survey was developed to collect information about factors influencing mothers’ decisions to return to work after the birth of their first child. The survey collected study eligibility information, demographics, education and income level, length of maternity leave taken, breastfeeding practices, and household composition from participants. Additionally, attitudes toward U.S. family leave polices and their impacts were assessed. Results: Mississippi mothers were negatively impacted financially by their first pregnancies. Many participants were not eligible for FMLA because they worked for small companies or lacked accrued employment time. Conclusion: The majority of women’s maternity leave was limited by the amount of paid time off they had accrued on their jobs. Most would have taken longer maternity leave if they had they been financially able to do so. Keywords: FMLA, maternity leave, working mothers, breastfeeding, maternal child health



Introduction In today’s society, more women are participating in the workforce, a trend that began during the middle of the 20th century (Aitken et al., 2015). Working women who plan to have a family must decide whether or not they will work during pregnancy and if they will return to the workforce after delivery. It is estimated that 67% of American mothers worked during their first pregnancy; of this group, 87% worked into their last trimester, and the majority of women worked full time (Guendelman et al., 2009).Various factors such as educational level, career experience, job security, and income level influence women’s decision to return to the workforce after having a child (Vahratian & Johnson, 2009). The decision to return to the workforce is also driven by the amount of parental leave available after delivery. Parental leave is the allotted time off from work for mothers and/or fathers after the birth of an infant (Ruhm, 2000). Parental leave can consist of paid or unpaid leave. The Family and Medical Leave Act (FMLA), which was passed in 1993, allows parents to take up to 12 weeks of unpaid, job-protected leave around the birth of a child or to provide family care if they meet eligibility requirements. Eligibility requirements consist of having worked at least 1,250 hours as a full-time employee within the past 12 months in a business that has 50 or more employees within a 75-mile radius ( Guendelman, Goodman, Kharrazi, & Lahiff, 2014; Raabe & Theall, 2016; Rossin, 2011). Smaller companies with fewer than 50 employees do not have to participate. Paid parental leave is a form of income provided by either the employer or government to replace and compensate the employee during leave around the birth or adoption of a child (Mariskind, 2017). Paid leave usually consists of accrued time off in the form of paid vacation days or paid sick leave. Thus, women who lack these employer-provided benefits would not have any paid maternity leave. Those who work for companies that do not participate in FMLA, or lack personal eligibility with regard to time worked, would not be eligible for unpaid leave, nor would they have the guarantee of their previous position when returning from leave. Returning to work quickly after childbirth is another increasing trend in the United States (Guendelman et al., 2014). This trend is concerning because longer maternity leaves are positively associated with positive birth outcomes and increases in both maternal and child health after delivery. A recent study concluded that shorter maternity leaves were associated with poor maternal and infant health outcomes, whereas more than 12 weeks’ leave was shown to improve the health of the infant and mother (Vahratian & Johnson, 2009; Ruhm, 2000). One cohort study showed that mothers who took more than 12 weeks’ maternity leave experienced less severe depression regardless of whether the leave was paid or unpaid (Chatterji & Markowitz, 2012). Mothers who received paid maternity leave had better overall physical and mental health than those who did not receive paid leave (Aitken et al., 2015; Cools, Fiva, & Kirkebøen, 2015). In addition, maternity leave during the last trimester has been associated with a reduction in cesarean deliveries and pre-term birth (Guendelman et al., 2014).



Breastfeeding habits also are positively associated with increased maternity leave. Mothers who took time off from work in the form of maternity leave were twice as likely to breastfeed as were mothers who did not take maternity leave (Aitken et al., 2015). Paid parental leave has been shown to increase the likelihood that women will attempt to breastfeed and the duration of breastfeeding. A study conducted in California, one of only five states in the U.S. with paid parental leave, found that women who had paid maternity leave breastfed twice as long as mothers who did not take leave (Appelbaum & Milkman, 2011). Breastfeeding for 6 months, with or without supplementation, has a protective effect against asthma by delaying or preventing the onset altogether (Abarca, Garro, & Pearlman, 2018). Further, infants who are breastfed have reduced rates of obesity, type 2 diabetes, and ear and respiratory infections (Centers for Disease Control and Prevention, 2018) In 2018, New York and Hawaii became the fourth and fifth states, respectively, in the U.S. to offer paid parental leave as a state benefit. Studies conducted in the other three states with established programs (Rhode Island, California, and New Jersey) found that paid parental leave for mothers led to increased job security and commitment and greater home stability when compared to mothers who did not have a paid leave option (Raabe & Theall, 2016). A 2011 study showed that participants who used the California Paid Family Leave program, which is funded by an employee-paid payroll tax, noticed an increase in productivity, morale, and performance of their employees (Lewis, Stumbitz, Miles, & Rouse, 2014). Mississippi does not currently offer paid parental leave as a state benefit. Although unpaid leave is available through the FMLA, only 38% of Mississippians are eligible. The remaining 62% of Mississippians (a) work for a company that has fewer than 50 employees, (b) have not met the required length of time on the job or hours worked, or (c) would see their family income drop to or below 200% of the federal poverty level if they took 12 weeks of unpaid leave (diversitydatakids.org, 2015). In 2016 and 2017, Mississippi experienced the highest infant mortality rates (death in the first 365 days of life) in the nation at 9.3 and 8.8 infant deaths per 1,000 live births, respectively (MSDH, 2016; MSDH, 2017). The state also experiences high rates of pre-term birth and low infant birthweight, two of the main contributors to infant mortality, statewide. The purpose of this study was to identify the factors that influenced Mississippi mothers to return to work or leave the workforce after the birth of their first child. Determinants explored include education and income level, length of maternity leave taken, breastfeeding practices, and household composition. Additionally, attitudes toward U.S. family leave polices and their impacts were assessed.



Methods In order to determine the impact of current U.S. family leave policies on mothers’ decision to return to work after the birth of their first child, a cross-sectional descriptive study was conducted. A comprehensive survey was developed consisting of approximately 35 questions that determined study eligibility and collected information on demographics, education and income level, length of time taken off before returning to work after the first child was born, breastfeeding practices, and household composition. Additionally, attitudes toward U.S. family leave polices and their impacts were assessed. Survey questions can be found in Figure 1. The survey was distributed online via Qualtrics (https://www.qualtrics.com/) and took approximately 20 minutes to complete. The study design and survey materials were approved by the University of Southern Mississippi Institutional Review Board. Informed consent was obtained from all participants prior to beginning the survey. In order to be eligible to participate, respondents had to be women between 18 and 60 years old, have at least one biological child at home, and reside in Mississippi. Eligibility requirements were confirmed through survey response. Potential participants were recruited via an informational card containing a link to the study survey. The card was distributed to local major retailers, childcare facilities, and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) Distribution Centers. Upon completion of the survey, participants were given an opportunity to be entered in a random drawing to win one of two $25 gift certificates as an incentive for participating in the survey. Participants were provided an email address to contact study staff directly if any questions arose during the survey or if clarification was needed. All data were analyzed using SPSS Version 23 (IBM Corp, 2015). Descriptive statistics for categorical variables were produced by calculating frequencies and percents in each category for relevant variables. Longitudinal variables were assessed for normality, and the mean and standard deviation or the median and range were reported as appropriate.

Results

Participants in the study included 309 adult mothers in Mississippi who had at least one biological child at home. Table 1 lists the demographic characteristics of participants. They ranged in age from 18 to 60 years old (�̅� = 33.5 ± 8.04). Almost half of women in the study were between the ages of 25 and 34 years old (48.2%, n=149). The majority of participants were Caucasian (88%, n=277) and married (77%, n=238). The most frequently reported level of education achieved was a bachelor’s degree (28.2%, n=87). The survey was open to residents in



all Mississippi counties. Table 2 shows the counties represented in the study. The largest pool of participants resided in Harrison County (35.3%, n=109).



Figure 1. Comprehensive survey to determine study eligibility and collect information on demographics, education and income level, length of time taken off before returning to work after the first child was born, breastfeeding practices, and household composition.

Eligibility Questions 1. Are you female? 2. Are you currently 18 years of age or older? 3. Do you have at least one biological child at home? 4. What is your date of birth? Demographic Questions 5. Please select the race you most identify with through biology or culture. 6. What is the highest level of school that you have completed? 7. Are you currently enrolled in school or a training / vocational program? 8. What is your marital status? 9. Which state do you reside in? 10. If you live in Mississippi: Which county do you live in? Household Composition 11. How many biological/ adopted/ step children/ other children do you have living with you currently? 12. Next, think about how many adults (18 and older) are currently living in your home. Think about their relationship to you. Are there

other adults living in your home besides yourself? How many adults live in your home that you would call your spouse, significant other, or partner? How many adults live in your home that you would call your parent? How many adults live in your home that you would call a relative that you have not yet counted? How many adults live in your home besides yourself that you have not yet counted?

Employment/Leave Questions 13. Now we will ask questions about your history in the workplace. We are trying to determine how having children impacted your ability

to stay in the workforce. Think back to the time BEFORE you become pregnant with your first biological child. Were you employed? 14. Thinking back to just before your FIRST pregnancy, when you consider all the locations where your employer operated, what was the

total number of people who work there? 15. How long had you been in that job prior to becoming pregnant with your first child? 16. Did you stop work (take time off) to have your first child? 17. How long into your pregnancy did you work during your FIRST pregnancy? 18. How much maternity/family leave did you take before returning to work after the birth of your first biological child? 19. If you took maternity/family leave around the time of your first pregnancy, how would you best describe that leave? 20. If you took maternity/family leave around the time of your first pregnancy, were you worried that your job might not be available after

you returned from maternity leave? 21. If you took maternity/family leave around the time of your first pregnancy, did your benefits impact the amount of time you took off? 22. Did you return to the workforce after maternity leave for your first child? 23. Did you return to work earlier than you would have liked due to financial reasons? 24. For your first child, did you try to breastfeed at all? 25. For your first child, if you started breastfeeding, when did you stop? 26. Did you stop breastfeeding because of the difficulty of breastfeeding while working? 27. Is there anything that would have made it possible/easier for you to continue breastfeeding/pumping? 28. How old was your first child when you returned to work? 29. Did the father of your first child take paternity leave (time off for fathers) from work around the time of your first pregnancy? 30. How much paternity/family leave did the father of your first child take before returning to work after the birth of your first biological

child? 31. If the father of your first baby took paternity/family leave around the time of your first pregnancy, how would you best describe that

leave? 32. What was your total household income before taxes around the time your first child was born? 33. Is there anything else you would like to tell us about your experience with working and having your FIRST CHILD? 34. What was your total household income before taxes around the time your FIRST child was born?



Table 1 Demographic Characteristics of Participants

Demographic Characteristics Frequency (%)

Age

18–24 41 (13.3%)

25–29 77 (24.9%)

30–34 72 (23.3%)

35–39 59 (19.1%)

40–44 28 (9.1%)

45–49 19 (6.1%)

50–55 10 (3.2%)

56+ 3 (1.0%)

Total 309 (100%)

Mean ± SD 33.50 ± 8.036

Race

Caucasian 272 (88.0%)

African American 25 (8.1%)

Asian or Pacific Islander 4 (1.3%)

American Indian or Alaska Native 2 (0.6%)

Other (please list) 6 (2.2%)

Total 309 (100%)

Marital Status

Single, never married 24 (7.8%)

Living with partner 19 (6.1%)

Married 238 (77.0%)

Widowed 2 (0.6%)

Divorced 21 (6.8%)

Separated 5 (1.6%)

Total 309 (100%)

Highest Level of Education

No school completed 1 (0.3%)

Some high school, no diploma 5 (1.6%)

High school graduate 21 (6.8%)

GED 9 (2.9%)

Some college credits, no degree 64 (20.7%)

Trade, technical, vocational training 16 (5.2%)

Associate’s degree 53 (17.2%)



Table 2 Participants’ Mississippi County of Residence

County Frequency (%) County Frequency (%)

Alcorn 1 (0.3%) Lowndes 1 (0.3%)

Calhoun 2 (0.6%) Madison 4 (1.3%)

Copiah 1 (0.3%) Marion 1 (0.3%)

Desoto 2 (0.6%) Marshall 2 (0.6%)

Forrest 15 (4.9%) Newton 1 (0.3%)

George 13 (4.2%) Panola 1 (0.3%)

Greene 1 (0.3%) Pearl River 5 (1.6%)

Hancock 10 (3.2%) Perry 4 (1.3%)

Harrison 109 (35.3%) Pontotoc 1 (0.3%)

Hinds 7 (2.3%) Rankin 12 (3.9%)

Itawamba 1 (0.3%) Scott 1 (0.3%)

Jackson 61 (19.7%) Stone 4 (1.3%)

Jones 6 (1.9%) Walthall 1 (0.3%)

Lafayette 7 (2.3%) Warren 20 (6.5%)

Lamar 9 (2.9%) Wayne 1 (0.3%)

Lauderdale 5 (1.6%) Total 309 (100.0%)

Table 3 shows data from participants’ responses pertaining to work history and work habits prior to their first pregnancy. As reported by respondents, 70.4% (n=216) of participants were full-time employees prior to their first pregnancy, and 17.3% (53) were part-time employees. Approximately one-fourth (26.7%, n=74) of participants reported that they were employed in their job less than a year, which indicates that they were ineligible to receive FMLA benefits. Further, 30% of employers had fewer than 50 employees, an indicator that they did not have to provide their employees with FMLA benefits. Overall, only one-fourth (30.1%, n=78) of participants met the eligibility requirements to receive FMLA benefits.

Bachelor’s degree 87 (28.2%)

Master’s degree 47 (15.2%)

Doctoral degree 6 (1.9%)

Total 309 (100%)



Participants were asked to answer questions related to the maternity leave they took associated with the birth of their first child (Table 4). Women who reported any employment on the previous question were asked if they had stopped working or had taken time off from work to have their first child. In this study, 87.4% (n=229) answered “yes,” while 12.6% (n = 33) answered “no.” Most participants reported working up until the time they went into labor with their first child (51.3%, n=134) or during their third trimester (31.8%, n=83). Approximately one-fourth of mothers (24.8%, n=65) did not return to the workforce after the birth of their first child. Participants’ length of maternity leave was highly variable. For example, 5.3% of participants reported that their maternity leave was less than 2 weeks long. The most frequently reported length was 6 weeks to 48 days (17.6%, n=46). Most women reported that their available employee benefits impacted the length of their maternity leave. The majority of mothers (60.7%, n=142) would have taken additional leave if they could, and 62.7% (n=158) reported that they returned to work earlier than they would have liked due to financial reasons. Approximately one-third (31.2%, n=78) of mothers returned to work when their infant was between 6 and 8 weeks old. In this cohort, 23.2% (n=60) of mothers had paid maternity leave in the form of accrued paid personal leave or paid leave provided by their employer. Approximately one-third utilized FMLA benefits, and 19.3% (n=50) used a combination of paid leave and FMLA benefits. Mothers were also asked to characterize the paternity leave taken by the infant’s father. A total of 75.2% (n=191) of participants responded that the father did take paternity leave and returned to work after the mother had the baby. Only 13% (n=33) stated that the father was not working during the time of their pregnancy. Approximately 1.2% (n=3) of fathers took paternity leave and did not return to work after the baby was born. Most fathers who did take paternity leave took less than 1 week of leave time (61.8%, n=120) or between 1 week and 13 days of leave time (26.2%, n=51). Table 3 Participants’ Work History and Habits

Participants’ Work History and Habits Frequency (%)

Think back to the time BEFORE you became pregnant with your first live biological child. Were you employed:

Part-time employed 53 (17.3%)

Full-time employed 216 (70.4%)

Out of work & actively looking 3 (1.0%)

Active duty military 7 (2.3%)

Unable to work 8 (2.6%)

Chose not to work 20 (6.5%)

Total 307 (100.0%)



How long had you been employed in that job prior to bringing home your first child?

1 to 3 months 31 (11.2%)

4 to 6 months 20 (7.2%)

7 to 11 months 23 (8.3%)

1 year 46 (16.7%)

2 years 54 (19.6)

3 to 5 years 78 (28.3%)

6 to 9 years 20 (7.2%)

10 years or more 4 (1.4%)

Total 276 (100.0%)

Thinking back to just before your FIRST pregnancy, when you consider all the locations where your employer operated, what was the total number of persons who work there?

2 to 9 20 (7.2%)

10 to 24 45 (16.3%)

25 to 49 18 (6.5%)

50+ employees 193 (69.9%)

Total 276 (100.0%)

Table 4 Characteristics of Mothers’ Maternity Leave and Fathers’ Paternity Leave

Characteristics of Mothers’ Maternity Leave and Fathers’ Paternity Leave Frequency (%)

Maternity Leave

Did you stop work (take time off) to have your first child?

Yes 229 (87.4%)

No 33 (12.60%)

Total 262 (100.0%)

How long into your pregnancy did you work during your FIRST pregnancy?

I stopped working during my first trimester. 20 (7.7%)

I stopped working during my second trimester. 24 (9.2%)

I stopped working during my third trimester. 83 (31.8%)

I stopped working when I went into labor. 134 (51.3%)

Total 261 (100.0%)

How much maternity/family leave did you take before returning to work after the birth of your first biological child?

I did not return to work after having the baby. 65 (24.8%)



Less than a week 5 (1.9%)

1 week to 13 days 9 (3.4%)

2 weeks to 20 days 8 (3.1%)


4 weeks to 34 days 12 (4.6%)


6 weeks to 48 days 46 (17.6%)


8 weeks to 62 days 22 (8.4%)


10 weeks to 76 days 8 (3.1%)

11 weeks to 83 days 1 (0.4%)

12 weeks to 90 days 35 (13.4%)

13 weeks to 98 days 2 (0.8%)

14 weeks to 104 days 1 (0.4%)

15 weeks to 111 days 1 (0.4%)

16 weeks or more 19 (7.3%)

Total 262 (100.0%)

If you took maternity/family leave around the time of your first pregnancy, did your benefits impact the amount of time you took off?

Yes, I would have taken more time off. 142 (60.7%)

Yes, I would have taken less time off. 7 (3.0%)

No, I took off the right amount of time for me and my child. 85 (36.3%)

Total 234 (100.0%)

Did you return to work earlier than you would have liked due to financial reasons?

Yes 158 (62.7%)

No 94 (37.3%)

Total 252 (100.0%)

How old was your first child when you returned to work?

1 week old 10 (4.0%)

2 weeks old 9 (3.6%)


4 weeks old 11 (4.4%)




6 weeks old 42 (16.8%)

7–8 weeks old 36 (14.4%)

9–11 weeks old 23 (9.2%)

3 months old 35 (14.0%)

4–6 months old 14 (5.6%)

7–9 months old 7 (2.8%)

10–11 months old 7 (2.8%)

1 year 8 (3.2%)

More than 1 year old 38 (15.2%)

Total 250 (100.0%)

If you took maternity/family leave around the time of your first pregnancy, how would you best describe that leave?

Unpaid leave through the Family Medical Leave Act (this is leave that you are eligible for if you work for a company with more than 50 employees. It provides up to 12 weeks of unpaid leave). 78 (30.1%)

Personal paid leave: This would be paid leave that you were eligible to take because you had saved up paid sick time or paid personal time. 43 (16.6%)

Company paid leave: Your company offered paid time off as a benefit of employment. 17 (6.6%)

A combination of paid and unpaid leave. 50 (19.3%)

Not applicable: I did not return to work after having the baby. 71 (27.4%)

Total 259 (100.0%)

Paternity Leave

Did the father of your first child take paternity leave from work around the time of your first pregnancy?

Yes, and he returned to work after I had the baby. 191 (75.2%)

No, he was not working then. 33 (13.0%)

Not applicable; father is not in the family. 27 (10.6%)

Yes, but he did not return to work after I had the baby. 3 (1.2%)

Total 254 (100.0%)

How much paternity/family leave did the father of your first child take before returning to work after the birth of your first biological child?

1 to 3 days 115 (59.0%)

4 to 6 days 5 (2.6%)

1 week to 13 days 51 (26.2%)

2 weeks to 20 days 15 (7.7%)







14 weeks to 104 days 1 (0.5%)

Total 195 (100.0%)

Table 5 provides data related to breastfeeding practices for the respondents’ first pregnancies. The percentage of mothers who attempted to breastfeed their infant was 84.4% (n=223). Among those who attempted to breastfeed their infants, 50.5% (n=100) of women stated that they stopped breastfeeding before they returned to work, while 49.5% (n=98) stopped breastfeeding after they returned to work. The percentage of women who stopped breastfeeding due to the difficulty of breastfeeding while working was 29.3% (n=61).

Table 5 Breastfeeding Practices During First Pregnancy

Participants’ Breastfeeding Practices Frequency (%)

For your first child, did you try to breastfeed at all? Yes 223 (84.4%)

No 40 (15.2%)

Total 263 (100.0%)

For your first child, if you started breastfeeding, when did you stop? Before I returned to work 100 (50.5%)

After I returned to work 98 (49.5%)

Total 198 (100.0%)

Did you stop breastfeeding because of the difficulty of breastfeeding while working?

Yes 61 (29.3%)

No 147 (70.7%)

Total 208 (100.0%)

Discussion

The main finding of this study was that Mississippi mothers were negatively impacted financially by their first pregnancies, and most would have taken more maternity leave if offered, had they been financially able to do so. More than half (53.1%) of the participants took fewer than 12 weeks of maternity leave around the birth of their first child; 10.3% took 4 weeks or less time off after having their first child. Half of the mothers who reported breastfeeding their infants (49.5%) stopped breastfeeding when they returned to work. Among all mothers who initiated breastfeeding, approximately one-third (29.3%) identified the difficulty of breastfeeding while working as the reason they stopped.



The first few months of parenthood after childbirth is commonly referred to as the fourth trimester (Verbiest, Tully, & Stuebe, 2017). During this time, new mothers are coping with physical recovery after delivery and determining how to best feed and care for their new infants. This can be a time of stress and fatigue for new families. Taking an abridged maternity leave may be acceptable for some mothers, but those who feel they need more maternity leave may not have the opportunity to take it due to financial hardship. Beyond the financial implications, returning to work earlier than a mother deems necessary can impact the health and well-being of the mother and child. Returning to work early may have overarching public-health implications. As of 2017, Mississippi was ranked as the highest state in the nation for infant mortality, or infant death that occurs in the first year of life. The leading cause of infant mortality in Mississippi is prematurity (less than 37 weeks gestation). The second leading cause of infant mortality in Mississippi is sudden unexpected infant death (SUID), which is death of an infant when the exact cause is not immediately known. Breastfeeding is especially beneficial for infants born prematurely and has been associated with significant reductions in SUID and neonatal illnesses (Mississippi State Department of Health, 2017). Mothers who have access to longer maternity leaves are more likely to breastfeed their infants and to do so for a longer period of time than mothers who return to work after shorter maternity leaves (less than 12 weeks). Further, paid maternity leave significantly predicts lower odds of rehospitalization after birth for the mother and the infant and higher odds of doing well with exercise and stress management after delivery (Jou, Kozhimannil, Abraham, Blewett, & McGovern, 2018). With so few individuals in Mississippi able to take advantage of even the unpaid leave granted by the FMLA, expanding access to paid maternity and family leave would greatly benefit women and families in Mississippi. Additionally, companies that offer paid leave may benefit financially. When paid leave was offered to employees, the rate at which mothers returned to work increased, and employers’ expenses were reduced because employee retention was increased (Gault, Hartmann, Hegewisch, Milli, & Reichlin, 2014). This study was entirely descriptive and, as such, is subject to several limitations. Though every effort was made to distribute the survey information cards widely, women who did not frequent areas that were canvassed did not have an opportunity to participate; thus, this study may not be representative of all mothers in Mississippi. For example, demographically, 59.2% of women in Mississippi are Caucasian and 37.8% are African American. Caucasians were overrepresented in our sample (88%). Another limitation was that the survey and survey information card were printed in English. This meant that non-English speaking individuals were excluded from participating. More generalizable longitudinal studies are warranted to determine the long-term association between parental leave and maternal and child well-being. Future research should associate length of parental leave to maternal and child outcomes in Mississippi. Policymakers



should recognize the significance of job security, both during and after pregnancy, for even the part-time worker (Guendelman et al., 2014). Practical implications include physicians’ emphasis of the importance of attending the post-delivery examination, which usually occurs 6 weeks postpartum. Best practice models that include home visits to provide lactation support and postnatal education about safe sleep environments should also be considered to provide support for both working and nonworking new mothers and their infants.

References

Abarca, N. E., Garro, A. C., & Pearlman, D. N. (2018). Relationship between breastfeeding and asthma prevalence in young children exposed to adverse childhood experiences. The Journal of Asthma, 13, 1–10. https://doi.org/10.1080/02770903.2018.1441869

Aitken, Z., Garrett, C. C., Hewitt, B., Keogh, L., Hocking, J. S., & Kavanagh, A. M. (2015). The maternal health outcomes of paid maternity leave: A systematic review. Social Science & Medicine, 130, 32–41. https://doi.org/10.1016/j.socscimed.2015.02.001

Appelbaum, E., & Milkman, R. (2011). Leaves that pay: Employer and worker experiences with paid family leave in California. Center for Economic and Policy Research. Retrieved from cepr.net/documents/publications/paid-family-leave-1-2011.pdf

Centers for Disease Control and Prevention. (2018). Breastfeeding: Why it matters. Retrieved from https://www.cdc.gov/breastfeeding/about-breastfeeding/why-it-matters.html

Chatterji, P., & Markowitz, S. (2012). Family leave after childbirth and the mental health of new mothers. The Journal of Mental Health Policy and Economics, 15(2), 61–76.

Cools, S., Fiva, J. H., & Kirkebøen, L. J. (2015). Causal effects of paternity leave on children and parents. Scandinavian Journal of Economics, 117. https://doi.org/10.1111/sjoe.12113

diversitydatakids.org. (2015). Working adults who are eligible for and can afford FMLA unpaid leave. Retrieved from http://www.diversitydatakids.org/data/ranking/529/working-adults-who-are-eligible-for-and-can-afford-fmla-unpaid-leave-share#loct=2&tf=17

Gault, B., Hartmann, H., Hegewisch, A., Milli, J., & Reichlin, L. (2014). Paid parental leave in the United States: What the data tell us about access, usage, and economic and health benefits. Retrieved from digitalcommons.ilr.cornell.edu/cgi/viewcontent.cgi?article=2608&context=key...

Guendelman, S., Goodman, J., Kharrazi, M., & Lahiff, M. (2014). Work-family balance after childbirth: The association between employer-offered leave characteristics and maternity leave duration. Maternal and Child Health Journal, 18(1), 200–208. https://doi.org/10.1007/s10995-013-1255-4

Guendelman, S., Pearl, M., Graham, S., Hubbard, A., Hosang, N., & Kharrazi, M. (2009). Maternity leave in the ninth month of pregnancy and birth outcomes among working women. Women’s Health Issues, 19(1), 30–37. https://doi.org/10.1016/j.whi.2008.07.007

IBM Corp. (Released 2015). IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY: IBM Corp.



Jou, J., Kozhimannil, K. B., Abraham, J. M., Blewett, L. A., & McGovern, P. M. (2018). Paid maternity leave in the United States: Associations with maternal and infant health. Maternal and Child Health Journal, 22(2), 216–225. https://doi.org/10.1007/s10995-017-2393-x

Lewis, S., Stumbitz, B., Miles, L., & Rouse, J. (2014). Maternity protection in SMEs: An international review, International Labor Organization. Retrieved from www.ilo.org/wcmsp5/groups/public/@dgreports/@dcomm/@publ/.../wcms_312783.pdf

Mariskind, C. (2017). Good mothers and responsible citizens: Analysis of public support for the extension of paid parental leave. Women’s Studies International Forum, 61, 14–19. https://doi.org/10.1016/j.wsif.2017.01.003

Mississippi State Department of Health. (2016). Infant Mortality Report 2016. Jackson, Mississippi.

Mississippi State Department of Health. (2017). Infant Mortality Report 2017. Jackson, Mississippi. Retrieved from https://msdh.ms.gov/msdhsite/_static/resources/7501.pdf

Raabe, P. H., & Theall, K. P. (2016). An analysis of paid family and sick leave advocacy in Louisiana: Lessons learned. Women’s Health Issues, 26(5), 488–495. https://doi.org/10.1016/j.whi.2016.07.003

Rossin, M. (2011). The effects of maternity leave on children’s birth and infant health outcomes in the United States. Journal of Health Economics, 30(2), 221–239. https://doi.org/10.1016/j.jhealeco.2011.01.005

Ruhm, C. J. (2000). Parental leave and child health. Journal of Health Economics, 19(6), 931–960. Retrieved from https://www.sciencedirect.com/science/article/pii/S0167629600000473/pdfft?md5=c77dc61acc99ace08343172af9b80886&pid=1-s2.0-S0167629600000473-main.pdf

Vahratian, A., & Johnson, T. R. B. (2009). Guest Editorial: Maternity leave benefits in the United States: Today’s economic climate underlines deficiencies. Birth, 36(3), 177–179. https://doi.org/10.1111/j.1523-536X.2009.00330.x

Verbiest, S. B., Tully, K. P., & Stuebe, A. M. (2017). Promoting maternal and infant health in the 4th trimester. Retrieved from https://beforeandbeyond.org/wp-content/uploads/2017/01/Zero-to-Three-Article-for-Dr.-Verbiest.pdf

Author Note

Jennifer Balcazar, Bachelor of Science Student (Allied Health, Management Emphasis), University of Southern Mississippi; Danielle Fastring, Department of Public Health, University of Southern Mississippi; and Avery Hilbert, Master of Public Health Candidate, University of Southern Mississippi. Correspondence concerning this article should be addressed to Danielle Fastring, University of Southern Mississippi, School of Health Professions, Box #5122, Hattiesburg, MS 39406. E-mail: [email protected]

SLPs and RTs: Team Approach to Caring for Patients


Speech-Language Pathologists and Respiratory Therapists:

Team Approach to Caring for Patients with Long-Term Tracheotomy

Javis M. Knott

Jackson State University

Celeste R. Parker

Xavier University of Louisiana

Background: Recent technological advances, together with growing social acceptance of patients with disabilities, has led to a realization of the importance of long-term management of technologically dependent and chronically ill patients with tracheostomies. This includes tracheostomy patients who are ventilator dependent, neurological patients, patients with severe illness such as stroke, and so forth. These patients are able to have a higher quality of life and communicate verbally due to advances in health care. One of the major advancements is communicating via a tracheostomy. Hence, this study will provide ways in which respiratory therapists (RTs) and speech-language pathologists (SLPs) can work together to make the process more efficient. Aim: The aim of this research article is to focus on a team approach utilizing the skills of speech-language pathologists and respiratory therapists to address communication issues for tracheostomy patients. Method: The authors reviewed historical and contemporary literature and computerized databases, and they also applied their collective 25-plus years of clinical and educational experience in the field. Results: The findings suggest that respiratory therapists and speech-language pathologists can work together to coordinate the most effective approach to helping patients with permanent tracheas regain full speech functionality and social adaptation. Conclusions: Rehabilitation of tracheostomy patients remains an important issue in modern medicine. There are a number of approaches to enhancing vocal speech in tracheostomy patients and ensuring full speech functionality and social adaptation. Successful speech therapies are based on coordinated interaction of respiratory and speech-language pathologists, nurses, caregivers, and patients. Findings regarding speech therapy for tracheostomy patients are based on a limited number of effective and controlled studies (Hess & Altobelli, 2014).



Introduction Recent technological advances, together with growing social acceptance of patients with disabilities, has led to a realization of the importance of long-term management of technologically dependent and chronically ill patients with tracheostomies. Infant, adolescent, and adult tracheostomy patients suffer from a number of risks associated with a complex invasive procedure (Hess & Altobelli, 2014; Lewarski, 2005). Communication difficulties associated with speech impairment are among those risks and can pose a threat to successful long-term management of tracheostomy patients and their social adaptation. A speech-language pathologist (SLP) is a health-care practitioner who specializes in the evaluation and treatment of communication and swallowing disorders. The components of speech production include phonation (producing sound); resonance; fluency; intonation (variance of pitch); and voice, including aeromechanical components of respiration. The components of language include phonology (manipulating sound according to the rules of a language); morphology (understanding and using minimal units of meaning); syntax (constructing sentences by using languages’ grammar rules); semantics (interpreting signs or symbols of communication to construct meaning); and pragmatics (social aspects of communication). Swallowing disorders include oropharyngeal and functional dysphagia in adults and children and feeding disorders in children and infants. SLPs are particularly crucial for pediatric patients because tracheostomy can lead to significant speech-development problems. Communication issues associated with tracheostomy can have a number of negative effects on a patient’s social life and well-being.

Respiratory therapists (RT) are specialized health-care practitioners who have earned a university degree and passed a national board-certifying examination. Respiratory therapists work most often in intensive-care and operating rooms, but they also are commonly found in outpatient clinics and home-health environments. Respiratory therapists are specialists and educators in cardiology and pulmonology. Respiratory therapists are also advanced-practice clinicians in airway management, establishing and maintaining the airway during trauma and intensive-care situations, and administering anesthesia for surgery or conscious sedation. In order to achieve the goals for evaluation and intervention for a tracheostomy patient, it is essential that speech-language pathologists, respiratory therapists, and other professionals work together as a team (Kobak, 2016; Hess & Altobelli, 2014; Hess, 2005). Therefore, it is crucial that the speech-language pathologist develops a strategic plan to address the communication issues and the respiratory therapist develops a plan to manage the tracheostomy.

Etiology and Concerns

Tracheostomy is a surgical operation that involves making an incision on the anterior side of the patient’s neck to directly open an airway. The created opening, tracheostomy, can be used as an



airway or a site to place a tracheostomy tube. This allows the patient to breath without using his or her mouth or nose. Tracheostomy patients form an etiologically diverse group that includes individuals with upper-airway deformities, individuals in need of lasting invasive ventilation, and individuals with a number of diseases that compromise the integral structure of the upper airway. Despite fundamental differences in the underlying conditions, long-term tracheostomy patients face similar challenges during their treatment and rehabilitation. The most common concerns can be grouped into the following categories (Hess & Altobelli, 2014; Siebens & Tippett, 1995):

• patient and caregiver education and training; • airway ventilation; • tracheostomy equipment management (installation, change, maintenance, check-ups); • humidification and suctioning needs; and • effective swallowing and vocal speech issues.

The latter is an important concern for respiratory therapists and speech-language pathologists to consider while providing a long-term care plan for tracheostomy patients.

Speech Enhancement and Rehabilitation

The practical and technological advances in handling respiratory failure has resulted in an increased number of tracheostomy patients. Most tracheostomy patients do not need speech therapy because they will have tracheostomies only for a limited time. The patients with long-standing tracheotomies can usually vocalize; however, a significant portion of them, particularly pediatric patients, need speech therapy (Kobak, 2016; Hess & Altobelli, 2014). Respiratory therapists, speech-language pathologists, and a surgical doctor should work together to determine if a patient is ready to use a speaking valve. Clinicians usually consider patients’ level of consciousness, ability to tolerate tracheostomy tube capping, cough effectiveness, and secretions as the most important factors in the decision to use a speaking valve (Hess & Altobelli, 2014). We will describe and analyze speech-enhancement practices for both ventilator-dependent patients and patients without mechanical ventilation. We will also discuss general therapeutic guidelines to prepare patients to successfully use their vocal cords. Ventilator-Dependent Patients There are a number of practices to assist vocal communication among tracheostomy patients dependent on ventilation; we will discuss talking tracheostomy tube, cuff down with a speaking valve, and cuff down without a speaking valve. Talking tracheostomy tube. The tube allows the patient to talk in a whispered voice. An important feature of the device is that it decouples breathing and talking functions because gas supplied by the ventilator passes through the larynx, enabling the patient to speak. Despite the



fact that this method allows the patient to vocally communicate, it is associated with a number of negative side effects. One of them is poor sound quality. The patient’s voice is very quiet, and volume can only be improved with higher air flows, which, in turn, can lead to damaged airways. Another pitfall is that an assistant is required to control the air flow pressure. The patient needs a period of training (by respiratory therapists and speech-language pathologists working together) in order to master vocal communication using a voice tube. Cuff down with a speaking valve. Using the speaking valve is based on the following principle: gas flows into the tube during inhalation and exits through the upper airways during exhalation. The method has been reported to improve various aspects of vocal communication in tracheostomy patients: speech flow and time, speech hesitancy, and so forth. Some patients might have difficulties adjusting to the use of a speaking valve. In that case, the assistance of a speech-language pathologist is needed (Hess, 2005; Kobak, 2016). Cuff down without a speaking valve. This technique is based on manipulations of the ventilator by a respiratory therapist to allow gas to escape through the upper airway during inspiration. This allows the patient to be able to speak during the inspiratory phase (Hess & Altobelli, 2014; Siebens & Tippett, 1995). In this phase, the speech-language pathologists teach/coach the patient on how to communicate while on the ventilator. It is important that the team monitors the patient for signs of distress at this time. Patients Without Mechanical Ventilation A tracheostomy tube can also be used in patients who are not mechanically ventilated. The following techniques can be used to address vocal communication issues in tracheostomy patients: cuff down finger occlusion and cuff down with speaking valve. Cuff down finger occlusion. The method is based on the following manipulation performed by the patient (or caregiver) when the cuff is down: the finger is placed on the proximal opening of the talking tracheostomy tube. This way, air is directed through the upper airways, which produce speech. The pitfall of this method is that many patients cannot coordinate their movements to master vocal speech with finger occlusion (Hess, 2005; Hess & Altobelli, 2014). Cuff down with a speaking valve. Air is directed through the upper airways by a speaking valve, which allows the patient to speak. This technique is most commonly used among patients and is relatively simple for them to master. Improvement of swallowing and olfaction has been reported among patients using the cuff down with a speaking valve method. This method, however, has a number of serious limitations, one of them being a high risk of aspiration. Therefore, the medical team should perform a thorough medical examination of the patient before prescribing this treatment (Hess, 2005; Hess & Altobelli, 2014).



SLP and RT Team Approaches to Training Patients/Caregivers The initial step of every successful speech-rehabilitation practice includes appropriate patient and caregiver training. Training should deal with the following important information blocks: basic airway anatomy, description of the tracheostomy invasion operation and its justification in the patient’s particular case, symptoms of respiratory and upper-airway distress, equipment installation and maintenance, and physician follow-up schedule. The training should be conducted for both adult and pediatric patients; ideally, this training should take place prior to the invasive procedure to prepare the patient and caregiver psychologically for possible outcomes (Lewarski, 2005; Hess & Altobelli, 2014). The long-term goal of speech therapy includes improved swallowing and vocal communication. There are a number of physiotherapeutic methods and device adjuncts to promote vocal communication in tracheostomy patients (Kobak, 2016; Lewarski, 2005). A crucial component of the speech-restoration process among long-term tracheostomy patients is teamwork between the patient and the care team (respiratory therapist, speech-language pathologist, and nurses) (Hess, 2005). Speech therapy for tracheostomy patients should focus on the following areas:

• educating the patient regarding the procedure and its effects on speech; • developing a strategy to address conditions of each specific patient (respiratory and

speech-pathology therapists, nurses, patient, caregivers); • choosing a correct communication-enhancement method and providing the patient with

appropriate training; • conducting speech development exercises (e.g., reading and talking to pediatric patients);

and • looking for an alternative communication technique if necessary (sign language,

electronic device, and so forth).

Conclusions

Rehabilitation of tracheostomy patients remains an important issue in modern medicine. There are a number of approaches to enhancing vocal speech in tracheostomy patients and ensuring full speech functionality and social adaptation. Successful speech therapies are based on coordinated interaction of respiratory therapists and speech-language pathologists, nurses, caregivers, and patients. The ability to speak is an important quality-of-life issue for tracheostomy patients.

Definitions

Tracheostomy is a surgical operation that involves making an incision on the anterior side of a patient’s neck to directly open an airway.



Tracheostomy tube is a tube that is placed through the tracheostomy opening to provide an airway and to remove secretions from the lungs. Speaking valve is a one-way valve that is attached to the end of the tracheostomy tube. It opens to allow air in through the tracheostomy during a breath in and then closes on the breath out, directing the air up through the larynx and out of the mouth in order to produce voice. Mechanical ventilator or respirator is a machine that improves the exchange of air between the lungs and the atmosphere. Aspiration is when food, stomach acid, or saliva is inhaled into the lungs.

References

Hess, D. R. (2005). Facilitating speech in the patient with a tracheostomy. Respiratory Care, 50(4), 519–525.

Hess, D. R., & Altobelli, N. P. (2014). Tracheostomy tubes. Respiratory Care, 59(6), 956–973. https://doi.org/10.4187/respcare.02920

Kobak, J. (2016, November 14). Swallowing and patients on mechanical ventilation: Something to chew on. Retrieved from https://dysphagiacafe.com/2016/11/14/swallowing-patients-mechanical-ventilation-something-chew/

Lewarski, J. S. (2005). Long-term care of the patient with a tracheostomy. Respiratory Care, 50(4), 534–537.

Tippett, D. C., & Siebens, A. A. (1995). Preserving oral communication in individuals with tracheostomy and ventilator dependency. American Journal of Speech-Language Pathology, 4(2), 55–61. doi:10.1044/1058-0360.0402.55

Author Note

Correspondence concerning this article should be addressed to Javis M. Knott, Healthcare Administration Program (HCA), School of Public Health, Jackson State University, Jackson, MS 39213. E-mail: [email protected]

Pre-K Obesity in Mississippi


Maternal, Child, and Parenting Factors Associated with Obesity Among Pre-Kindergarten Children in Mississippi

Jerome R. Kolbo, Angel Herring, Hwanseok Choi, Bonnie L. Harbaugh, Elaine Fontenot Molaison, Olivia Ismail, Lindsey Hardin, and Nichole Werle


Background: Obesity among children and youth has been consistently assessed among public school students in Mississippi since 2005. Significant declines in the prevalence of obesity among elementary students over the past decade suggest that changes may be occurring prior to entry into public school. Purpose: The purpose was to collect anthropometric data on a weighted, representative sample of children ages 3 to 5 years in licensed childcare facilities across Mississippi, and to correlate maternal, child, and parenting characteristics to obesity. Methods: The Body Mass Index was calculated using measured height and weight data. Results: A total of 14.12% of the 1,728 children were obese. Differences were not noted by age or gender but were significant by race, with 16.73% of Black children and 9.22% of White children categorized as obese (p < 0.0001). Obesity was significantly correlated with breastfeeding, hours of sleep, hours of child’s screen time, parent’s perception of the child’s weight gain, child’s birth weight, mother’s diabetes, and type of delivery. Conclusion: These findings provide a more complete picture of children’s health and factors impacting children’s health at an early age, and they can be of great value in future policy-making efforts to address unhealthy weights among children in Mississippi. Keywords: obesity, child, maternal, parenting, disparities



Introduction Between 2005 and 2015, Body Mass Index (BMI) and BMI percentiles have been calculated six times on Mississippi’s public school students in grades K-12 using measured height and weight data for weighted, representative samples in the Child and Youth Prevalence of Obesity Studies (CAYPOS; Kolbo et al., 2008; Kolbo et al., 2006; Kolbo et al., 2012; Kolbo et al., 2016; Molaison, Kolbo, Speed, Dickerson, & Zhang, 2008; Molaison et al., 2010; Zhang et al., 2014). Two similar studies using identical methods have been conducted among Mississippi’s Head Start preschoolers aged 3 to 5 years (Harbaugh, Bounds, Kolbo, Molaison, & Zhang, 2009; Harbaugh et al., 2011). In Mississippi’s most recent CAYPOS, the prevalence of overweight, obesity, and overweight and obesity combined remained higher than national averages, yet rates neither increased nor decreased significantly since 2005 (p = 0.6904; Kolbo et al. 2016). The combined prevalence of overweight and obesity for all students in grades K-12 was 43.4%, compared to 43.9% in 2005. As with all previous CAYPOS, there was no difference between boys and girls (p = 0.570). However, the prevalence of obesity in 2015 was again significantly higher among Black students (p < 0.001) than among White students. Similar to 2011 and 2013, in the 2015 CAYPOS, there was a significant difference in obesity prevalence by grade level (p = 0.0029), with the lowest prevalence again among the elementary students. Unlike all previous years, the highest prevalence of obesity was among high school students, and a linear increase on overweight among high school students between 2005 and 2015 was observed (p = 0.0152). The significant linear decrease in obesity prevalence among elementary school students observed from 2005 to 2013 continued to 2015 (p = 0.0209). The combined prevalence of overweight and obesity among elementary school students also showed a significant linear decline (p = 0.0002). While the CAYPOS is methodologically sound in the ongoing surveillance of obesity among public school students in grades K-12, one of the limitations is the ability to determine if these significant declines noted at the elementary level in Mississippi public schools are due to something occurring at the elementary grade level or whether students are arriving in public school with BMIs that are already lower than in previous years. Numerous studies have recently called for earlier assessment and comprehensive treatment of factors associated with childhood obesity (Farley & Dowell, 2014; Graversen et al., 2014; Kumanyika, Swank, Stachecki, Whitt-Glover, & Brennan, 2014; Pan, McGuire, Blanck, May-Murriel, & Grummer-Strawn, 2015; Wang, Gortmaker, & Taveras, 2011) and racial disparities (Balistreri & Van Hook, 2011) at a time when significant declines among the pre-K population are being reported (CDC, 2013; Farley & Dowell, 2014; Lo et al., 2014; Ogden, Carrol, Kit, & Flegal, 2014; Wen, Rifas-Shiman, Kleinman, & Taveras, 2012). A number of maternal, child, and parenting factors that appear to be associated with obesity at an early age have been identified in these and other recent studies. Maternal factors include the mother’s weight gain during pregnancy, tobacco use, the



development of gestational diabetes, gestational age of her child at delivery, and type of delivery (Durmus et al., 2013; Linabery et al., 2013; Nehring et al., 2013; Oken, Levitan, & Gillman, 2008). Parenting factors include whether or not the child was breastfed, age of introduction to solid foods, hours of sleep, and active play versus screen time (Anderson, Economos, & Must, 2008; Feig, Lipscombe, Tomlinson, & Blumer, 2011; Flores & Lin, 2013; Seach, Dharmage, Lowe, & Dixon, 2010). Child factors include the child’s birth weight and weight gain in infancy (Rooney, Mathiason, & Schauberger, 2011). The purpose of this study was to associate the prevalence of obesity with such variables among a pre-K population in order to provide additional insight into, and a more comprehensive understanding of, differences in race, gender, and grade levels being observed over the past decade in grades K-12 through the CAYPOS.

Methods

Support and approval for this study were obtained from the Mississippi Department of Health (MSDH), the MSDH Childcare Regulation and Licensure Division, the MSDH Child Care Advisory Council, and Head Start of Mississippi. The study received Institutional Review Board approval through the Human Subjects Committee at the University of Southern Mississippi and MSDH in September 2016. Data collection occurred between October 2016 and August 2017. A total of 94,975 children were enrolled in 1,517 Mississippi childcare centers licensed by MSDH at the onset of the study. The sampling frame consisted of 46,411 children in 1,241 of these licensed centers offering childcare to children ages 3 to 5 years. Consistent with sampling methods employed in the previous CAYPOS, the sample design was a two-stage stratified probability design. In the first stage, 100 licensed childcare facilities were randomly selected. A systematic sample of centers was drawn with probability proportional to the enrollment of children ages 3 to 5 years in each center. In the second stage, where centers had fewer than 50 children ages 3 to 5 years, all were selected. In centers where there were more than 50 children ages 3 to 5 years, classrooms were randomly selected within the sampled centers using equal probability systematic sampling. The sample was designed to result in a self-weighting sample so that every eligible child had an equal chance of selection, improving the precision of the estimates. The weighting process developed sample weights to assure the weighted sample estimates accurately represented the pre-K population in licensed childcare centers in Mississippi. Every eligible child was assigned a base weight, which was equal to the inverse of the probability of selection for the child. Adjustments were made to the initial weights to remove bias from the estimates and reduce the variability of the estimates. Center directors were sent a letter requesting their participation in the study, with an invitation to receive training on the study that included free lunch and reimbursement for travel. Follow-up



letters, e-mail, phone calls, and visits to the center informed center directors of the study. Center directors completed consent forms prior to participation in the study. Each center received $100 for participating in the study, which included assisting with the sampling of classrooms, encouraging parents to participate in the study, distributing parent consent forms and parent surveys, and coordinating a time for a trained MSDH representative to collect heights and weights of children whose parents consented to participation in the study. The MSDH representatives traveled to each of the centers to collect the data and return collected data to the project principal investigator (PI). After parental consent was obtained, the parent survey was used to obtain information on six variables. The questions and response sets (in parentheses) in the parent survey included:

• “From birth until your child was 1 year old, how did your child’s weight gain compare to other children their age?” (The child gained less weight than other children their age; The child gained the same as other children; The child gained more weight than other children their age);

• “Was your child breastfed?” (Yes; No; Don’t know); “If yes, how many months old was your child when breastfeeding was stopped?” (open-ended);

• “How many months old was your child when they first had baby food or other solid foods?” (open-ended);

• “How many total hours of sleep does your child currently get each night at home?” (open-ended);

• “How many total hours does your child usually watch TV or play on a computer or a phone each day at home?” (open-ended); and

• “How many total hours of exercise or physical activity (running around, jumping, dancing, etc.) does your child usually get each day at home?” (open-ended).

Consent was required for the trained MSDH representative to then collect the child’s height and weight, and for the MSDH Office of Vital Records to merge the child’s BMI data with official birth record data (mother’s weight gain, gestational diabetes, smoking, gestational age of child at delivery, type of delivery, and child’s birth weight). Measurements were performed by placing the weight scale on a hard, smooth surface. Children’s weight was measured using an MS7301 Automatic Step-On Digital Scale. The scale was calibrated to zero before use and recalibrated after every 10th child. Children’s height was measured using an HM200P PortStad Portable Stadiometer. Children were weighed and measured in a location where the information gathered would be confidential, and a center representative remained present, affording a measure of safety with two adults present. The MSDH representative recorded height and weight, rounded to the nearest whole inch or quarter pound. Heights and weights were added only to the survey form that had been initially completed by the parent.



After heights and weights were collected at each center, parents who consented received a $20 gift card. If heights and weights were collected on all eligible children sampled for participation in the center, the center received a $20 gift card. Consent forms, surveys, and recorded heights and weights for each center were placed in a sealed envelope at the center and then hand-delivered by the MSDH representative to the principal investigator for data entry. After data entry, the electronic file was sent to Westat Research Corporation for weighting and determination of BMI. Once Westat returned the file, the new data file with the BMI data was then submitted to the Office of Vital Records for matching with official birth record data. The matched file was returned to the PI for analysis. Data Analysis BMI was computed by Westat for each child based on height (in meters) and weight (in kilograms). The height in feet and inches was first converted to meters. The weight in pounds was then converted to kilograms. BMI was calculated using the Statistical Analysis System 9.3® (SAS) program (gc-calculate-BIV.sas) as follows: BMI = Weight (in kg) / [Height (in m)]. BMI values were checked to ensure the results were biologically plausible, using the limits developed by the Centers for Disease Control and Prevention (CDC, 2015). BMI classifications are age- and gender-specific as specified by the CDC. Children were classified as:

• underweight (BMI is less than the 5th percentile); • normal weight (BMI is equal to or greater than the 5th but less than the 85th

percentile); • overweight (BMI is equal to or greater than the 85th but less than the 95th

percentile); or • obese (BMI is equal to or greater than the 95th percentile).

SAS was used to calculate weighted estimates and standard errors. PROC FREQ procedure was used to compare prevalence of child obesity among different subgroups: gender (male and female); race (Caucasian, African-American, Other); and age (3, 4, and 5 years). Chi-square tests for categorical variables and t-tests and ANOVA for continuous variables were performed to test differences among obesity groups using the significance level 0.05. The Satterthwaite t-test was used if the equal variance assumption was violated as indicated by Levene’s test at α = .05.

Results

Characteristics of Pre-K Obesity Study Participants A total of 1,728 surveys with height and weight information were received from 69 (75%) of the 92 eligible centers. The response rate of the children was 90% (1,728 participating divided by 1,911 total sampled children). Thus, the overall response rate was 68%, above the threshold of 60% required to obtain weighted estimates.



After calculation of the BMI, the data were submitted to the Office of Vital Records at MSDH for matching with official birth record data. Of the 1,728 BMI data records, 1,588 records (92%) were able to be matched with official birth record data, which provided information on six of the twelve variables. Of those, 35 children were either 2 or 6 years at the time of the study so were excluded from analyses. A total of 1,553 birth records (1,588 - 35 = 1,553) were available for analysis. Consequently, BMI data were correlated with the six parent survey variables on the original sample of 1,728 children. The 175 records that could not be matched (1,728 - 1,553 = 175) with the official birth records were treated as missing in correlations between BMI and the six birth record variables. The sample of 1,728 children consisted of 873 males (50.52%) and 855 females (49.48%). A total of 567 (33.49%) were age 3, 778 (45.95%) were age 4, and 348 (20.56%) were age 5. There were 586 White students (33.91%), 998 Black students (57.75%), and 144 students from other racial/ethnic backgrounds (8.33%). The number of students in other race categories (e.g., Asian, Latino, Other) was too small for separate analysis and was not included in the comparison analyses. Results Based on Subgroups of Participants As a group, 14.12% of the children were classified as obese (Table 1). The prevalence of obesity did not differ by gender or age. A total of 15.01% of males were classified as obese. Of the females, 13.22% were obese. Differences in obesity prevalence by gender were not statistically significant (p = 0.601). In children age 3 years, 11.99% were classified as obese. Among children age 4 years, 14.40% were obese, and among 5-year-olds, 16.95% were obese. Differences in the prevalence of obesity by age were not statistically significant (p = 0.376). In terms of race, 9.22% of White children were classified as obese. Among Black children, 16.33% were obese. Obesity prevalence among Black children was significantly higher than among White children (p < 0.001).



Table 1 Percent Obese by Subgroup of Participants

Subgroups

Obese n (%)

Overweight n (%)

Normal n (%)

Underweight n (%)

All 244 (14.12%) 287 (16.61%) 1139 (65.91%) 58 (3.36%) Black 167 (16.73%) 157 (15.73%) 629 (63.03%) 45 (4.51%) White 54 (9.22%) 104 (17.75%) 421 (71.84%) 7 (1.19%) Male 131 (15.01%) 142 (16.27%) 574 (65.75%) 26 (2.98%) Female 113 (13.22%) 145 (16.96%) 565 (66.08%) 32 (3.74%) Age 3 68 (11.99%) 92 (16.23%) 391 (68.96%) 16 (2.82%) Age 4 112 (14.40%) 134 (17.22%) 508 (65.30%) 24 (3.08%) Age 5 59 (16.95%) 55 (15.80%) 220 (63.22%) 14 (4.02%)

Maternal, Child, and Parenting Variables Correlated with Obesity Breastfeeding. A total of 712 (41.86%) mothers reported that they did breastfeed, while 989 (58.14%) reported that they did not. Differences by child’s age and gender were not significant. However, differences in breastfeeding were significant by race (p < 0.001). A total of 49.3% of White mothers reported breastfeeding, compared to 39.8% of Black mothers (refer to Table 2). Obesity was negatively correlated to breastfeeding (p = 0.0046) (refer to Table 3). Among obese children, 33.47% were breastfed, while 66.53% were not breastfed.



Table 2 Bivariate Analyses Including Age, Gender, and Race

Variable 1 Variable 2 Test Statistic

p

Race White Black

Other

Breastfed 351 (49.3%) 283 (39.8%) 78 (10.9%)

Non-breastfed 227 (22.9%) 700 (70.8%)

62 (6.3%)

χ2 =

164.583

< .0001

Race White Black

Other

Sleep Hours (M, SD) 9.341 (1.28) 8.041 (1.23) 8.95 (1.18)

F =

106.49

< .0001

Race White Black

Other

Screen Hours (M, SD) 1.87 (1.07) 3.15 (1.86) 2.26 (1.35)

F =

125.07

< .0001

Age 3 4 5

Screen Hours (M, SD) 2.50 (1.67) 2.73 (1.73) 2.71 (1.72)

F = 3.155

0.043

Perceived Weight Gain Less

Same More

Male 129 (15.0%) 599 (69.8%) 130 (15.2%)

Female 151 (17.8%) 600 (70.8%) 97 (11.4%)

χ2 =

6.468

0.039

Perceived Weight Gain Less

Same More

White 114 (19.7%) 410 (70.8%)

55 (9.5%)

Black 139 (14.1%) 694 (70.2%) 155 (15.7%)

Other 27 (19.4%) 95 (68.4%) 17 (12.2%)

χ2 =

18.601

< .001

Gender Male

Female

Birth Weight in Grams (M, SD) 3,170.2 (563.1) 3,011.0 (550.7)

t =5.64

< .0001

Race White Black

Other

Birth Weight in Grams (M, SD) 3,242.5 (539.4) 3,011.3 (556.3) 3,066.9 (581.3)

F =

29.095

< .0001

Gender Male

Female

Physical Activity Hours (M, SD) 3.58 (1.69) 3.37 (1.71)

t =2.56

0.0107



Race White Black

Other

Physical Activity Hours (M, SD) 3.28 (1.58) 3.64 (1.75) 3.18 (1.73)

F =

10.686

< .0001

Race White Black

Other

Mother’s Weight Gain (M, SD) 40.96 (96.76) 28.28 (35.72)

49.55 (136.04)

F = 7.966

< .0001

Age 3 4 5

Mother’s Weight Gain (M, SD) 43.12 (121.95) 29.37 (17.03) 28.42 (15.03)

F = 6.544

< .0001

Age 3 4 5

Mother’s Tobacco Use 456 (71.25%) 149 (23.28%)

35 (5.47%)

χ2 =

667.09

< .0001

Gender Male

Female

Mother’s Tobacco Use 317 (47.46%) 351 (52.54%)

χ2 =

4.091

0.043



Table 3 Bivariate Analyses with Obese Group (BMI percentiles ≥ 95 %)

Variable Test Test Statistic p-value

Breastfeeding χ2 8.0331 0.0046 *

Hours of Sleep t-test 4.25 < .0001 *

Hours of Screen Time t-test 2.59 0.0096 *

Hours of Physical Activity t-test -0.15 0.8791

Introduction of Solid Foods t-test 0.45 0.6538

Perceived Weight Gain Satterthwaite t-test 8.30 < .0001 *

Mother’s Tobacco Use χ2 1.0795 0.2988

Mother’s Weight Gain Satterthwaite t-test 0.43 0.6687

Mother’s Diabetes χ2 9.4403 0.0021 **

Child’s Birth Weight t-test 3.32 0.0009 *

Gestation Age of Child t-test 0.10 0.9213

Method of Delivery χ2 12.3782 0.0004 *

* Statistically significant ** Statistically significant, yet n = less than 50

Hours of sleep. The average number of hours reported by mothers was 8.77 hours, with a standard deviation of 1.31 hours. The number of hours did not differ significantly by child’s age or gender. However, there was a significant difference by race (p < 0.001). The mean number of hours among White children was 9.34, compared to 8.04 among Black children. The number of hours of sleep was negatively correlated to obesity (p < 0.001). Among obese children, the average number of sleep hours at home was 8.44 (SD = 1.31), while the average number of hours of sleep at home for non-obese children was 8.82 (SD = 1.31). Screen time. The average number of hours mothers reported as their child’s screen time was 2.64 hours, with a standard deviation of 1.70 hours. The number of hours did not differ significantly by gender, but there was a significant difference by child’s age (p = 0.043) and race (p < 0.001). Children aged 3 watched 2.50 hours (SD = 1.67), children aged 4 watched 2.73 hours (SD = 1.73), and children aged 5 watched 2.71 hours (SD = 1.72). White children watched an average of 1.87 hours (SD = 1.07), and Black children watched an average of 3.15 hours (SD



= 1.86). Obesity was positively correlated with screen time (p < 0.001). Obese children watched an average of 2.95 hours (SD = 1.76), and children who were not obese watched an average of 2.59 hours (SD = 1.68). Perceived weight gain. Perceived weight gain did not differ by child’s age but differed by gender (p = 0.014) and race (p < 0.001). A higher percentage of females were viewed as gaining less weight that males (17.8% versus 15.0%), and a higher percentage of males were viewed as gaining more weight than females (15.2% versus 11.4%). A higher percentage of White children were viewed as gaining less weight than Black children (19.7% versus 14.1%), and a higher percentage of Black children were viewed as gaining more weight than White children (15.7% versus 9.5%). Obesity was positively correlated to perceptions of weight gain (p < 0.001). Among obese children, 6.79% of their mothers perceived their child’s weight gain as less than others, 11.84% perceived their child’s weight gain as the same as others, and 35.68% perceived their child’s weight gain as more than others. Child’s birth weight. Children’s birth weights were obtained from MSDH official birth records. The average weight was 3,090.46 g, or approximately 6.80 lb. Birth weight did not differ by child’s age (p = 0.2901) but did differ significantly by gender (p < 0.001) and race (p < 0.001). The average birth weight among boys was 3,170.2 g (6.97 lb) compared to 3,011.0 g (6.62 lb) among girls. The average birth weight among White children was 3,242.5 g (7.13 lb) compared to 3,011.3 g (6.62 lb) among Black children. Obesity was positively correlated with birth weight (p = 0.0009). Obese children had an average birth weight of 7.00 lb, while non-obese children had an average birth weight of 6.76 lb. Delivery method. The delivery method was obtained from MSDH official birth records. Two categories capturing vaginal birth were collapsed into one variable, and two categories of C-section were collapsed into another variable. A total of 62.26% were vaginal deliveries, while 37.74% were C-sections. Delivery method did not differ by gender, child’s age, or race. Obesity was significantly correlated (p = 0.0004) to delivery method. Among those children who were delivered via vaginal birth, 12.06% were obese, while 18.54% of those delivered via C-section were obese. Maternal, Child, and Parenting Variables Not Correlated with Obesity Introduction of solid foods. The average age of introduction was 5.54 months, with a standard deviation of 2.07 months. The number of months did not differ significantly by child’s age, gender, or race. Obesity was not significantly correlated with the introduction of solid foods (p = 0.6538). Physical activity. The average number of hours was 3.48 hours, with a standard deviation of 1.70 hours. The number of hours did not differ significantly by child’s age. However, there was a



significant difference by gender (p = 0.011) and race (p < 0.001). The average number of hours among boys was 3.58 (SD = 1.69), and the average number of hours among girls was 3.37 (SD = 1.71). Among White children, the average number of hours was 3.28 (SD = 1.58), and the average number of hours among Black children was 3.64 (SD = 1.75). Obesity was not significantly correlated to physical activity (p = 0.8898). Mother’s weight gain. Mother’s weight gain was obtained from MSDH official birth records. Mothers gained an average of 33.82 lb. Weight gained did not differ by the child’s gender (p = 0.2257) but did differ by the child’s age (p = 0.0015) and race (p = 0.0004). Among children aged 3, mother’s weight gain was 43.12 lb (SD = 121.95), while among children aged 4, mother’s weight gain was 29.37 lb (SD = 17.03), and among children aged 5, mother’s weight gain was 28.42 lb (SD = 15.03). Among White children, mother’s weight gain was 40.96 lb (SD = 96.76); among Black children, mother’s weight gain was 29.37 lb (SD = 35.72). Obesity was not related to mother’s weight gain. Mother’s tobacco use. Mother’s tobacco use was obtained from MSDH official birth records. Most (61.34%) did not smoke during pregnancy, while 38.66% did. Tobacco use did not differ by race (p = 0.1806) but did differ by child’s age (p < 0.001) and gender (p = 0.043). Among children aged 3, 73.25% of mothers reported tobacco use, while among children aged 4, 23.28% reported tobacco use, and among children aged 5, 5.4% reported tobacco use. Among girls, 52.54% of mothers reported tobacco use, and among boys, 47.46% of mothers reported tobacco use. Obesity was not related to mother’s tobacco use. Mother’s diabetes. Mother’s diabetes data was obtained from MSDH official birth records. Only 43 mothers (2.49%) were reported to have diabetes during the pregnancy. Diabetes did not differ by the child’s gender, age, or race. While obesity was significantly correlated to mother’s diabetes, this finding is not considered reliable given the low number (n = 43) included. Gestation. The obstetric estimate of gestation was obtained from MSDH official birth records. The mean number of weeks was 38.03. Gestation did not differ by gender, child’s age, or race. Obesity was not significantly correlated to gestation (p = 0.9213).

Discussion

Obesity was present in 14.12% of the 1,728 children in this study. Differences were not noted by age or gender, but they were significant by race, with 16.73% of Black children and 9.22% of White children measuring as obese (p < 0.0001). These findings are consistent with national data indicating Black children have a higher probability of becoming obese than do their non-Hispanic and non-Black peers, which has held true for more than a decade (Rossen, 2012). Compared to White children, Black children have BMI trajectory profiles associated with higher



risks for subsequent obesity, including a younger age of onset and classification. These disparities are noted by the preschool years, with a particular increase in prevalence noted in Black females (Wang, Gortmaker, & Taveras, 2011; Wen et al., 2012). Prior Mississippi studies on childhood obesity have found that a high percentage of students are already overweight by the first grade, with the prevalence trending an increase with each subsequent grade (Molaison et al., 2008). The higher obesity rate in Mississippi’s preschool children at 14.2%, as compared to the national rate of 13.9%, could be connected to the rurality and lower socioeconomic status of the state. Though modest declines in obesity prevalence for most racial groups of low-income children ages 2 to 4 years have been reported, the prevalence rate for obesity still remains higher for low-income children (Pan et al., 2015). Unfortunately, no measures of parental or household income or other measures of socioeconomic status were collected in this study. Evidence suggests children who live in more metropolitan or urban areas are less likely to be overweight and obese due to greater access to healthy food options and healthy eating information (Hansstein, 2016; Kimbo, Brooks-Gunn, & McLanahan, 2007). Nationally, both gender and age differences have been associated with higher prevalence rates for obesity (BMI > 95th percentile) and severe obesity (BMI > 1.2 x 95th percentile) linked to boys and varying by age group (Lo et al., 2014). Obesity was negatively correlated with breastfeeding. This finding is consistent with national examinations of breastfeeding and childhood obesity that indicate breastfeeding increases the likelihood of being normal weight by 5.2% (p < 0.001) and decreases the likelihood of being obese by 8.6% (p < 0.001) (Hansstein, 2016). High birth weight status has been connected to childhood obesity, with reports showing children with high birth weights were 2.5 times as likely to be overweight or obese (Kimbro, Brooks-Gunn, & McLanahan, 2007). In addition, this study also found significant correlations between obesity and hours of sleep, hours of child’s screen time, parent’s perception of the child’s weight gain in the first year, mother’s diabetes, and the method of delivery. Obesity was not correlated with the physical activity of the child, timing of the introduction of solid foods, mother’s tobacco use during pregnancy, mother’s weight gain during pregnancy, or gestational age of the child at delivery. Kimbro, Brooks-Gunn, and McLanahan (2007) also found no significant correlation between physical activity and childhood obesity. However, Hillier, Pedula, Vesco, Oshiro, and Ogasawara (2016) found that excessive maternal weight gain (>40 lb) was associated with a more than 15% increased risk of childhood overweight and obesity among normal birth weight children, a finding that was not supported in this analysis. This study was limited to a random sample of children ages 3 to 5 years from a random sample of licensed childcare facilities across the state. As such, it did not include children who are in unlicensed childcare or at home. Differences by age, gender, and race were examined, but not



combined factors such as gender and race. The study included a number of maternal, child, and parenting factors recently associated with childhood obesity. It is likely that other factors individually or collectively played a role in these findings, as nationally representative samples suggest the relationship between collective health factors and other life variables seem to influence overweight and obesity differently across genders and age categories (Balistreri & Van Hook, 2011). Another limitation is that the information collected on many of the variables included in this study came solely from the mother’s self-report. Information the mothers provided at birth (e.g., tobacco use) and through the six questions in the survey were not confirmed by any other source. Also, the sample size limited the analysis of the variables in relation to correlation with mother’s diabetes, but this finding was not considered reliable given the low number (n = 43) of mothers whose birth records indicated that they had diabetes. However, evidence exists to support a link between gestational diabetes and a greater prevalence for developing childhood overweight and obesity during the first decade of life (p < 0.0001), with the risk of childhood overweight and obesity nearly 30% or higher with a maternal diagnosis of gestational diabetes (Hillier et al., 2016). Additional analyses need to be conducted to more thoroughly explore the relationships between childhood obesity and maternal, child, and parenting factors. For example, do the differences in weight status classification (i.e., overweight, normal weight, or underweight) vary significantly by age, gender, or race? Do the maternal, child, and parenting variables correlate differently to the different weight status classifications? Also, how do the maternal, child, and parenting variables correlate with each other? For example, does breastfeeding, or the length of time one breastfeeds, correlate with other maternal health variables such as tobacco use or other child variables such as sleeping or the introduction of solid foods? Are sleeping and screen time associated and, thereby, impacting the child’s weight status? Future research could be conducted on these same children over time or new cohorts of pre-kindergarten children as new policies and interventions are enacted. While prevention is ideal, resolution of unhealthy weights at a young age is highly desirable, since early intervention and correction lessen long-term exposure to chronic weight-related illnesses and may instill healthy eating and activity habits that extend to later childhood and adulthood. Preschoolers are in a unique developmental period of cognitive growth, and increasing social exposures to new, non-family environments (such as Head Start and licensed preschools) can make them receptive to positive messages from both family and school. In addition to the physical health-related outcomes associated with childhood obesity, childhood obesity is associated with psychosocial and academic risks. Obese children are more likely to experience low self-esteem, depression, anxiety, behavioral problems, and poor body image;



endure teasing, social discrimination, and emotional distress; and often underachieve academically (Daniels, 2006; Kihm & Rolling, 2014; Kuhl et al., 2012; Lang, 2012; Larsen et al., 2006). Obese children miss four times as much school as their normal-weight peers and score significantly lower in kindergarten-level math and reading test scores when compared to non-overweight children (Datar et al., 2004; Satcher, 2005).

State-specific obesity prevalence surveillance has been highlighted as key in determining the need for and impact of state- and local-level obesity-prevention strategies (CDC, 2013). For this reason and the many negative health-related obesity associations listed earlier, it is imperative to continue surveillance of the substantially high obesity prevalence in our preschoolers and children, which will have a significant impact on early onset of obesity-related illnesses, the longevity of lifetime exposure to those illnesses, as well as adult obesity and other morbidities (Harbaugh et al., 2011; Ogden, Carroll, Kit, & Flegal, 2014; Wang, Gortmaker, & Taveras, 2011). Some states have witnessed a recent, slight decline in childhood obesity in their youngest children, with explanations focusing on many potential levels. Among those levels, day care, school food policies, and media messages have been identified as influential (Farley & Dowell, 2014). Though Mississippi children have historically been among the most obese children in the nation, the recent MSDH Physical Activity in the Child Care Setting (PACCS) initiative has targeted the reduction of high-prevalence obesity among 2- to 5-year-olds through the implementation of guidance and standards for childcare facilities to better their physical activity and nutrition environments (MSDH, 2017). With an increasing number of children in this age range spending their days in childcare facilities, it is logical to target childcare facilities’ physical activity opportunities and nutritional content of meals. These standards have been nationally recognized for their focus on child nutrition, outdoor exposure, and limits on screen time. These standards are currently influencing state policy, which require daily, structured physical activity in all licensed facilities; an action plan to help facilities work toward the improved goals and requirements for physical activity; and an Allies for Quality Care program, which will provide direct assistance for selected childcare facilities to improve their learning environments and nutritional offerings for Mississippi’s youngest children (Mississippi State Department of Health, 2017).

References Anderson, S. E., Economos, C. D., & Must, A. (2008). Active play and screen time in U.S.

children aged 4 to 11 years in relation to sociodemographic and weight status characteristics: A nationally representative cross-sectional analysis. BMC Public Health, 8(366), 1–13.



Balistreri, K. S., & Van Hook, J. (2011). Trajectories of overweight among U.S. school children: A focus on social and economic characteristics. Maternal Child Health Journal, 15(5), 610–619.

Centers for Disease Control and Prevention. (2011). Obesity in K-8 students – New York City, 2006–07 to 2010–11 school years. Morbidity and Mortality Weekly Report, 60(49), 1673–1678.

Centers for Disease Control and Prevention. (2013). Vital signs: Obesity among low-income, pre-school-aged children – United States. Morbidity and Mortality Weekly Report, 62(31), 629–634.

Centers for Disease Control and Prevention. (2015). A SAS program for the CDC growth charts. Retrieved from: https://www.cdc.gov/nccdphp/dnpao/growthcharts/resources/sas.htm

Daniels, S. R. (2006). The consequences of childhood overweight and obesity. The Future of Children, 16(1), 47–67.

Durmus, B., Arends, L. R., Ay, L., Hokken-Koelega, A. C., Raat, H., Hofman, A., … Jaddoe, V. W. (2013). Parental anthropometrics, early growth, and the risk of overweight in pre-school children: The Generation R study. Pediatric Obesity, 8(5), 339–350.

Datar, A., Sturm, R., & Magnabosco, J. L. (2004). Childhood overweight and academic performance: National study of kindergartners and first-graders. Obesity Research, 12(1), 58–68.

Farley, T. A., & Dowell, D. (2014). Preventing childhood obesity: What are we doing right? American Journal of Public Health, 104(9), 1579–1583.

Feig, D. S., Lipscombe, L. L., Tomlinson, G., & Blumer, I. (2011). Breastfeeding predicts the risk of childhood obesity in a multi-ethnic cohort of women with diabetes. The Journal of Maternal-Fetal & Neonatal Medicine, 24(3), 511–515.

Flores, G., & Lin, H. (2013). Factors predicting severe childhood obesity in kindergartners. International Journal of Obesity, 37(1), 31–39.

Graversen, L., Sørensen, T. I. A., Petersen, L., Sovio, U., Kaakinen, M., Sandbaek, A., … Obel, C. (2014). Preschool weight and body mass index in relation to central obesity and metabolic syndrome in adulthood. PLoS ONE, 9(3), 1–9.

Hansstein, F. V. (2016). The impact of breastfeeding on early childhood obesity: Evidence from the National Survey of Children’s Health. American Journal of Health Promotion, 30(4), 250–258.

Harbaugh, B., Bounds, W., Kolbo, J. R., Molaison, E., & Zhang, L. (2009). Prevalence estimates of overweight among Head Start preschoolers. The Journal of Pediatric Nursing, 24(5), 350–359.

Harbaugh, B., Kolbo, J. R., Molaison, E., Hudson, G. M., Zhang, L., & Wells, D. (2011). Obesity and overweight prevalence among a Mississippi low-income preschool population: A five-year comparison. International Scholarly Research Network (ISRN) in Nursing, 2011, 1–7.



Hillier, T. A., Pedula, K. L., Vesco, K. K., Oshiro, C. E. S., & Ogasawara, K. K. (2016). Impact of maternal glucose and gestational weight gain on child obesity over the first decade of life in normal birth weight infants. Maternal and Child Health Journal, 20(8), 1559–1568.

Kihm, H. S., & Rolling, P. (2014). The preschool nap as a protective factor in the fight against childhood obesity: An exploratory study. Journal of Family and Consumer Sciences, 106(4), 40–50.

Kimbro, R. T., Brooks-Gunn, J., & McLanahan, S. (2007). Racial and ethnic differentials in overweight and obesity among 3-year-old children. American Journal of Public Health, 97(2), 298–305.

Kolbo, J. R, Armstrong, M. G., Blom, L., Bounds, W., Molaison, E., Dickerson, H., … Zhang, L. (2008). Prevalence of obesity and overweight among children and youth in Mississippi: Current trends in weight status. Journal of the Mississippi State Medical Association, 49(8), 231–237.

Kolbo, J. R., Penman, A. D., Meyer, M. K., Speed, N. M., Molaison, E. F., & Zhang, L. (2006). Prevalence of overweight among elementary and middle school students in Mississippi compared with prevalence data from the Youth Risk Behavior Surveillance System. Preventing Chronic Disease, 3(3), 1–10.

Kolbo, J. R., Zhang, L., Molaison, E. F., Harbaugh, B., Hudson, G. M., Armstrong, M. G., & Werle, N. (2012). Prevalence and trends in overweight and obesity among Mississippi public school students, 2005–2011. Journal of the Mississippi State Medical Association, 53(5), 140–146.

Kolbo, J. R., Zhang, L., Werle, N., Molaison, E. F., Harbaugh, B., Kirkup, M., & Walker, E. (2016). Overweight and obesity prevalence and trends among Mississippi public school students: A decade of data between 2005 and 2015. Journal of the Mississippi State Medical Association, 57(10), 1–9.

Kuhl, E. S., Clifford, L. M., & Stark, L. J. (2012). Obesity in preschoolers: Behavioral correlates and directions for treatment. Pediatric Obesity, 20(1), 3–29.

Kumanyika, S. K., Swank, M., Stachecki, J., Whitt-Glover, M. C., & Brennan, L. K. (2014). Examining the evidence for policy and environmental strategies to prevent childhood obesity in Black communities: New directions and next steps. Obesity Reviews, 15(4), 177–203.

Lang, J. E. (2012). Obesity, nutrition, and asthma in children. Pediatric Allergy, Immunology, and Pulmonology, 25(2), 64–75.

Larsen, P., Nelson, M. C., Page, P., & Popkin, B. M. (2006). Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics, 117(2), 417–424.

Linabery, A. M., Nahhas, R. W., Johnson, W., Choh, A. C., Towne, B., Odegaard, A. O., … Demerath, E. W. (2013). Stronger influence of maternal than paternal obesity on infant



and early childhood body mass index: The Fels Longitudinal Study. Pediatric Obesity, 8(3), 159–169.

Lo, J. C., Maring, B., Chandra, M., Daniels, S. R., Sinaiko, A., Daley, M. F., & Greenspan, L. C. (2014). Prevalence of obesity and extreme obesity in children aged 3–5 years. Pediatric Obesity, 9(3), 167–175.

Molaison, E. F., Kolbo, J. R., Speed, N., Dickerson, E., & Zhang, L. (2008). Prevalence of overweight among children and youth in Mississippi: A comparison between 2003 and 2005. Retrieved from http://www.mshealthpolicy.com

Molaison, E. F., Kolbo, J. R., Zhang, L., Harbaugh, B., Armstrong, M. G., Rushing, K., … Green A. (2010). Prevalence and trends in obesity among Mississippi public school students, 2005–2009. Journal of the Mississippi State Medical Association, 51(3), 1–6.

Mississippi State Department of Health (2018). Physical activity in the child care setting. Retrieved from https://msdh.ms.gov/msdhsite/index.cfm/30,0,183,526,html

Nehring, I., Chmitorz, A., Reulen, H., Von Kries, R., & Ensenauer, R. (2013). Gestational diabetes predicts the risk of childhood overweight and abdominal circumference independent of maternal obesity. Diabetic Medicine, 30(12), 1449–1456.

Ogden, C. L., Carrol, M. D., Kit, B. K., & Flegal, K. M. (2014). Prevalence of child and adult obesity in the United States, 2011–2012. Journal of the American Medical Association, 311(8), 806–814.

Oken, E., Levitan, E., & Gillman, M. (2008). Maternal smoking during pregnancy and child overweight: Systematic review and meta-analysis. International Journal of Obesity, 32(2), 201–210.

Pan, L., McGuire, L. C., Blanck, H. M., May-Murriel, A. L., & Grummer-Strawn, L. M. (2015). Racial/ethnic differences in obesity trends among young low-income children. American Journal of Preventive Medicine, 48(5), 570–574.

Rooney, B. L., Mathiason, M. A., & Schauberger, C. W. (2011). Predictors of obesity in childhood, adolescence, and adulthood in a birth cohort. Maternal and Child Health Journal, 15(8), 1166–1175.

Rossen, L., M., & Schoendorf, K. C. (2012). Measuring health disparities: Trends in racial−ethnic and socioeconomic disparities in obesity among 2- to 18-year old youth in the United States, 2001–2010. Annals of Epidemiology, 22(10), 698–704.

Satcher, D. (2005). Healthy and ready to learn. Educational Leadership, 63(1), 26–30. Seach, K. A., Dharmage, S. C., Lowe, A. J., & Dixon, J. B. (2010). Delayed introduction of solid

feeding reduces child overweight and obesity at 10 years. International Journal of Obesity, 34(10), 1475–1479.

Wang, C. Y., Gortmaker, S. L., & Taveras, E. M. (2011). Trends and racial/ethnic disparities in severe obesity among U.S. children and adolescents, 1976–2006. International Journal of Pediatric Obesity, 6(1), 12–20.

Wen, X., Kleinman, K., Rifas-Shiman, S. L., Gillman, M. W., & Taveras, E. M. (2012). Childhood body mass index trajectories: Modeling, characterizing, pairwise correlations,



and socio-demographic predictors of trajectory characteristics. BMC Medical Research Methodology, 12(38) 1–13.

Zhang, L., Kolbo, J. R., Kirkup, M., Molaison, E. F., Harbaugh, B., Werle, N., & Walker, E. (2014). Prevalence and trends in overweight and obesity among Mississippi public school students, 2005–2013. Journal of the Mississippi State Medical Association, 55(3), 80–87.

Author Note

Jerome R. Kolbo, School of Social Work, University of Southern Mississippi; Angel Herring, Department of Child and Family Studies, University of Southern Mississippi; Hwanseok Choi, Department of Public Health, University of Southern Mississippi; Bonnie L. Harbaugh, College of Nursing, University of Southern Mississippi; Elaine Fontenot Molaison, Department of Nutrition and Food Systems, University of Southern Mississippi; Olivia Ismail, School of Social Work, University of Southern Mississippi; Lindsey Hardin, School of Social Work, University of Southern Mississippi; and Nichole Werle, School of Social Work, University of Southern Mississippi. Funding for this study was provided by the Center for Mississippi Health Policy. This study would not be possible without the support and assistance of the Childcare Regulation and Licensure Division and Office of Vital Records in the Mississippi State Department of Health. The authors wish to thank the center directors and staff of the licensed childcare centers across Mississippi who were so instrumental in gathering vital information from parents and children. Correspondence concerning this article should be addressed to Jerome R. Kolbo, School of Social Work, University of Southern Mississippi, 118 College Drive #5114, Hattiesburg, Mississippi 39406. E-mail: [email protected]

Student-Run Free Clinics


Training the Next Generation of Primary-Care Physicians:

Are Student-Run Free Clinics (SRFCs) the Way to Go?

Tobe Momah University of Mississippi Medical Center

Rita Momah

Jackson State University

William Replogle University of Mississippi Medical Center

Elizabeth McClain

William Carey University

Makayla Merritt William Carey University

Background: The consensus over the last 20 years is that increased availability of primary care reduces the overall cost of healthcare and improves mortality and morbidity rates by as much as 1.44 fewer deaths per 10,000 people (American College of Physicians, 2008; Shi, Starfield, Kennedy, & Kawachi, 1999). However, not enough physicians are going into primary care to meet the need for improved and increased access. By 2020, the expanded Title VII program goal is to produce a physician workforce that is at least 40 percent primary care (Jackson et al., 2014). Estimates show that 74% of U.S. medical school graduates go into non-primary care specialties (Pugno, McGaha, Schmittling, DeVilbiss, & Kahn, 2007), but a shortage of 45,000 primary care physicians remains (U.S. Department of Health and Human Services, 2013). Purpose: In light of this dire forecast, this paper builds upon a proposal by Campos-Outcalt and Senf (1999) (where a student-run clinic affected a minority of students’ residency choices) to see if the use of student-run free clinics among students at a state academic medical center would build medical students’ interest in primary care and influence their choice of residency. Methods: A retrospective analysis of volunteers at the Jackson Free Clinic from 2012 to 2017 in Jackson, Mississippi, was carried out. Results: Results indicated a statistical correlation between the fourth-year medical students’ frequency of volunteering at JFC and their choice of primary care in residency. However, for first-, second-, and third-year medical students, there was no significant correlation between volunteer time at JFC and choosing a primary care residency. Keywords: free clinic, primary care, underserved



Introduction

Student-run free clinics (SRFCs) have been in operation since 1965, when David Smith started the Haight Ashbury Free Clinics near the University of California, San Francisco (Smith et al., 2014; Xu, 2013). This has since morphed into 207 entities, sponsored by students and faculty from 111 different U.S. university medical schools (Simpson & Long, 2007). These are led by students, staffed by licensed physicians, and supported by the local community. It is an archetype of clinical outreach to the uninsured that enables poor and indigent individuals to see a physician and obtain some of their medications for free. The Jackson Free Clinic (JFC) in Jackson, Mississippi, is the only SRFC in the state. It has been pursuing the delivery of high-quality health services to individuals with inadequate access in Jackson since its opening in 1999. It sees an average of 1,150 patients a year (A. M. Mercier, personal communication, August 14, 2017), and more than 80% of University of Mississippi Medical Center (UMMC) students volunteer their time on a weekly basis. It is community-supported through various fund-raising efforts. Physicians and residents from the UMMC Psychiatry, Family Medicine, and Internal Medicine departments coordinate care and provide medical coverage for patients. UMMC is the state of Mississippi’s only academic medical center and trains more than 70% of all physicians (primary care and non-primary care) educated in the state (Wolfe, 2017). As the state with the lowest number of primary care physicians per capita (8.3 physicians per 10,000 Mississippians, compared to 12.7 physicians per 10,000 patients nationally; Krause, 2015), Mississippi needs to add at least 1,300 primary-care physicians over the next 8 years to reach the national average.

Method

Participants The student volunteers are from all 4 years (M1–M4) of the UMMC medical school. In December 2016, JFC stopped receiving dental student or faculty volunteers. Volunteers from the occupational therapy, physical therapy, nutrition, and pharmacy schools do participate. The average number of student volunteers per week is between 20 and 30, while physician volunteers number between three and five weekly. Procedure This study was conducted at the Jackson Free Clinic (JFC) in Jackson, Mississippi. Institutional Review Board (IRB) approval was not required as no human specimens were involved in this study. At the time of the study, JFC was staffed by volunteer faculty and residents from the UMMC Departments of Psychiatry, Internal Medicine, and Family Medicine who alternated



Saturdays supervising the clinical aspect of JFC’s care. The clinic is open only on Saturdays from 11:30 a.m. to 4 p.m. and has an average of approximately 20 patients per week. The students and volunteer physicians document their names and departments in a log book at the SRFC administrative office before commencing their activities for the day. Statistical analysis was done on the 123 and 133 students from the UMMC medical school 2016 and 2017 graduating classes, respectively, to see if there was a correlation between participation in JFC during their medical school years and matching into a primary-care residency program as part of the annual MATCH program by Electronic Residency Application Service (see Table 1). Data from the 2016 and 2017 MATCH programs was compared to the frequency of volunteering at JFC using the statistical program SPSS 22.0. It analyzed the frequency of volunteerism and residency choice using Pearson’s chi-squared values (Table 2). The chi-squared analysis was then used to determine the correlation between volunteer status and choice of residency.

Table 1 Percentage of Medical Students Who Went into Primary Care

1-Primary care: 0-Not primary care

Frequency Percent Valid

percent Cumulative

percent Valid 0 106 41.4 41.4 41.4

1 150 58.6 58.6 100.0 Total 256 100.0 100.0

Table 2 Attendance of Medical Students at Jackson Free Clinic (2012–2017)

Statistics M1 M2 M3 M4 N Valid 256 256 256 256

Missing 0 0 0 0 Mean .58 .77 1.65 .71 Minimum 0 0 0 0 Maximum 38 13 13 13



Results

All categorical data, including demographic information, is represented in numbers and percentages, and a P value of less than 0.05 was considered statistically significant. Our data collection rate was 100%, as, between 2012 and 2017, every JFC clinical activity, including names of student volunteers, was documented in an easily retrievable log book. The sample was then analyzed in regards to the percentage of students who matched in primary care versus those who matched in non-primary care residencies. Table 1 shows that the majority of UMMC medical students (58.6%) in the 2016 and 2017 MATCH programs went into primary care, with 41.2% going into non-primary-care residencies and other specialties. The mean number of times students volunteered at JFC was highest in the third year of medical school (Table 2). When JFC student volunteers were analyzed for their choice of specialty, the correlation between going into a primary care residency (or not) and the total times volunteered in the 4 years of medical school were non-significant (r = .042, and p = .502; Table 3). There was a significant correlation between time volunteered in the M4 year (r = .142, p <0.023) and choosing primary care (obstetrics/gynecology, internal medicine, family medicine, pediatrics, and med-pediatrics) as a residency. Other results indicated internal medicine (28%), pediatrics (10.9%), family medicine (10.8%), and emergency medicine (8.6%) were the most popular residencies chosen by UMMC medical students in 2016 and 2017 (Table 4). There was no positive correlation between volunteering at the JFC and a MATCH into emergency medicine or pediatrics, as was seen in earlier studies (Weinreich, Kafer, Tahara, & Frishman, 2005).



Table 3 Correlation Between Going into Primary Care Residency (or Not) and Total Times Volunteered at Jackson Free Clinic

Correlations

1-Primary care: 0-Not

primary care Total Pearson correlation .042

Sig. (2-tailed) .502 N 256

M1 Pearson correlation -.045

Sig. (2-tailed) .475 N 256

M2 Pearson correlation .037

Sig. (2-tailed) .561 N 256

M3 Pearson correlation .041

Sig. (2-tailed) .518 N 256

M4 Pearson correlation .142*

Sig. (2-tailed) .023 N 256

Year Pearson correlation -.062

Sig. (2-tailed) .320 N 256

*Correlation is significant at the 0.05 level (2-tailed).



Table 4 Residency Choices by UMMC Medical Students

Residency

Frequency Percent Cumulative

Percent Valid Anesthes 13 5.1 5.1

Derm 5 2.0 7.0 EM 22 8.6 15.6 ENT 4 1.6 17.2 FM 26 10.2 27.3 Gen Sx 17 6.6 34.0 IM 72 28.1 62.1 Med Peds 2 0.8 62.9 MedPeds 11 4.3 67.2 NeuroSx 2 0.8 68.0 ObGyn 9 3.5 71.5 OMFS 1 0.4 71.9 Opth 1 0.4 72.3 Opth 3 1.2 73.4 OPTH 1 0.4 73.8 Ortho 7 2.7 76.6 Ortho Sx 1 0.4 77.0 Pathology 5 2.0 78.9 PedGenetics 1 0.4 79.3 Peds 28 10.9 90.2 PEds 2 0.8 91.0 Plastic Sx 2 0.8 91.8 Psych 7 2.7 94.5 Radiation 1 0.4 94.9 Radiolo 1 0.4 95.3 Radiology 8 3.1 98.4 RadOnco 1 0.4 98.8 Transition 2 0.8 99.6 Urology 1 0.4 100.0 Total 256 100.0



Discussion

The JFC data indicate that those who participated actively in JFC during their fourth year of medical school were more willing to choose primary-care residencies at a statistically significant rate. The confounding variable, however, was that the choice of a medical residency is typically decided at the beginning of medical students’ fourth year when applications and letters of recommendation are being uploaded to the ERAS system by the M4 students. This study corroborates the importance of JFC among those who volunteered in their fourth year of medical school and eventually went into primary-care residencies, but it does not show that JFC volunteerism influenced choices between M1 and M3 years. This study suggests greater emphasis must be paid to M4 medical students who volunteer at SRFCs in their fourth year. They are a potential recruitment pool for future primary-care residents. Any institution, therefore, without an SRFC structure may be at a disadvantage during interviews. For example, a program with an SRFC in place will showcase its availability to those interviewing and prioritize M4 SRFC volunteers for the ERAS MATCH. Even though UMMC mandates that third-year medical students volunteer at least one week in the year for JFC, there is no evidence from this study that this policy has increased recruitment into primary care. UMMC has consistently produced more primary-care physicians than specialists, compared with the national average, and this has been a result of innovative programs such as the Mississippi Rural Physicians Scholarship Program (MRPSP; Helseth, 2014) and the Office of Mississippi Physician Workforce (OMPW; Morris, 2015). The MRPSP, which was authorized by the Mississippi Legislature in 2007, sponsors up to 64 medical students ($2.1 million) annually with the requirement that they return to a rural area in Mississippi to practice primary care post-residency for at least 4 years. OMPW, on the other hand, was authorized by the Mississippi Legislature in 2012 with the mandate to create more family residency programs and increase physician workforce availability in Mississippi. These two programs have helped to maintain a steadily growing number of primary-care physicians in Mississippi. However, the number of primary-care physicians UMMC produces is skewed when those who enter non-primary-care internal-medicine residencies are taken into consideration. In a recent graduation ceremony of the UMMC internal medicine program, the chair of the department confirmed that less than 10% of graduates were continuing as primary-care physicians (Dr. T. Tonore, personal communication, June 2017). The unequal distribution of primary-care physicians and internal-medicine-produced sub-specialists has been attributed to the widening influence of specialists in major academic centers like UMMC (Kirch & Patel, 2017).



Medicine is currently at a crossroads. SRFCs have been proposed by several authors as a tool to rekindle medical students’ interest in primary care (Starfield, Shi, & Macinko, 2005). With decreased reimbursement, longer hours, and an aging workforce (50% of primary-care physicians are above 50 years of age; Council on Graduate Medical Education, 2010), medical students’ interest in primary care is at an all-time low. SRFCs are not overwhelmed with the insurance bureaucracy and financial problems that primary-care physicians tend to be. They, thus, afford medical students an opportunity to see primary care in its ideal manifestation—a relationship between the physician and his or her patients. SRFCs could be an opportunity to showcase primary-care medicine to the coming generations of medical students. How well SRFCs are implemented may determine how many choose to go into primary-care medicine. This study shows that SRFCs are a selling point to fourth-year medical students interested in primary care. They also provide an opportunity to identify those with a genuine interest in primary-care medicine, and give medical students hands-on learning experiences outside their traditional learning environment.

References

American College of Physicians. (2008). How is a shortage of primary care physicians affecting the quality and cost of medical care? Philadelphia: American College of Physicians; 2008: White Paper. (Available from American College of Physicians, 190 N. Independence Mall West, Philadelphia, PA 19106.)

Campos-Outcalt, D., & Senf, J. (1999). A longitudinal, national study of the effect of implementing a required third-year family practice clerkship or a department of family medicine on the selection of family medicine by medical students. Academic Medicine, 74(9), 1016–1020. doi: 10.1097/00001888-199909000-00016

Council on Graduate Medical Education. (2010). Twentieth report: Advancing primary care. Retrieved from https://www.hrsa.gov/

Helseth, C. (2014, May 15). Mississippi scholarship program helps create more rural physicians. The Rural Monitor. Retrieved from https://www.ruralhealthinfo.org/rural-monitor/mississippi-physician-scholarship/

Jackson A., Baron R. B., Jaeger J., Liebow, M., Plews-Ogan, M., & Schwarts, M. D. (2014). Addressing the nation’s physician workforce needs: The Society of General Internal Medicine (SGIM) recommendations on graduate medical education reform. Journal of General Internal Medicine, 29(11), 1546–1551. doi: 10.1007/s11606-014-2847-4

Kirch, D., & Patelle, K. (2017). Addressing the physician shortage: The peril of ignoring demography. Journal of the American Medical Association, 317(19), 1947–1948. doi:10.1001/jama.2017.2714



Krause, D. D. (2015). Data lakes and data visualization: An innovative approach to address the challenges of access to health care in Mississippi. Online Journal of Public Health Informatics, 7(3). Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4731224/

Morris, M. (2015, February 2). Mississippi needs right programs, right places. The Daily Journal. Retrieved from http://www.djournal.com/lifestyle/health/mississippi-needs-right-programs-right-places/article_fe91334d-5501-5d8b-bfbb-e7c78681db0b.html

Pugno, P. A., McGaha, A. L., Schmittling, G. T., DeVilbiss, A., & Kahn, N. B. (2007). Results of the 2007 National Resident Matching Program: Family medicine. Family Medicine, 39(8), 562–571.

Shi L., Starfield, B., Kennedy, B., & Kawachi, I. (1999). Income inequality, primary care, and health indicators. Journal of Family Practice, 48(4), 275–284.

Smith, S., Thomas, R., Cruz, M., Griggs, R., Moscato, B., & Ferrara, A. (2014). Presence and characteristics of student-run free clinics in medical schools. Journal of the American Medical Association, 312(22), 2407–2410. doi:10.1001/jama.2014.16066

Simpson, S. A., & Long, J. A. (2007). Medical student-run health clinics: Important contributors to patient care and medical education. Journal of General Internal Medicine, 22, 352–356.

Starfield, B., Shi, L., & Macinko, J. (2005). Contribution of primary care to health systems and health. The Milbank Quarterly, 83(3), 457–502

U.S. Department of Health and Human Services, Health Resources and Services Administration, National Center for Health Workforce Analysis. (2013). Projecting the supply and demand for primary care practitioners through 2020. Rockville, Maryland: U.S. Department of Health and Human Services.

Weinreich, M., Kafer, I., Tahara, D., & Frishman, W. (2005). Participants in a medical student-run free clinic and career choice. Journal of Contemporary Medical Education, 3(1).

Wolfe, A. (2017, August 4). New school of medicine hopes to address doctor shortage in Mississippi. The Clarion Ledger. Retrieved from https://www.clarionledger.com/story/news/politics/2017/08/04/new-school-medicine-hopes-address-doctor-shortage-mississippi/525436001/

Xu, J. (2013, November 12). Letting medical students run the clinic. The Atlantic. Retrieved from https://www.theatlantic.com/health/archive/2013/11/letting-medical-students-run-the-clinic/281241/



Author Note Tobe Momah, Department of Family Medicine, University of Mississippi Medical Center; Rita Momah, Department of Health Policy and Management, Jackson State University; William Replogle, Department of Biostatistics, University of Mississippi Medical Center; Elizabeth McClain, Medical Education, William Carey University; and Makayla Merritt, Preclinical Sciences, William Carey University. Correspondence concerning this article should be addressed to Tobe Momah, Department of Family Medicine, University of Mississippi Medical Center (UMMC), Clinical Science Building, 2500 North State Street (fourth floor), Jackson, MS 39216. E-mail: [email protected]

Rural Medical Scholars Program


Rural Medical Scholars Program: Filling the Gap for Health-Care and Public Health Leaders in Mississippi

Ann Sansing, David R. Buys, Marion W. Evans, Laura Downey, and Jasmine Harris-Speight Mississippi State University

The Rural Medical & Science Scholars program aims to help rising high school seniors determine if they want to pursue health-related careers. The program shapes students’ interest in and understanding of medicine, health-related disciplines, and other science, technology, engineering, and mathematics (STEM) fields. The program combines didactic, observational, and practical learning during a summer semester. Participants earn seven college credits to jump-start a health or STEM career. We report on descriptive statistics since the program’s inception in 1998. The program has matriculated 401 students, of whom approximately 71% have chosen health-related careers in nursing, physical or occupational therapy, dentistry, pharmacy, public health, or medical research. Others are pursuing science-based careers in chemical, biological, or mechanical engineering; information technology; and science-based educational fields. The scholars learn independence and soft skills such as time management, study skills, effective communication, relationship-building skills, and critical thinking. These skills will benefit them in their academic and professional careers. The program has been successful in promoting medical and STEM-related fields while, at the same time, helping to fill the gap for health-care and public-health leaders in Mississippi. Keywords: workforce, public-health infrastructure, physician shadowing, high school career



Introduction

The Mississippi State University (MSU) Extension Service provides early experiences to rising high school seniors in an effort to fill gaps in the number of rural health-care providers. MSU Extension has a history dating to 1914 as part of the Cooperative Extension Service, which is associated with the land-grant university system. In addition to mission areas in agriculture and natural resources, Extension addresses family and consumer sciences (FCS), 4-H youth development, and community resource development (CRD). The need for more health-care providers, which is a perennial issue in Mississippi, sits at the intersection of those FCS, 4-H, and CRD program areas. With additional health-care providers and public health leaders, Mississippi would be better positioned to address its last-place status in health outcomes (United Health Foundation, 2017). In 1998, MSU Extension began offering the Rural Medical Scholars (RMS) program in response to Mississippi’s low number of physicians per capita. Additional health-care providers are needed, given that Mississippi has 6.4 active primary-care physicians per 10,000 population, which is below the national average of 9 physicians per 10,000 population (American Association of Medical Colleges, 2017). The primary goal of RMS has been to “grow local docs” by identifying competitive rising high school seniors interested in the field of medicine. The program has been offered for 18 years since 1998; in 2008 and 2009, funding was suspended during the Great Recession. RMS scholars are routinely admitted to MSU for one summer term, during which they earn seven pre-medicine college credits; shadow health-care providers; and engage in workshops emphasizing preventive medicine, health-behavior change, and professional development. The collective experiences in RMS help participants solidify an undergraduate focus on pre-medical science. Those students who remain on the pre-medical science track contribute to a change in economic and social outcomes in rural areas. According to a national report (American Medical Association, 2018), each physician in Mississippi contributes an average total of $1.8 million in economic output to their community or region. RMS funding has originated from several sources during the program’s history (Carew, Cossman, & Sansing, 2011). From 1998 to 2006, a United States Department of Agriculture/Cooperative State Research, Education, and Extension Service grant supported RMS. This funding assured recruitment from the state’s 15 community college regions to maintain an equitable geographic disbursement of scholars. Since this federal funding ended in 2017, funding sources have included the Mississippi Institute for the Improvement of Geographic and Minority Health, Mississippi State Office of Rural Health, Office of the Provost at MSU, Appalachian Regional Commission, and designated donations from the Mississippi Rural Healthcare Association, Mississippi Rural Physicians Scholarship Program, and



CREATE/Wellspring Toyota. For the last 6 years, MSU Extension has been the primary program sponsor.

Methods

Theoretical Models Informing Program Development Theoretically driven design helps ensure the success and sustainability of programs like the RMS program. Toward that end, program staff have identified the Theory of Reasoned Action/Planned Behavior (TRA/TPB) as a useful model for organizing and explaining components of RMS’s development, evaluation, and refinement (Azjen & Driver, 1991). The TRA/TPB defines the links between beliefs, attitudes, norms, intentions, and behaviors. The program takes into account that scholars’ knowledge, learned attitudes, beliefs, and aspirations change throughout the program and that scholars develop a positive attitude toward rural medical care. Within the TRA/TPB model, attitudes drive intent, and intent is believed to drive behavior. Program Components and Curriculum Recruitment and enrollment. Historically, RMS participants have been recruited from across Mississippi via press releases, social media, and radio and web-based marketing. RMS program staff send materials to high school counselors and health-science program coordinators who share program information with rising high school seniors. Additionally, MSU Extension agents and partnering organizations share the recruitment materials with potential program applicants. Applicants submit a biographical statement including academic performance and out-of-school activities, and essays detailing their interest in RMS. Scholars are selected by a blind-review panel using a rubric for the scoring process. Counselors and tutors. RMS staff select counselors who serve as mentors and reside with the scholars in the dorms throughout the summer term; staff also hire tutors who facilitate mandated study sessions to reinforce good study habits, time management, and successful completion of coursework. On-campus living. Scholars live in a residential dorm at MSU during the week and return home on weekends. Scholars depart campus on Friday afternoons and return on Sunday evenings to resume the program. This experience helps scholars prepare for life as college students. Orientation. Before classes begin, scholars attend a 3-day orientation, which includes multiple workshops focused on communication skills, teamwork, study skills, and critical thinking. Individual topics relevant to the life of a physician are also presented during the workshop. Over the past 2 years, additional workshop components were added on the topics of preventive medicine and health-behavior change.



Academic coursework. Scholars enroll in introductory biology (with lab) and a sociology course taught by MSU faculty. The didactic courses are taught in the morning, and the biology lab takes place in the afternoon. At the end of the program, which is a 4-week summer session, scholars earn seven college credits. Both courses meet pre-medicine curriculum requirements for a medical career. Most scholars perform well in the classroom, which is a positive indicator that they will do well in medical or professional school. The experiential/observational and experiential/practical workshops are in addition to the didactic seven hours of coursework. Additional learning experiences include shadowing physicians, Junior Master Wellness Volunteer training, visiting the University of Mississippi School of Medicine, and participating in a video documentary review and reflection, Scholars in the Kitchen, and Scholars in the Lab. RMS is a rigorous and intensive program; however, scholars continually report that the program helped them answer that all-important question, “Is medicine the career choice for me?” Shadowing. Scholars shadow primary-care physicians and a limited number of specialists. Shadowing gives scholars the opportunity to experience the day-to-day work of a physician and offers communication skill-building in a professional setting. Shadowing is a critical component to the program’s success. Scholars finish with 12 to 15 hours of physician shadowing experiences, which boosts their total number of hours over many of their peers when applying for medical school. Junior Master Wellness Volunteer program. Scholars began training and earning certification as Junior Master Wellness Volunteers in 2016. This training enables them to return to their communities and provide accurate health information to promote healthy choices and lifestyle changes. Each scholar is required to earn at least 24 hours of community service over the following academic year. This certification helps enhance scholars’ résumés for professional health-care careers. Scholars form relationships with their local MSU Extension agents, who assist with community-service projects. The curriculum features a volunteer component, a social media component, and a curriculum modified from the University of Mississippi Medical Center’s Community Health Advocate program. Additional modules relevant to school-based issues and a toolkit for carrying out community-action projects are also included in the curriculum. Visit to academic medical center. Additionally, scholars visit the University of Mississippi School of Medicine. The visit includes an interactive question-and-answer session with the dean of admissions to gain insight on future admission to medical school. They visit with primary-care physicians and hear personal stories of their journeys into medical school. Scholars also engage in medical-simulation activities that introduce them to the rigors of physicians’ work.



Video documentary review and reflection. Scholars review two video documentaries, MD: The Making of a Doctor (1987) and Doctors’ Diaries (2009), and work in assigned teams to reflect on the lives of a variety of physicians through their medical-school and residency experiences and into their careers and personal lives. The activity allows scholars to imagine themselves in their chosen field and, at the end of RMS, answer the question, “What have you learned about becoming a physician, and what does it mean to you?” They provide written reflections and deliver oral presentations to their peers as part of this assignment. Scholars in the Lab. This is a 1.5-hour experiential/practical learning experience that provides a tour of the MSU College of Veterinary Medicine. This experience includes an overview of the One Health concept, which refers to the commonalities between human and animal health. Faculty at the College of Veterinary Medicine teach a hands-on workshop using pigs’ feet to teach three types of suturing skills. Scholars in the Kitchen. This observational/practical learning workshop emphasizes the connection between nutrition and overall health. Scholars provide an ingredient list and recipes for the preparation of a nutritious evening meal and are required to plan using a designated budget. Scholars determine whether to prepare one large meal for everyone or to work in small groups to prepare group meals. This experience teaches cooking skills, teamwork, and communication skills. Scholars can use the skills obtained in this workshop to prepare nutritious meals in their dorms during the program. Social media connections. Scholars exchange phone numbers and connect via social media and group-texting platforms during the first week of the program. During orientation, many of the activities are strategically designed to promote interaction and group bonding. The social-media and phone-based communication approaches provide opportunities for the group to engage and further develop relationships. Counselors and scholars report staying connected after the program on GroupMe, Facebook, and other platforms. Evaluation Methods As part of ongoing evaluation efforts, scholars continually respond to didactic, experiential/observational, and experiential/practical learning experiences through both written, quantitative evaluation and qualitative reflections. Reflective learning is a mechanism that allows the learner to think back about the experience by reflecting on words and feelings that bring clarity to concepts taught, help generate new knowledge and ideas, and allow explanation of ideas to someone who was not present. Thus, scholars are required to submit reflection papers on various components of the program. To make the activity more challenging and competitive, communications experts review the documents and award first, second, and third place for the best papers. Students choose one reflection to share at the end of the program celebration dinner,



when awards are presented. They also can use this document to share about their experiences during interviews for professional or graduate school. Participants complete evaluations to assess program satisfaction and intention to pursue health-care careers. The evaluation component features four primary questionnaires administered throughout the program. The initial questionnaire is administered as a pre-test to evaluate scholars’ baseline knowledge and expectations for the program. A second questionnaire is used to evaluate various presenters and presentations. The third questionnaire is administered for each shadowing experience. A final questionnaire is administered at the end of the program that summarizes the experience as a whole. Analysis As noted previously, RMS program staff maintain records of participants’ county of origin, as well as their progress through college and post-graduate studies, including medical or other professional school. Staff also record demographic data about the participants and report on univariate statistics. Furthermore, staff compile their reflective journals and responses to components of the program and conduct open coding and thematic analysis of the assignments. Two coders analyze the data and compare themes for consistency and agreement. In particular for this manuscript, we report on results from the question, “What was the greatest benefit you personally gained from participating in the program?” Additionally, program staff maintain ongoing evaluation with participants through social media. The Mississippi State University Institutional Review Board has reviewed and approved the evaluation efforts of the Rural Medical Scholars program.

Results

Description of Scholars Since the program’s inception, a total of 401 scholars have participated in the program (Table 1). Of these, 157 (40%) were males and 244 (60%) were females. In addition, 103 (26%) participants represented minorities; specifically, through 2018, participants have included 15 Asian males; 23 Asian females; 22 Black males; 33 Black females; 5 Hispanic males; and 5 Hispanic females. Scholars have come from 67 of Mississippi’s 82 counties (Figure 1). Forty-six scholars have completed or are actively enrolled in medical school. Of these, 36 are currently practicing physicians (residency or private practice). In addition, 29 are in primary care and 15 are practicing in Mississippi. Eighty-two scholars have yet to complete college.



Table 1 Number of Rural Medical Scholars Since 1998 Number of scholars since 1998 401

Males 157 (39%)

Females 244 (61%)

Minorities 103 (26%)

Asian males 15

Asian females 23

Black males 22

Black females 33

Hispanic males 5

Hispanic females 5

Number of counties represented 67



Figure 1. Rural Medical Scholars 1998–2018 participants’ home counties.

Follow-up evaluation shows that 80% of former RMS medical-school graduates entered primary-care residency programs and 17% chose family medicine as their area of specialty. In addition, while 26% of all participants are from minority races/ethnicities, 30% of RMS scholars who have gone to medical school are from minority races/ethnicities, suggesting that our program is providing an important boost for minority scholars. Furthermore, approximately 71% of the scholars are currently studying for or engaged in other health-related careers such as nursing, pharmacy, dentistry, physical or occupational therapy, counseling, medical research, and public



health. Others are pursuing careers in mechanical, biological, or chemical engineering; technology; or math professions. Only a small number (less than 10%) of scholars decide to pursue careers outside of the sciences. Qualitative Impact Assessment At the conclusion of each RMS year, scholars are asked, “What was the greatest benefit you personally gained from participating in the program?” Their responses, combined with informal follow-up assessment, indicate the most important aspects of the program are the social networks scholars build, the independence and soft skills they develop, and the shadowing they experience. Social networks. Scholars consistently report meaningful relationships built with other scholars and their counselors. Some of them have attended the same college and even roomed together. They often find themselves attending the same medical schools and stay connected for many years. Scholars and counselors report that the dynamic established during RMS continues and that scholars contact their former counselors for advice; these counselors act as mentors throughout scholars’ college, medical school, and residency years. Independence and soft skills. Scholars also comment that, because they have the opportunity to earn seven college credits and experience college life a year early, they develop a sense of independence and learn time management, study skills, and critical thinking. They also learn effective communication skills and have daily opportunities to implement them with program staff, peers, and course faculty. Value of shadowing. The shadowing component is overwhelmingly a highlight of the program; scholars learn from and enjoy these real-life experiences. Shadowing, coupled with the visit to the University of Mississippi School of Medicine, make a medical or other health-care career seem more achievable. This visit is often reported as pivotal for scholars’ choosing medicine or choosing another career. Overall program themes. Following are quotes from RMS alumni that represent key themes of the overall program.

“RMS solidified my interest in medicine, specifically primary care. You could say that, without RMS, I would not be where I am today!" “This program has allowed me to realize that there are far more options for careers in healthcare than I could have ever known. RMS has provided me with opportunities of a lifetime, endless resources, and the motivation to pursue my dreams of making a difference in the lives of others.”



“RMS began my interest in a career in the medical field and launched my trajectory into my current career as a quality engineer, ensuring medical devices meet regulatory requirements for the safety of the patient.” “Those who can, should. RMS showed me that I CAN.”

Discussion and Conclusions

The Rural Medical Scholars program has helped high school seniors determine if they want to pursue health-related careers. As noted previously, the theory of planned behavior/ theory of reasoned action drove development and evaluation of this program. The theory notes that intention predicts behavior, and this program helps students solidify their intention regarding pursuit of medical school. They also learn life skills such as teamwork, building relationships, community service and engagement, and self-efficacy related to performing various tasks. A social media network helps keep the group connected and, therefore, reinforces various normative beliefs regarding the attainment of a medical education among the group after completing the RMS program. The Rural Medical Scholars program, which could be replicated in other states, may help address health-care workforce shortages across the United States. RMS expanded its scope and name in 2018 to Rural Medical & Science Scholars. The program continues to focus on “growing local docs” but offers additional experiences in science, technology, engineering, and mathematics (STEM), as well as other opportunities in the health-care field. Approximately 71% of the scholars have chosen some type of health-related career, while other scholars have indicated an interest in or pursued other STEM careers. The expansion of the program’s vision will help ensure a strong and passionate workforce for the long-term goals of improving access to healthcare and improving Mississippi’s science-based economy. The MSU Rural Medical and Science Scholars program has the ability to produce locally educated physicians and perhaps scientists specializing in health-related fields. Future research should report on the outcomes of scholars who do not go into medical or health-care fields. Researchers should also track those scholars who become health-care providers but do not practice in rural settings.

References

American Association of Medical Colleges. (2017). Mississippi Profile. 2017 State physician workforce book. Retrieved from https://www.aamc.org/data/workforce/reports/484392/2017-state-physician-workforce-data-report.html



American Medical Association. (2018). The national economic impact of physicians national report. Fairfax, VA: IQVIA. Retrieved from https://www.physicianseconomicimpact.org/

Azjen, I., & Driver, B. L. (1991). Prediction of leisure participation from behavioral, normative, and control beliefs: An application of the theory of planned behavior. Leisure Science, 13, 185–204. doi:10.1080/01490409109513137

Carew, B., Cossman, J., & Sansing, A. (2011). Rural medical scholars: A pipeline program for Mississippi’s future physicians. Journal of the National AHEC Organization, 27(1), 23.

Doctors’ diaries [Television broadcast]. (2009, April 7). WGBH Boston (Producer). Arlington, VA: Public Broadcasting Service.

MD: The making of a doctor [Television broadcast]. (1987). WGBH Boston (Producer). Arlington, VA: Public Broadcasting Service.

United Health Foundation. (2017). America’s health rankings: A call to action for individuals and their communities. Annual report, 2017. Retrieved from https://www.americashealthrankings.org/learn/reports/2017-annual-report

Author Note

Ann Sansing, Department of Food Science, Nutrition, and Health Promotion, Mississippi State University Extension Service; David R. Buys, Department of Food Science, Nutrition, and Health Promotion, Mississippi State University Extension Service and Mississippi Agricultural and Forestry Experiment Station; Marion W. Evans Jr., Department of Food Science, Nutrition, and Health Promotion, Mississippi State University; Laura Downey, School of Human Sciences, Mississippi State University; Jasmine Harris-Speight, Department of Food Science, Nutrition, and Health Promotion, Mississippi State University. Correspondence concerning this article should be addressed to David R. Buys, Assistant Extension and Research Professor, Department of Food Science, Nutrition, and Health Promotion, Mississippi State University Extension Service and Mississippi Agricultural and Forestry Experiment Station, P.O. Box 9805, Mississippi State, Mississippi 39759. E-mail: [email protected]

Journal of Public Health in the Deep South

Aim and Scope The Journal of Public Health in the Deep South is a peer-reviewed, open-access, online journal focused on disseminating high-quality research and/or effective applied techniques to academicians, educators, and practitioners. Topics addressed include aging; alcohol; chronic disease; climate change; communicable disease; community water fluoridation; ebola; environmental health; financial health; global health; gun violence; health administration; health disparities; health education; health equity; health promotion; health reform; health in all policies; health rankings; health community design; healthy housing; high school graduation; injury and violence prevention; lead contamination; maternal and child health; medical care; mental health; nursing; nutrition; office professionals; physical activity; preparedness; prescription drug overdose; public health accreditation; public health education, policy, and research; public health standards; racism and health; reproductive and sexual health; rural health; school-based health care; social determinants of health; social work; stress; substance misuse; tobacco; transportation; vaccines; zika; and other public health-related topics. JPHDS is the first publication to specifically focus on public health issues in this geographical space and with an appeal to both research and applied public health professionals and educators. JPHDS will initially be hosted on the MPHA website and have an early focus on research, practice, and teaching emanating from Mississippi. Types of Articles Published Several types of articles in the content area listed above are considered appropriate for the journal: full manuscripts, research briefs (overview of ongoing projects, preliminary research, or limited findings), multimedia presentations (digital interviews, lectures, and presentations), and practice and pedagogy. Individuals who are accepted to present at the Mississippi Public Health Association Annual Conference are encouraged to submit their work to JPHDS. Additionally, the journal will accept editorials that express points of view on health issues. Editors may directly invite individuals to submit a manuscript. Frequency of Publication The Journal of Public Health in the Deep South will be published at least once a year, with occasional intermittent special issues. Open-Access Policy JPHDS is a fully open-access journal, meaning that all works published in the journal are freely available to read, download, copy, print, and share/transmit. ISSN In process Publication Agreement JPHDS requires that all authors sign a publication agreement prior to online publication of an accepted manuscript. For More Information Visit https://mpha2.wildapricot.org/

journal of public health in the deep south · survey designed by mathematica policy research was...

Documents