measuring the magnitude of oral health inequalities within ... · greatly indebted to my thesis...
TRANSCRIPT
Measuring the Magnitude of Oral Health Inequalities within and between the United States and Canada from 1970 to
2009
by
Julie W. Farmer
A thesis submitted in conformity with the requirements for the degree of Masters of Science in Dental Public Health
Graduate Department of Dentistry University of Toronto
© Copyright by Julie W Farmer (2015)
ii
Measuring the Magnitude of Oral Health Inequalities within and
between the United States and Canada from 1970 to 2009
Julie W. Farmer
Masters of Science in Dental Public Health
Graduate Department of Dentistry
University of Toronto
2015
Abstract
Objectives: To compare the magnitude of, and contributors to, income-related inequalities in
oral health outcomes within and between Canada and the United States over time. Methods:
The Concentration Index (CI) was used to estimate income-related oral health inequalities from
two Canadian and two American cross-sectional surveys, and decomposed to determine the
potential contributors to inequalities. Results: Pro-poor inequalities were observed in oral
disease outcomes and pro-rich inequalities in oral health outcomes, with greater magnitude in the
United States. Decreases in inequalities for edentulism and increases in decayed teeth were
observed in both countries over time. Inequalities in filled teeth decreased in the United States
and increased in Canada. Socioeconomic characteristics contributed greater to inequalities than
demographic characteristics, with greater contributions of income over time. Conclusions: Oral
health inequalities have persisted over time in Canada and the United States and are associated
with age, sex, education, and have varied over time.
iii
Acknowledgments
This project would not have been possible without the advice and support from many people. First
and foremost, I would like to acknowledge those who helped bring this research into fruition. I am
greatly indebted to my thesis supervisor, Dr. Carlos Quiñonez, for his tremendous supervision,
guidance, and support on this project, and for providing me with opportunities to grow as a
researcher over the past two years. I would like to acknowledge the members of my MSc. thesis
advisory committee for their support; Dr. Logan McLeod for his invaluable methodological and
statistical support, as well as Dr. Arjumand Siddiqi and Dr. Vahid Ravaghi for sharing their
expertise and insight on this research. I would also like to give thanks to Joanna Jacobs and the
team at the Statistics Canada Research Data Centre at the University of Toronto.
Secondly, I would like to thank my family for their love and support throughout this endeavor, and
to Mike and my friends for helping maintain a balance between my work and personal life.
My sincerest thanks to my colleagues in dental public health: Dr. Jodi Shaw, Dr. Faahim Rashid,
Dr. Sonica Singhal, Dr. Noha Gomaa, and Dr. Sojung Lee. I am grateful to have met each and
every one of you and to have developed friendships that will last far beyond the duration of this
program.
Finally, I would like to thank the Canadian Foundation for Dental Hygiene Research and
Education(CFDHRE), the Dr. George and Nancy Vasiga Scholarship in Dental Public Health, and
the Canadian Association of Public Health Dentistry(CAPHD) Dr. James Leake Student Bursary
for their financial support.
iv
Table of Contents
Abstract ........................................................................................................................................... ii
Acknowledgments........................................................................................................................... ii
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
List of Appendices ......................................................................................................................... ix
Introduction .................................................................................................................................1
1.1 Statement of the Problem .....................................................................................................2
1.2 Central Research Question ...................................................................................................3
1.3 Objectives ............................................................................................................................3
1.4 Conceptual Framework ........................................................................................................3
Background .................................................................................................................................6
2.1 Dental Care Systems in Canada and the United States ........................................................6
2.1.1 Financing of Dental Care .........................................................................................6
2.1.2 Dental Insurance Coverage ......................................................................................8
2.1.3 Dental Networks & Reimbursement Systems..........................................................8
2.1.4 Dental Service Delivery Environment .....................................................................9
2.2 Societal and Economic Contexts .......................................................................................11
2.3 Oral Health Inequalities .....................................................................................................14
2.3.1 Oral Health Inequalities in Canada ........................................................................14
2.3.2 Oral Health Inequalities in the United States .........................................................14
2.3.3 Comparing Inequalities between Canada and the United States ...........................15
2.4 Using the Concentration Index to Measure Oral Health Inequalities ................................16
Methodology .............................................................................................................................20
3.1 Design Overview ...............................................................................................................20
v
3.2 Data Sources ......................................................................................................................20
3.2.1 Nutrition Canada National Survey 1970-1972 ......................................................21
3.2.2 Canadian Health Measures Survey 2007-2009 ......................................................21
3.2.3 Health and Nutrition Examination Survey 1971-1974 ..........................................22
3.2.4 National Health and Nutrition Examination Survey 2007-2008............................23
3.3 Variables ............................................................................................................................24
3.3.1 Outcome variables .................................................................................................24
3.3.2 Socioeconomic status .............................................................................................27
3.3.3 Control and predictor variables ..............................................................................27
3.4 Analysis..............................................................................................................................29
3.4.1 Income Quintiles ....................................................................................................30
3.4.2 Indirect Standardization .........................................................................................30
3.4.3 Concentration Curves and Concentration Index ....................................................31
3.4.4 Decomposition of the Concentration Index ...........................................................33
3.4.5 Weighting of Data ..................................................................................................34
3.4.6 Significance Testing...............................................................................................34
3.5 Ethical Considerations .......................................................................................................34
3.6 Data Limitations.................................................................................................................35
Results .......................................................................................................................................36
4.1 Sample characteristics ........................................................................................................36
4.2 Concentration Indices ........................................................................................................39
4.2.1 Redistribution of the Concentration Index .............................................................40
4.3 Decomposition Analysis ....................................................................................................41
4.3.1 Canada 1970-1972 .................................................................................................41
4.3.2 Canada 2007-2009 .................................................................................................44
vi
4.3.3 United States 1971-1974 ........................................................................................47
4.3.4 United States 2007-2008 ........................................................................................50
4.4 Summary Points .................................................................................................................53
Discussion .................................................................................................................................56
5.1 Key findings .......................................................................................................................56
5.1.1 Oral Health Inequalities in Canada and the United States .....................................57
5.1.2 Age and Oral Health ..............................................................................................60
5.1.3 Sex and Oral Health ...............................................................................................61
5.1.4 Income and Oral Health .........................................................................................61
5.1.5 Education and Oral Health .....................................................................................63
5.2 Recommendations ..............................................................................................................63
5.3 Limitations .........................................................................................................................66
5.4 Concluding Remarks ..........................................................................................................66
References ......................................................................................................................................68
Appendix A: Statistics Canada Microdata Research Contract ......................................................77
Appendix B: Characteristics of surveys used in analyses. .............................................................78
Appendix C: Changes in outcome based on tooth counts (NCNS 1970-1972). ............................79
Appendix D: Changes in outcome based on income variable (NHANES 2007-2008). ................82
Appendix E: Significance Testing .................................................................................................86
vii
List of Tables
Table 1 Comparative framework to analyze oral health inequalities. .......................................... 13
Table 2. Consistency of common variables collected through different surveys (outcomes). ..... 26
Table 3. Consistency of common variables collected through different surveys
(explanatory/control variables). .................................................................................................... 28
Table 4. Description of variables used in analysis. ....................................................................... 29
Table 5. Income Quintiles. ............................................................................................................ 30
Table 6. Analysis sample characteristics (%). .............................................................................. 36
Table 7. Oral health outcomes of sample population. .................................................................. 38
Table 8. Observed and Expected Concentration Indices.1,2 .......................................................... 39
Table 9. Concentration Indices. .................................................................................................... 39
Table 10. Percentage redistribution requirements (%). ................................................................ 40
Table 11. Decomposition Results, Canada 1970-1972. ................................................................ 42
Table 12. Decomposition Results Canada 2007-2009. ................................................................. 45
Table 13. Decomposition Results United States 1971-1974. ....................................................... 48
Table 14. Decomposition Results United States 2007-2008. ....................................................... 51
viii
List of Figures
Figure 1. Andersen's emerging model of health care services (Andersen, 2008) ........................... 4
Figure 2. Operational Model - adapted from Andersen’s emerging model of health care services
(Andersen, 2008)............................................................................................................................. 5
Figure 3. Dental financing typologies (Adapted from Burau & Blank, 2006) ............................... 6
Figure 4. Lorenz curve .................................................................................................................. 17
Figure 5. Line of equality and concentration curve ...................................................................... 31
Figure 6. Concentration indices by explanatory variable, Canada 1970-1972 ............................. 43
Figure 7. Aggregate contribution to income-related inequality, Canada 1970-1972 ................... 44
Figure 8. Concentration indices by explanatory variable, Canada 2007-2009 ............................. 46
Figure 9. Aggregate contribution to income-related inequality, Canada 2007-2009 ................... 47
Figure 10. Concentration indices by explanatory variable, United States 1971-1974 .................. 49
Figure 11. Aggregate contributions to income-related inequality, United States 1971-1974 ...... 50
Figure 12. Concentration indices by explanatory variable, United States 2007-2008 .................. 52
Figure 13. Contribution to income-related inequality, United States 2007-2008 ......................... 53
ix
List of Appendices
Appendix A: Statistic Canada Microdata Research Contract 77
Appendix B: Characteristics of surveys used in analysis 78
Appendix C: Changes in outcome based on tooth count (NCNS 1970-1972) 79
Appendix D: Changes in outcome based on income variable (NHANES 2007-2008) 82
Appendix E: Significance Testing 86
1
Introduction
Oral diseases have been shown to negatively impact an individual’s quality of life, leading to
impaired chewing, decreased appetite, sleep problems, and poor school and work performance
(Sheiham, Conway, & Chestnutt, 2015). In addition, treatment of preventable oral disease, such as
dental caries, can have a significant financial impact at the individual and societal level
(Department of Health and Human Services [DHHS], 2000). For example, emergency department
visits for preventable dental conditions produce substantial drainage of resources from the health
care sector. In the US, it is estimated that oral diseases are the fourth most expensive disease to
treat, with curative dental care producing a significant economic burden for many developed
countries (DHHS, 2000). Most importantly, it is well known that oral diseases disproportionately
affect lower income individuals.
Allin and colleagues (2007) suggest that equality in health relies on three principles: equal health
outcomes, equal access to health care for those in equal need of health care, and equal utilization
of health care for those in equal need of health care. In this context, socioeconomic status is
regarded as an important determinant of dental care utilization and oral health, where lower income
individuals often receive less treatment leading to worse oral health outcomes than their higher
income counterparts (Ravaghi, Quiñonez, & Allison, 2013b). Addressing inequalities in oral health
requires a complete understanding of the distribution of oral health or disease along the income
gradient, and identification of how different factors contribute to these inequalities, over time.
Evidence suggests that societal and environmental factors, as well as individual characteristics,
can influence inequalities in health (Mackenbach, 2003). These include the level of health
expenditure, insurance coverage of health care, public/private delivery mix, accessibility, and
extent of inter-sectorial policies (Mackenbach, 2003). By comparing health outcomes between
countries with different health care, social, and political systems, this enables an understanding of
how societal and environmental factors may contribute to health inequalities; such analyses have
been conducted in health and dental literature (Bhandari, Newton, & Bernabe, 2015; Guarnizo-
Herreno, Watt, Pikhart, Sheiham, & Tsakos, 2013; Siddiqi, Ornelas, Quinn, Zuberi, & Nguyen,
2013).
2
With similarities in the social, economic, and historical contexts in Canada and the United States,
as well as differences in social and health policies, it has been suggested that comparing these two
countries has significant implications for understanding how such differences shape inequalities
in health (Prus, 2011; Siddiqi & Hertzman, 2007). Comparative analyses have previously been
performed between these countries using the Joint Canada-United States Survey of Health
(JCUSH); findings from these studies identify how societal differences have contributed to
inequalities in self-rated health among individuals of different sociodemographic and
socioeconomic characteristics (Prus, 2011; Siddiqi, Ornelas, et al., 2013). Further, longitudinal
analyses of health outcomes between Canada and the United States have identified how changes
societal factors, such as the degree of income inequality, equality in the provision of social goods,
and extent of social cohesiveness influence health inequalities over a 20-year period (Siddiqi,
Kawachi, Keating, & Hertzman 2013).
In this regard, while previous studies have examined the differences in clinical oral health
outcomes among low, middle, and high income Canadians and Americans at different time
periods, they have not examined the distribution of these outcomes across the income gradient or
the contribution of demographic or socioeconomic characteristics to inequalities (Elani, Harper,
Allison, Bedos, & Kaufman, 2012). Moreover, there is need to determine and compare outcomes
between and within countries over time in order to more fully understand the potential effects of
societal and environmental factors on oral health inequalities. Given the similarities and
differences, as well as changes in labour markets and oral health care systems in Canada and the
United States over the past 40 years, there is an opportunity to determine whether these changes
have impacted income-related inequality in oral health outcomes over this time period.
1.1 Statement of the Problem
The magnitude of income-related oral health inequalities has not been previously compared within
and between Canada and the United States over time. There is also no knowledge of the potential
influence that changes in social policy, dental care markets, and dental care systems have had on
the magnitude of income-related oral health inequalities in both countries since the 1970s.
3
1.2 Central Research Question
What is the magnitude of income-related oral health inequalities in Canada and the United States
in the 1970s and 2000s?
1.3 Objectives
i. To compare the magnitude of oral health inequalities in Canada in the 1970s and 2000s.
ii. To compare the magnitude of oral health inequalities in the United States in the 1970s and
2000s.
iii. To compare the magnitude of oral health inequalities between Canada and the United States
over time.
iv. To determine the contributors to income-related oral health inequalities in Canada and the
United States.
v. To determine changes in the contributors to income-related oral health inequalities in
Canada and the United States.
1.4 Conceptual Framework
An adaptation of Andersen’s emerging model of health care services served as the conceptual
framework for this study. Figure 1 outlines the four interconnected components of this model:
contextual and individual characteristics, health behaviours, and outcomes (Andersen, 2008).
Previous adaptations of the Andersen model have been used to examine oral health outcomes,
using three individual characteristics: (i) predisposing factors that exist prior to illness, (ii)
enabling factors that affect the availability and accessibility of resource, and (iii) need factors that
are either perceived or clinically determined (Andersen, 1995; Baker, 2009; Ramraj, Azarpazhooh,
Dempster, Ravaghi, & Quiñonez, 2012; Thompson, Cooney, Lawrence, Ravaghi, & Quiñonez,
2014). It has also been adapted in previous dental research to aid in determining sociodemographic
variables related to the utilization of dental services (Ramraj et al., 2012; Thompson et al., 2014).
4
Figure 1. Andersen's emerging model of health care services (Andersen, 2008)
Andersen’s emerging model suggests that health practices and health utilization can influence both
perceived and evaluated health outcomes, as well as satisfaction of care (Andersen, 1995). As
shown in Figure 1, feedback loops are used to show the interrelatedness of all components in the
production of health outcomes (Andersen, 1995).
Figure 2 details the modified model used as the conceptual framework for this research. The intent
of this research is not to test the components of this model, but rather to provide a basis for
understanding health outcomes (perceived and evaluated) and their relation to demographic,
socioeconomic, behavioural, and attitudinal determinants, as well as the environment (Andersen,
1995).
5
Figure 2. Operational Model - adapted from Andersen’s emerging model of health care services
(Andersen, 2008)
The adapted model serves as a guide to examine how different dental care systems (Canada and
the United States), their corresponding environments (social and political), and individual
characteristics (predisposing, enabling, and need factors) affect clinical oral health outcomes along
the socioeconomic gradient in both countries over time. As the purpose of this study involves
quantifying the magnitude of income-related oral health inequalities in Canada and the United
States at different time periods, as well as determining contributors to such inequalities, this
framework also aims to identify variables that may contribute to these inequalities.
6
Background
Since the 1970s, inequalities in oral health have been identified in both Canada and the United
States (Elani et al., 2012). Oral health inequalities may not only be shaped by the oral health care
system, but also by the social and political factors affecting one’s ability to pay for and access
dental care services. Understanding how oral health inequalities have persisted over the past four
decades requires a review of the changes in dental care financing, delivery, and organization in
Canada and the United States, as well as a review of measurement tools used to quantify oral health
inequalities. Thus, a brief historical review of dental care systems in Canada and the United States
is presented below.
2.1 Dental Care Systems in Canada and the United States
2.1.1 Financing of Dental Care
There are three main typologies of financing health care systems that are recognized internationally
(Figure 3). These typologies aid in explaining how dental care is financed in Canada and the United
States. The basic sources for financing dental care systems are taxation (general taxes or specified
health tax), insurance (paid by individuals and/or employers), and direct payment from individuals
(Burau & Blank, 2006; Burt & Eklund, 2005).
Figure 3. Dental financing typologies (Adapted from Burau & Blank, 2006)
Social equity models of financing aim to provide universal coverage, where funds are obtained
primarily through taxation; on the other end, patient sovereignty models often obtain financing
through employer or individual purchase of private insurance or out-of-pocket payments (Burau
& Blank, 2006; Burt & Eklund, 2005). Dental care services in Canada and the United States
represent some combination of the above, yet are predominately privately delivered and financed
7
either through private insurance or out-of-pocket payment (US DHHS, 2000; Health Canada,
2010). As will be detailed later, both countries have some form of public dental programming
available to ensure that certain vulnerable populations, who are not able to afford dental care
services, can access care.
At the time of the first national surveys in the 1970s, dental care systems in Canada and the United
States exhibited variations from what is seen today. Canadian spending on dental care over the
past 40 years has risen. In the 1970s, $56 million was spent on dental care, with 20% of all costs
attributed to public dental care programming (Quiñonez et al., 2007); this has risen to
approximately $703 million in 2010. As of 2010, approximately 5.3% of dental expenditures were
publicly financed, with 52.1% and 42.6% financed through private insurance and out-of-pocket
payments, respectively (Canadian Institute for Health Information, 2012).
In the United States, dental care was predominately paid for through out-of-pocket payments in
the 1960s and 1970s (U.S. DHHS, 2000). Payment through private insurance began to increase in
the mid-1980s, with private dental insurance and out-of-pocket payments dominating the financing
of dental care today (U.S. DHHS, 2000). Over time, public share of dental care has remained low,
with all states providing coverage to child Medicaid recipients, yet only 45 states providing some
form of dental coverage to adult Medicaid recipients (McGinn-Shapiro, 2008). In 2011, of the total
dental care expenditures in the United States, 48.6% was paid through private dental insurance,
41.6% through out-of-pocket payments, and 9.3% publicly (Centers for Medicare and Medicaid
Service, 2013).
When comparing against ten countries of the Organisation for Economic Co-operation and
Development (OECD), the mean per capita spending on dental care in Canada was $300.5, which
was the second highest rank after the United States in 2009 (Organisation for Economic Co-
operation and Development [OECD], n.d). While mean per capita spending has increased, the
public financing of dental care has remained stagnant since 2000 in Canada. Among OECD
countries, Canada had the second lowest percentage of dental care paid through the public sector
(5.4%) (Ramraj, Weitzner, Figueiredo, & Quiñonez, 2014). The United States faired higher in
mean per capita spending ($333.3) and percent provided by public financing (9.5%), but was also
third lowest for the public financing of dental care overall (Ramraj et al., 2014). Nevertheless, the
8
public financing of dental care in the United States has increased by 3.5 percentage points since
the 2000s, yet has remained relatively stagnant in Canada (Ramraj et al., 2014).
2.1.2 Dental Insurance Coverage
Dental insurance operates by decreasing the cost incurred by the recipient through a reduction in
price paid at the point of service (Baldota & Leake, 2004). Public dental care programs in Canada
and the United States are often tied to provincial and state welfare programs, where the amount of
coverage for dental care is dependent on the availability of transfer payments from federal to
provincial/state governments and individual jurisdictions. Although on a broader timeline the
public share of dental service costs in the United States has plateaued over time, there have been
recent improvements in the availability of coverage for dental care, primarily through increases in
funding for the national Child Health Insurance Program (CHIP) and Medicaid programs
(Cosgrove, 2008; Edelstein, 2010).
Dental coverage for Medicaid recipients is only mandated for children, where individual states are
responsible for determining and setting coverage to other socially marginalized groups, such as
low-income non-elderly adults and pregnant women (McGinn-Shapiro, 2008). As mentioned
previously, 45 states have some form of dental coverage for adults under Medicaid, with the
majority of those only covering emergency or limited dental services (McGinn-Shapiro, 2008).
For adults aged 21-64 in 2007, 60% had private dental insurance, 5% had public insurance and
35% had no form of dental insurance coverage (McGinn-Shapiro, 2008).
Similarly in Canada, there is no national mandate on government responsibility for covering dental
care services specifically for social assistance recipients or children; however, all provinces and
territories have some form of dental coverage for these vulnerable groups with jurisdictional
variations in coverage and eligibility (Quiñonez et al., 2005). Today, 6% of Canadians are publicly
insured, 62% have private insurance and 32% rely on out-of-pocket payments or self-insurance
(Health Canada, 2010).
2.1.3 Dental Networks & Reimbursement Systems
Systems of reimbursement for dental professionals can vary within and between countries. In
general there are three basic systems of payment: fee for service (paid a fee for each item of work
provided); capitation (payment based upon the number of patients registered under care); and
9
salary (employers pay an annual income for the services provided by practitioners). The system of
reimbursement may be defined by a network or relationship providers have with their patients
(Burt & Eklund, 2005).
There are known advantages and disadvantages of different mechanisms for reimbursing dental
providers, all of which have implications for potential care received by patients. Fee-for-service
mechanism are known for providing treatment-focused approaches that may lead to over-treatment
and little focus on prevention; salaried and capitation reimbursement often lead to under-treatment
due to the lack of financial incentives to provide care (Burt & Eklund, 2005).
In Canada, providers are predominately reimbursed through fee-for-service mechanisms, where
patients may choose their provider. In the United States, mixed methods of networks and
reimbursement systems have existed, such as managed care and indemnity programs. In managed
care systems, individuals are able to purchase into a care plan organized by a health care company
that includes dental care provided by contract dentists. In the 1990s, the dental market began to
change in the United States as well, with employers offering managed care dental reimbursement
systems such as health maintenance organization (HMO), dental preferred provider organizations
(PPO), and dental referral networks, all of which cannot operate under fee-for-service
reimbursement (U.S. DHHS, 2000). In the United States, with managed care reimbursement
systems, patients receive care through an approved list of providers, whereas those that are covered
under indemnity insurance programs often have more freedom in provider choice (U.S. DHHS,
2000).
In Canada, most patients have the freedom to choose their provider, regardless of their type of
coverage. Canada’s private dental insurance system has remained as a fee-for-service system with
fluctuations in limits on annual maximums and services as well as increases in deductibles,
coinsurance, and co-payments over the same time period (Quiñonez & Grootendorst, 2011;
Quiñonez et al., 2005).
2.1.4 Dental Service Delivery Environment
Dental care in Canada and the United States is predominantly delivered in private practice settings.
In Canada, approximately ninety-two percent of dental care delivery occurs through traditional
practice settings with some public health clinics available but sparse across provinces (Canadian
10
Dental Association, 2010). Albeit minimally, in more recent years, dental care delivery has shifted
to non-traditional practices such as residential care homes, long term care facilities, and mobile
clinics through delivery by dentists and dental hygienists.
Similar to Canada, public dental clinics have been sparse in comparison to private dental clinics
in most communities in the United States (U.S. DHHS, 2000). In addition, there have been recent
changes in legislation to allow for non-dental personnel in non-dental settings to provide
preventive dental care to Medicaid recipients, such as fluoride varnish applications, which has
been initiated in 34 states as of 2009 (Mandal, Edelstein, Ma, & Minkovitz, 2014).
The variation in the supply of dentists and allied dental professionals in Canada and the United
States may also impact oral health outcomes. In the 1960s, dental supply shortages were well
recognized in the United States; reimbursement programs were implemented in the mid to late
1970s for recent graduates by the federal government as a means to increase supply in underserved
areas (Wall, 2012). Despite these efforts, challenges continue to persist in retaining dentists in rural
and remote areas today (U.S. DHHS, 2000). Further, the use of allied dental professionals, such as
dental hygienists, in providing access to care was and still is limited to providing services in-office,
under the supervision by a dentist in most states. Although discussions began in 1960s of the role
of dental auxiliary in increasing access to care, these initiatives have remained dormant until
recently where certain states now have authorized independent dental hygienist licenses
(Waldman, 1980).
A different story emerged in Canada, with the use of dental therapists in two Prairie Provinces and
northern communities beginning in 1972 (Canadian Dental Therapists Association, 2012;
Quiñonez et al., 2007). Although the focus was children, the provision of care allowed for
increased access to preventive and basic restorative services for those who could not seek regular
care primarily due to geographical barriers. Unfortunately though, the dental therapy model has
waned in recent years and is at risk of failing, as most dental therapists now work in private dental
practice under the supervision of dentists in Newfoundland and Labrador (Uswak & Keller-
Kurysh, 2012). Further, with changes to legislation, dental hygienists now have the authority to
work independently from dentists beginning in 1995 in Canada, and in 1998 in the United States
(Mertz, 2008).
11
Overall, a review of the oral health care systems and changes to these systems in Canada and the
United States provides insight into the potential for influence on income-related inequality in oral
health over time. For example, given the reliance on fee-for-service and employment-based
insurance schemes, it may be suspected that lower income individuals may have a greater share of
oral disease than higher income counterparts and the greatest barriers to care, with little change
over time. However, this may be mitigated by the potential increased access to preventive oral
health care services and improvements in oral health behaviours over time.
2.2 Societal and Economic Contexts
Changes to the societal and economic conditions of Canadian and American populations may also
shape inequalities in oral health. The ability of an individual to afford dental care is determined by
the price of the service and/or the amount of disposable income available to be able to pay for a
service (Douglass & Cole, 1979). In turn, the economic characteristics within a society, such as
income distribution and job status, may affect access and obtainment of optimal oral health care.
In the 1970s and over time, Canada has predominately been more equal in terms of income
inequality compared to the United States (Table 1). However, income inequality in Canada has
risen, while the United States has sustained a steady rise since the 1970s (Table 1). This may
suggest that, with changes in the cost of care, lower income individuals may have experienced
more financial barriers to care over time.
As dental insurance is often linked to employment-based benefits, changes in labor markets may
also affect one’s ability to access oral health care. In both countries in the late 1980s, the number
of standard full-time jobs started to decrease, leading to more part-time, temporary job positions
that often provided no health benefits, especially dental insurance (Bhatti, Rana, & Grootendorst,
2007; Blumberg & Holahan, 2004). Further, due to differences in social safety nets for unemployed
populations, and their associated health benefits, unemployment rates in both countries may also
impact income-related oral health inequalities. In the 1970s, higher unemployment rates were
exhibited in the United States (8.5%) compared to Canada (6.9%), and by 2008, unemployment
rates fell to relatively equal size in both countries (6.1% and 5.8%, respectively) (Table 1). Again,
due to changes in government funded dental programs and differences in eligibility requirements
12
for low income and unemployed populations, changes in labor markets may have affected the
number of individuals that are able to afford dental care through out of pocket payment.
The level of educational attainment has also been attributed to differences in oral health outcomes
between populations (Schwendicke et al., 2015). This may be related to differences in lifestyle
choices among those of differing educational backgrounds (Galobardes, Lynch, & Smith, 2007).
Importantly, measures of educational attainment between the Canada and the United States reveal
lower high school completion rates in Canada (37.7%) than the United States (64.1%) in the 1970s,
with an overall rise in completion rates by 2006 (84.6% and 85.5%, respectively) (Table 1). Table
1 provides a comparative framework outlining changes to oral health care systems, as well as social
and economic conditions in Canada and the United States. Overall, comparing societal and
economic changes in Canada and the United States over time provides insight into how these
factors may have influenced inequalities in oral health.
13
Table 1 Comparative framework to analyze oral health inequalities.
Canada United States
1970s 2000s 1970s 2000s
Oral health systems
Major source of financing
dental care
Out of pocket (OOP) and
private insurance OOP and private insurance OOP and private insurance OOP and private insurance
Dental Insurance Coverage a
62% privately insured
6% publicly insured
32% uninsured
a
60% privately insured
5% publicly insured
35% uninsured
Dental networks &
reimbursement systems Open
Fee-for-service
Open
Fee-for-service
Open
Fee-for-service
Open and Managed Care
Mix of fee-for-service and
capitation
Service delivery
environment Private practice
Predominately private
practice with some non-
traditional practice
Private practice
Predominately private
practice with some non-
traditional practice
Societal and Economic Contexts
Income distributionb,c
Gini (G): 0.304
P90/P10: 4.1
(1976)
G: 0.321
P90/P10: 4.1
(2008)
G: 0.316
P90/P10: 4.8
(1974)
G: 0.378
P90/P10: 5.9
(2008)
Employment Statusb
Full-time
Unemployment rate
(UR): 6.9% (1975)
Non-standard
UR: 6.1% (2008)
Full-time
UR: 8.5% (1975)
Non-standard
UR: 5.8% (2008)
Education
(High school completion of
population >25)
37.7 (1976)e
84.6 (2006)f
64.1 (1976)g
85.5 (2006)g
a Information not available b OECD.Stats. 2015. Income Distribution Database. Retrieved from: http://www.oecd.org/std c Gini coefficient of disposable income post (taxes and transfers); P90/P10 disposable income decile ratio d OECD.Stats. 2015. Unemployment rate aged 15 and over, all persons. Short-term Labour market Statistics. Retrieved from http://www.oecd.org/stde e Statistics Canada. 1976. Population: Demographic Characteristics. Level of Schooling by Age Groups. 1976 Census of Canada. Catalogue 92-827. Bulletin 2.8, Table 30 f Statistics Canada. 2006. Population: Demographic Characteristics. Level of Schooling by Age Groups. 2006 Census of Canada. Catalogue no. 97-564-XCB2006009 g US Census Bureau. 1974-2002. March Current Population Survey 2003-2014. Annual Social and Economic Supplement to the Current population survey.
http://www.census.gov/hhes/socdemo/education/data/cps/index.html
14
2.3 Oral Health Inequalities
2.3.1 Oral Health Inequalities in Canada
Inequalities in access to dental care and oral health care outcomes have been identified previously
in Canada. Allin (2008) found pro-rich inequity in the probability of a dental visit across all
provinces, indicating income as a predictor of inequalities in dental care visits. Ravaghi and
colleagues (2013) identified that the poor often receive less preventive treatments and postpone
curative treatments, potentially leading to the development of more severe oral health problems
and more untreated decay in adult Canadian populations. They also found sex differences for
clinical indicators of oral health by using the concentration index with data obtained from the
2007-2009 Canadian Health Measures Survey (CHMS); low-income women accounted for more
of the burden of caries, missing teeth, and oral pain than their male counterparts (Ravaghi et al.,
2013b). These findings suggest that factors related to income-related oral health inequalities vary
in distribution among socioeconomic groups between men and women. Inequalities in accessing
dental care have also been reported for low- and middle-income populations, with cost often cited
as the main barrier to obtaining care (Ramraj, Sadeghi, Lawrence, Dempster, & Quiñonez, 2013).
The prevalence of oral diseases, such as dental caries, periodontal disease, and edentulism, have
reportedly declined since the 1970s (Health Canada, 2010). However, several subgroups within
Canada still experience a greater share of the burden of oral disease and lower inability to access
to dental care; these include aboriginal populations, Canadians from lower and middle-income
families, the working poor and those on remote and isolated areas (Federal, Provincial, and
Territorial Dental Working Group [FPTDWG], 2013). In general, Canadian studies have
identified potential associations with education, complementary insurance, and region in inequities
in the utilization of dental care and oral health outcomes (Allin, 2008).
2.3.2 Oral Health Inequalities in the United States
Inequalities in oral health in the United States have been well documented. Ongoing surveillance
at the national level through the National Health and Examination Survey has enabled monitoring
of oral health trends since the 1960s. Results from these measures have identified inequalities in
oral health, despite declines in the presence of oral diseases. They reveal that lower income adults
are twice as likely to have worse oral health than adults living in non-poor households (Bloom,
15
Simile, Adams, & Cohen, 2012). In addition, adults with Medicaid are more likely than uninsured
adults to experience toothaches or have missing teeth or broken fillings (Bloom et al., 2012).
The absolute prevalence difference in edentulism between low and high socio-economic positions
has remained unchanged over the last three decades in the United States (Cunha-Cruz, Hujoel, &
Nadanovsky, 2007). In addition, tooth retention tends to vary by poverty level with complete tooth
retention higher for adults living above the poverty line than those living at or below the poverty
line among adults aged 25-64 (Dye et al., 2007). In addition, wealth effects on dental care
utilization have been identified with utilization of dental care tending to decrease with a decline in
income and/or wealth (Manski et al., 2012).
2.3.3 Comparing Inequalities between Canada and the United States
In both countries, inequalities have been defined as poorer oral health outcomes in groups of lower
socioeconomic status, with socioeconomic status regarded as an important determinant of oral
health, dental care utilization, and access to dental care. Horizontal inequalities – where individuals
with equal treatment needs do not have equal access – have been observed in the Canada and the
United States, and in countries worldwide (Grignon, Hurley, Wang, & Allin, 2010; Guarnizo-
Herreno, Watt, Pikhart, Sheiham, & Tsakos, 2013; A. E. Sanders et al., 2009; Somkotra &
Detsomboonrat, 2009). Comparisons of oral health inequalities between countries have been
primarily performed across European countries (Bernabe & Sheiham, 2014; Bhandari, Newton, &
Bernabe, 2015; Guarnizo-Herreno, Tsakos, Sheiham, & Watt, 2013; Guarnizo-Herreno, Watt, et
al., 2013; Guarnizo-Herreno, Watt, Pikhart, Sheiham, & Tsakos, 2014; Listl, 2015; R. Manski et
al., 2015). Indeed, to date, only one study has examined inequalities in oral health between Canada
and the United States.
Here, Elani and colleagues (2012) compared differences in clinical oral health outcomes between
Canadians and Americans by using absolute and simple measures of inequality. They examined
how the prevalence of oral health or disease varied within and between these populations by
income, place of birth, and education. They found that differences among edentulism by income
status were more prevalent in Canada than in the United States. Over time, declines were observed
in edentulism and absolute socioeconomic inequalities in Canada and the United States. There was
also better progress in narrowing absolute inequality among place of birth, education and income
in Canada in comparison to the United States (Elani et al., 2012). These inequalities may likely be
16
influenced by the increased social inequalities of the past 20 years due to processes of economic
crisis and reorganization of social welfare policies in Canada and the United States (Wall, Vujicic,
& Nasseh, 2012). However, the extent to which individual- and societal-level factors influence
these inequalities have not been previously explored.
2.4 Using the Concentration Index to Measure Oral Health Inequalities
Measuring health inequalities provides a way to describe the distribution of health or disease
within a population, as well as to monitor changes in distributions over time (Galobardes, Lynch,
& Smith, 2007). Information obtained from these measures can be used to inform and evaluate
health policies. However, it is important to note that the approach to measuring inequalities should
be dependent on the objectives to be pursued (Mackenbach & Kunst, 1997).
Simple measures of inequalities, such as differences between observed frequencies of health across
different income groups, provide a straightforward interpretation of identifying inequalities within
a population (Regidor, 2004b). Despite the simplicity of comparing differences in observed health
measures, they are limited to reporting differences in outcomes between subgroups in populations
and do not illustrate the share of health across different socioeconomic groups within a population
(Regidor, 2004b). Therefore, these measures do not truly estimate the disproportionality of disease
or health across differing levels of socioeconomic status.
More complex measures incorporate the socioeconomic distribution within a population into
measures of health inequalities; these include the concentration index, the relative index of
inequality, and the slope index of inequality, to name a few (Regidor, 2004a). The concentration
index is considered a valuable tool in measuring socioeconomic inequality in health since it
satisfies the following principles: i) it reflects the socioeconomic dimension to inequalities in
health; (ii) it reflects the experience of the entire population; and (iii) it is sensitive to changes in
the distribution of the population across socioeconomic groups (Wagstaff, Paci, & Van Doorslaer,
1991).
The concentration index is a method adapted from the concepts of the Lorenz curve and Gini index.
As identified in Figure 4, the Lorenz curve plots the cumulative proportion of individuals by level
17
of health ranked in increasing order on the x-axis, against the cumulative total proportion of health
within these individuals on the y-axis. The diagonal line identifies the distribution of health if it
were equally distributed across a population. A line that deviates from the diagonal indicates that
health is unequally distributed across individuals, such that some individuals have more health
than others (Regidor, 2004b).
Figure 4. Lorenz curve
The Gini index is derived from the Lorenz curve and is a value ranging from 0 (diagonal line) to
1 (health is concentrated in a single person). The concentration index is an adaptation of this
concept as it measures the distribution of health across the distribution of socioeconomic level
within a population. Further information on the concentration index is provided in Chapter 3
(Section 3.4.3).
Health inequalities measured by the concentration index express inequality as a function of
differences between shares of some health outcome compared with shares of the population, which
can be measured across ordered social groups, and therefore reflects the social gradient of disease
(Konings et al., 2010). This allows for comparison of socioeconomic inequality in health over time
and between different places, as when individuals are ordered by socioeconomic level, the size
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100
Cu
mu
lati
ve
hea
lth
(%
)
Cumulative population ranked by health (%)
Line of Equality Lorenz Curve
18
and sign of the concentration index depends only on the gradient observed between socioeconomic
level and health (Regidor, 2004a).
As mentioned previously, while Elani et al. (2012) discovered the narrowing of oral health
inequalities in Canada and the United States over time, the results of this study were limited to
measuring absolute differences in oral health outcomes between the rich and poor, and did not
consider the gradient in health across all social groups (Elani et al., 2012; Jagger, Sherriff, &
Macpherson, 2013). In addition, their results did not reveal heterogeneity in outcomes between
sexes. By using the concentration index, research has found sex differences in oral health outcomes
and age differences in dental caries along the income gradient (Bernabe, Delgado-Angulo,
Murasko, & Marcenes, 2012; Ravaghi et al., 2013b).
The concentration index has also been used in oral health research to examine inequalities in
general health and oral health outcomes, as well as differences in inequalities between clinical and
self-reported oral health outcomes (Ravaghi, Quiñonez, & Allison, 2013a; Ravaghi et al., 2013b;
Shen, Wildman, & Steele, 2013). The advantage of using the concentration index in health
inequalities research is that it can be decomposed to explain the distribution of health by a set of
factors that may systematically vary with socioeconomic status, such as education, insurance
coverage, and sociodemographic factors (O'Donnell Owen, Adam, & Lindelow, 2007).
Decomposition of the concentration index reveals how far inequalities in health can be explained
by the inequalities present in other explanatory variables, which can be used by policy makers
(O'Donnell Owen et al., 2007). The drivers of income-related oral health inequalities have been
previously studied in Canada and the United States.
The contribution of individual- and health care system characteristics have been examined and
compared for self-reported health-related quality of life in Canada and the United States (McGrail,
Van Doorslaer, Ross, & Sanmartin, 2009). Decomposition analyses revealed that income was a
large contributor to income-related health inequalities in both countries and health care system
factors contributed greater to income-related health inequalities in the United States compared to
Canada (McGrail et al., 2009). These findings suggest that through decomposition analysis,
potential contributors to income-inequalities and the differences of these contributors may be
identified between countries.
19
Somoktra & Detsomboonrat (2009) decomposed the concentration index to determine the effect
of demographic and socioeconomic factors as drivers of dental care utilization in Thailand. They
found that income and non-need determinants (geographic location, insurance, education)
principally contributed to the pro-poor public sector utilization, unlike pro-rich private sector
utilization (Somkotra & Detsomboonrat, 2009). Shen and colleagues (2013) decomposed the
concentration index to determine contributors to income-related inequalities in health and oral
health in the United Kingdom (Shen et al., 2013). Through decomposition, their findings revealed
the contributions of age, education, and occupational status to inequalities in self-reported and
clinical measures of oral health in adults in the United Kingdom.
To date, few studies have used and decomposed the concentration index to quantify the extent to
which these potential factors contribute to oral health inequalities in Canada or the United States
(Ravaghi et al., 2013a; Ravaghi et al., 2013b). It can be hypothesized that demographic,
socioeconomic and behavioural factors may have an influence on oral health outcomes in addition
to income. Therefore, further analysis should consider how socio-demographic and lifestyle
choices may influence the distribution of income-related oral health inequalities among income
groups in Canada and the United States.
20
Methodology
3.1 Design Overview
An observational study using data from cross-sectional national datasets from Canada and the
United States at two points in time was performed [to examine the magnitude of, and contributors
to income-related oral health inequalities]. The concentration index (CI) method was used to
determine income-related inequality for three clinical oral health outcomes in each dataset. The CI
was subsequently decomposed to determine contributors to income-related oral health inequalities
for clinical and self-reported oral health outcome variables. Quantitative comparisons were made
through use of unpaired t-tests.
3.2 Data Sources
Data was used from four sources; two American and two Canadian. The two American sources are
the Health and Nutrition Examination Survey 1971-1975 (HANES I) and the National Health and
Nutrition Examination Survey 2007-2008 (NHANES). Data from the two Canadian sources are
the Nutrition Canada National Survey 1970-1972 (NCNS) and Canadian Health Measures Survey
2007-2009 (CHMS).
Information was collected from four data sources. Both datasets from the United States (HANES
I and NHANES) were publicly accessed through the Center for Disease Control and Prevention
data website. The NCNS Canadian dataset was accessed through the Statistics Canada under the
Data Liberation Initiative (DLI) license. Access to the CHMS required approval from the Statistics
Canada Research Data Centre (RDC) in Toronto and accessed through the University of Toronto
RDC (Appendix A).
Analysis focused on individuals aged 20 to 74 years. Those who were less than 20 and over 75
years of age, those who did not complete household questionnaires and/or clinical examinations,
or had missing data were excluded from analysis. The subsequent paragraphs provide a description
of each dataset; further information can be found in Appendix B.
21
3.2.1 Nutrition Canada National Survey 1970-1972
The NCNS was conducted between October 1970 and September 1972. The survey collected
information from 19,590 individuals 0 to >100 years old. The survey attempted to estimate the
prevalence of nutritional diseases in the Canadian population and determine food type and quantity
consumed by Canadians. The survey was designed to collect representative estimates from
metropolitan, urban, rural residents and from “low-income” and other income groups. Indians
living in provinces and territories and Inuit living in four settlements in Northwest Territories were
included in this survey. The sampling allowed for representation of five regions: Atlantic
(Newfoundland, Prince Edward Island, New Brunswick, and Nova Scotia); Quebec, Ontario,
Prairies (Manitoba, Saskatchewan and Alberta); and British Columbia. Indians living on
reservations and crown lands were also sampled from the following regions: Maritime (PEI, New
Brunswick, and Nova Scotia); Quebec; Ontario; Prairie (Manitoba, Saskatchewan and Alberta);
British Columbia; Yukon and Northwest Territories. With the exception of the samples of
expectant women, and phase I of the youth survey, the survey was statistically designed to produce
probability samples. Data was collected over a two year phase and obtained in two stages: (i)
household interview; (ii) clinical examination at the Nutrition Canada Clinic. Household
interviews provided information on general demographic, food handling and food preparation
information. Several clinical measures were used in the Nutrition Canada clinic, including oral
health measures. The Nutrition Canada physician and nurses made notes on medical history and
conducted a medical check-up to determine the state of health of each person who visited the clinic.
The dental examiner performed a similar function in checking the dental health of each participant.
3.2.2 Canadian Health Measures Survey 2007-2009
Data were collected by Statistics Canada between March 1, 2007 and March 31, 2009 and contains
information from household residents, age 6 to 79. Collection occurred in two stages: (i) household
interview, and (ii) clinical examination. The household interview collected information on a
respondent’s demographic characteristics, socioeconomic status, and health behaviours. The
clinical examination collected clinical measures of respondents’ physical health (including an oral
health examination). The clinical examination was conducted in a CHMS mobile examination
22
centre. The oral health examination collected direct physical measurements of oral health using a
mouth mirror and explorer, and calibrated dentists/examiners.
The sample frame divided Canada into 257 potential collection sites, each with a population of
greater than 10,000. The region (British Columbia, Prairies, Ontario, Quebec, Atlantic) and
urban/rural nature of each site was identified and then 15 sites were systematically selected in
proportion to the size of their population. Within each site, households with known household
composition (based on the 2006 census) were divided into six strata to obtain sufficient numbers
of respondents in each of the targeted age groups. A random sample of households from each
stratum was taken. Within a selected household, one or two respondents were selected. All five
regions were representative. Residents of Indian Reserves and Crown Lands, institutions, certain
remote regions or in areas with low population densities, and full-time members of the Canadian
Forces were excluded from the sampling frame.
Of the 8,772 households selected for the CHMS, 69.6% agreed to participate; 88.3% of them
responded to the household interview, and of those, 84.9% visited the mobile examination centre.
The overall response rate was 51.7%. A comprehensive consent process was employed.
Participation was voluntary and respondents could opt out of any part of the survey at any time.
The final CHMS sample size is 5,604 respondents and is representative of approximately 96.3%
of the Canadian population. To account for the CHMS’s complex survey design, Statistics Canada
produces survey weights that represent a survey respondent’s contribution to the target population.
The survey weights are computed using an initial weight representing a respondent’s inverse
probability of selection. The initial weight is then adjusted to account for survey specifics (such as
nonresponse).
3.2.3 Health and Nutrition Examination Survey 1971-1974
The First National HANES I was conducted between April 1971 and June 1974 on nationwide
probability sample of approximately 32,000 persons, ages 1-74 years, from the civilian, non-
institutionalized population of the coterminous United States, excepting those persons residing on
Indian reservations. The HANES I sample was selected so that certain population groups thought
to be at high risk of malnutrition (persons with low incomes, preschool children, women of
23
childbearing age, and the elderly) were oversampled. On completion of the survey, 31,973 sample
persons had been interviewed; of these, 23,808 people were examined. Adjusted sampling weights
were computed within 60 age-sex-race categories in order to inflate the sample so as to closely
reflect the U.S. civilian non-institutionalized population 1-74 years of age at the midpoint of the
survey.
Information about all of the examined persons in HANES I was obtained by means of several
measures including a household interview and dental examination. The dental examiners attempted
to derive their findings uniformly by following a written set of objective standards in which they
had been carefully trained. The standards were guidelines that, in effect, narrowed the range of
examiner variability by eliminating many of the borderline or questionable conditions that are
frequently a source of disagreement.
3.2.4 National Health and Nutrition Examination Survey 2007-2008
The National Health and Nutrition Examination Survey, 2007-2008 (NHANES 2007-2008) was
carried out between January 2007 and December 2008 and contains data for 10,149 individuals 0-
80 years old. First, the eligible sample for the survey and tasks related to survey operations and
data management were performed. The NHANES survey design is a stratified, multistage
probability sample of the civilian non-institutionalized U.S. population.
The stages of sample selection were: 1) selection of Primary Sampling Units (PSUs), which are
counties or small groups of contiguous counties; 2) segments within PSUs (a block or group of
blocks containing a cluster of households); 3) households within segments; and 4) one or more
participants within households. A total of 15 PSUs are visited during a 12-month time period. A
brief description of the data collection procedures follows.
Beginning in 2007, oversampling occurred for the entire Hispanic population, persons 60 and
older, Blacks, and low-income persons. In addition, for each of the race/ethnicity domains, the 12-
15 and 16-19 year age domains were combined and the 40-59 year age minority domains were
split into 10-year age domains 40-49 and 50-59. This has led to an increase in the number of
participants aged 40+ and a decrease in 12-19 year olds from previous cycles. Participants aged 5
years and older were eligible for a tooth count and basic screening examination (BSE).
24
Additionally, persons aged 25 years and older were eligible for a brief denture questionnaire and
functional contacts assessment.
The oral health examination component assessed the prevalence of oral conditions and diseases,
such as edentulism, denture use, dental sealants, and dental caries. A concurrent set of questions
was administered during the household interview to assess issues related to oral health quality of
life. Non-dental professionals who were trained to administer the oral health screening assessments
conducted the NHANES 2007-08 oral health exam.
3.3 Variables
3.3.1 Outcome variables
Table 2 lists clinical and self-reported outcome variables collected from each survey. All national
surveys included in this study included clinical oral health examinations of each survey participant.
In three of the four surveys, individual tooth measures were recorded for the number of decayed
(D), missing (M), filled (F) and sound teeth (T) (DMFT). The DMFT is used to measure the
prevalence and extent of dental caries in a population; it expresses the total number of teeth
affected by caries as a value ranging from 00-32 or 00-28. Components of the DMFT provide
information on current and previous dental disease. Measures of decayed and filled teeth indicate
the level of untreated and treated dental disease and serve as surrogate indicators of oral disease.
Two surveys collected clinical oral health information from tooth counts 00-32; the CHMS
excluded third molar counts in their collection of DMFT. In order to ensure consistency between
surveys, both 00-32 and 00-28 tooth counts were included in analysis, where applicable. Appendix
C provides information on differences in outcomes for tooth counts 00-32 and 00-28 for the NCNS
1970-1972 as it was the only dataset available to make comparisons. As mentioned previously, the
NHANES 2007-2008 survey performed an oral health BSE that did not include individual tooth
counts for decayed and filled teeth and therefore DMFT counts were not reported. The BSE
provided information on the presence or absence of one or more decayed or restored(filled) teeth.
It is a less resource intensive form of assessment and has been used for surveillance in the United
States in more recent years. Fortunately, DMFT counts employed in other surveys can be recoded
25
to produce the similar outcomes: the presence of one or more decayed [or filled] teeth, allowing
for comparisons between surveys.
The presence of edentulism, or number of missing teeth, is a true endpoint of oral disease, and was
clinically recorded in all four surveys. Clinically derived treatment needs were reported in both
Canadian surveys, and were inclusive to prevention, restorative, surgery, periodontal, endodontic,
prosthodontics, and urgent needs. In addition to clinical outcomes, self-reported measures were
reported through household questionnaires conducted through personal interviews. Some
questionnaires contained measures of oral health behaviours (dental visits, hygiene care) and oral
health-related quality of life.
26
Table 2. Consistency of common variables collected through different surveys (outcomes).
Type of
Outcome Variable Description
Canada
1970-1972
Canada
2007-2009
United States
1971-1974
United States
2007-2008
Clinical Edentulism
Binary:
“0” – Dentate
“1” – Edentulous ✓ ✓ ✓ ✓
Clinical Presence of
Decayed Teeth
Binary:
“0” – no decay present
“1” – >1 decayed tooth present;
Continuous: total number of teeth with decay
(00-32 or 00-28)
✓ ✓ ✓ ✓
Clinical History of Dental
Decay
Binary:
“0” – no filled teeth present
“1” – >1 filled tooth present
Continuous: total number of filled teeth (00-32
or 00-28)
✓ ✓ ✓ ✓
Clinical History of Dental
Disease
Continuous: total number of decayed, missing,
and filled teeth (00-32 or 00 – 28) ✓ ✓ ✓
Clinical Treatment status
Binary:
“0” – No treatment needed
“1” – >1 treatment required (excludes
preventive treatment)
✓ ✓
Self-reported Dental visit in the
past 12 months
Categorical:
“0” – Less than 12 months
“1” – within 1 to 2 years
“2” – within 2 to 5 years
“3” – over 5 years ago
✓ ✓
Self-reported
Satisfaction with
appearance of
teeth
Binary:
“0” – satisfied
“1” – dissatisfied ✓ ✓
Self-reported Oral Pain in the
past 12 months
Categorical:
“1” – present
“2” – no oral pain in the past year ✓ ✓
27
3.3.2 Socioeconomic status
Income was used as a proxy measure for socioeconomic status. Alternative indicators such as
educational attainment and occupational status tend to be stable or provide little variation among
adults. In addition, among retirees, a measure of occupation status tends to lose its significance.
All four surveys provide ordinal variables for income. Household and family income were used as
measures of socioeconomic status in our analyses. Total household income and household size
were reported in the NHANES 2007-2008 and CHMS 2007-2009 surveys, whereas total family
income and family size were reported in both surveys conducted in the 1970s.
As the concentration index requires a ranked measure of socioeconomic status, income variables
unique to each survey were used. Due to differences in reporting income in four surveys,
comparisons between household income and family income were made utilizing the NHANES
2007-2008 in order to examine viability of comparisons (Appendix D).
3.3.3 Control and predictor variables
Table 3 outlines all potential control and predictor variables from each survey. Socio-demographic
variables of sex, age, and education were used as controls for analysis of income-related oral health
inequalities. Sex was dichotomized into male and female. For consistency across surveys, age
groups were categorized into children & youth (6-19), young adults (20-39), middle-aged adults
(40-59), and older adults (60-74). Education was reported as the highest level of education
achieved by the head of household and dichotomized to “less than highschool” and “high-school”.
28
Table 3. Consistency of common variables collected through different surveys (explanatory/control variables).
Type of
Outcome Variable Description
Canada
1970-1972
Canada
2007-2009
United States
1971-1974
United States
2007-2008
SES Ranking Income Ordinal ✓ ✓ ✓ ✓
Demographic
and Controls
and Predictors
Sex
Binary:
“0” – Male
“1” – Female ✓ ✓ ✓ ✓
Age Continuous: (0-79) ✓ ✓ ✓ ✓
Education
Binary:
“0” – Less than high school
“1” – High school ✓ ✓ ✓ ✓
Marital
Status
Binary:
“0” – Married
“1” – Single
✓ ✓ ✓
Dental
Insurance
Do you have insurance or a government program
that covers all or part of your dental expenses?
“0” – Private
“1” – Public
“2” – None
✓ ✓
Employment
Status
Categorical:
“0” – Full-time
“1” – Part-time
“2” – Unemployed
“3” – Retired
✓ ✓ ✓
Smoking
Status
Do you smoke cigarettes regularly?
“1” – Yes
“0” – No ✓ ✓
Do you smoke cigarettes daily, occasionally or
not at all?
“1” – Yes
“0” – No
✓
Do you now smoke cigarettes?
“1” – Yes
“0” – No
✓
29
3.4 Analysis
Table 4 outlines the variables used in the analysis. Measures of edentulism, the presence of one or
more decayed teeth, and the presence of one or more filled teeth served as outcome measures.
Dummy variables were produced for all socioeconomic, and control/predictor variables.
Quantitative analysis of income-related oral health inequalities in and between datasets was
conducted through use of STATA MP/dual core Software.
Table 4. Description of variables used in analysis.
Variable Description of codes
Clinical
Outcomes
Edentulism edent
o 0 - Dentate
o 1 - Edentulous
Presence of one or more decayed
teeth
decay
o 0 - No decayed teeth
o 1 - At least one decayed tooth
Presence of one or more filled
teeth
filled
o 0 - No filled teeth
o 1 - At least one filled tooth
Socioeconomic
Ranking
Income
incA (lowest income quintile)
incB (lower middle quintile)
incC (middle quintile)
incD (upper middle quintile)
incE (highest quintile)
Income (rank) Ordinal
Household/Family size hhsize (continuous)
Control/
Predictor
Sex male
female
Age Groups
ageB (20-39years)
ageC (40-59years)
ageD (60-74years)
Age/Sex Interaction
mageB (males 20-39years)
mageC (males 40-59years)
mageD (males 60-74years)
Educational Attainment eduA1 - less than high school
eduA2 - high school graduate
30
3.4.1 Income Quintiles
In order to make comparisons across datasets, income variables were categorized into quintiles
based on each survey’s distribution of income (Table 5).
Table 5. Income Quintiles.
Percent of population
ranked by income
Lowest Income 20
Lower Middle Income 20-40
Middle Income 40-60
Upper Middle Income 60-80
Highest Income 80-100
As the concentration index is based upon individual survey distribution, quintiles were formulated
for each group to capture the true picture of income distribution at that time and place rather than
control for inflation. To examine income gradients, each oral health outcome was compared across
quintiles. Family and household size were used as controls.
3.4.2 Indirect Standardization
When examining income-related oral health inequalities, it is important to control for confounding
effects of other variables, such as age and sex. Indirect standardization generates estimates of oral
health outcomes that would be observed in the population independent of differences in need
variables, such as age, sex, and education (O'Donnell Owen et al., 2007).
Since all oral health outcomes in our analysis were binary, multivariate logit models were
estimated for each outcome on a set of control variables by using equation 1:
Equation 1 𝑦𝑖 = 𝑔(𝛾 + 𝛿𝑥𝑗𝑖 + 𝜆𝑧𝑗𝑖) + 휀𝑖
where yi is observed health, 𝛾 is the intercept, xji is the jth control variable, g() is the logistic
function relating to yi to the linear combination of 𝛾 and xji, zk are non-confounding variables, and
휀𝑖 is the random error term. Need-predicted outcomes were produced from parameter estimates
31
and an individual’s observable characteristics (xji). Indirectly standardized outcomes(𝑦𝑖𝐼𝑆) were
then calculated as the difference between the observed outcome(𝑦𝑖), need-predicted outcome (𝑦𝑖𝑋),
plus the average need-predicted outcome(�̅�) of the sample (Equation 2).
Equation 2 `𝑦𝑖𝐼𝑆 = 𝑦𝑖 + 𝑦𝑖
𝑘 − �̿�
3.4.3 Concentration Curves and Concentration Index
The CI is used to quantify the magnitude of income-related inequality for each health outcome,
which is derived from a concentration curve (CC). A CC plots the cumulative proportion of the
population ranked by socioeconomic status from lowest to highest against the cumulative
proportion of the oral health outcome (Figure 5).
Figure 5. Line of equality and concentration curve
As described by Wagstaff et al. (1991) the x-axis of the concentration curve represents the
cumulative proportion of individuals ranked by socioeconomic status, beginning with those who
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100
Cu
mu
lati
ve
ora
l hea
lth
(%
)
Cumulative population ranked by socioeconomic status (%)
Line of Equality Concentration Curve
32
have the lowest level of socioeconomic status and ending with those who have the highest; the y-
axis represents the cumulative total proportion of oral health (or disease) in a population. A
diagonal line represents the line of equality, where deviation above or below the line indicates
measures of inequality; the farther the CC is from the diagonal the greater the degree of inequality.
For measure of oral health, if the CC falls under the diagonal, oral health is concentrated in those
with higher socioeconomic status; if the curve is above the diagonal line, oral health is concentrated
in those with lower socioeconomic status. A CC that falls under the diagonal is given a positive
value and vice versa.
The CI is derived from the CC in order to quantify the magnitude of inequality in socioeconomic
status among clinical oral health outcomes (e.g., number of decayed teeth). The CI is calculated as
twice the area between the CC and the line of inequality. The CI is a value ranging from -1 to +1.
For measures of oral health, if all oral health is concentrated in the person with the highest
socioeconomic level, the index will have a value of +1, whereas if oral health is concentrated in
the person with the lowest socioeconomic level it will have a value of -1. Therefore, the closer the
CC is to the diagonal and the closer the value is to 0, then the greater the oral health equality for a
given oral health measure.
For binary health outcomes the possible values of the concentration index are limited by the mean
(p) of the distribution and are equal to p-1 and 1-p, respectively (Wagstaff, 2005). As the mean
increases, the range of possible values of the concentration index shrinks, which has implications
for judging the health outcomes of a binary variable. Therefore, to permit comparison of the
concentration index for binary outcomes with those of other outcomes the concentration index can
be normalized so that the bounds will be between -1 and +1 (Wagstaff, 2005).
The CI is derived using the convenient linear regression methods. Observed and expected CIs are
calculated for each oral health outcome. The difference between the observed and expected CIs
are calculated to produce avoidable concentration indices (equation 3). These describe the
magnitude of avoidable income-related inequality in a health variable.
Equation 3 CI = CIob – CIexp
Once derived, CIs were multiplied by 75 to determine the percentage of the outcome variable that
would need to be redistributed from the richer half to the poorer half of the population in order to
33
reach a value of zero (equality) (Koolman & van Doorslaer, 2004). The redistribution scheme is
indicated for large samples that use convenient linear regression methods to compute the CI, where
the percentage to be redistributed is equal to ¾ of the CI (Koolman & van Doorslaer, 2004).
Equation 4 Redistribution = CI*75
For example, a CI of 0.10 would indicate that the health outcome is concentrated among the rich
and that 7.5% (0.10*75) of the health outcome would have to be redistributed to the poor in order
to eliminate inequality.
3.4.4 Decomposition of the Concentration Index
The CIs for each outcome were then decomposed to determine the contributors (e.g., education,
income, oral health practices) to income-related oral health inequalities. A variable’s contribution
to income-related inequality is based on: (i) a variable’s effect on the outcome, and (ii) how
unequal the distribution of a given variable is across income.
A linear regression model that relates the oral health variable, y, to a set of K determinants, xk:
Equation 5 𝑦1 = 𝛼 + ∑ 𝛽𝑘𝑥𝑘𝑖 + 휀𝑖𝑘 ,
where 𝛽𝑘 are coefficients and 휀 is the random error. It is assumed everyone in the selected sample
experiences the same coefficient vector, 𝛽𝑘. Using equation 2 and the relationship between x and
y, the concentration index can be calculated as:
Equation 6 𝐶 = ∑(𝛽𝑘𝑥𝑘̅̅ ̅/𝜇)𝐶𝑘 + 𝐺𝐶𝜀/𝜇
where μ̅ is the mean of 𝑦 , 𝑥𝑘̅̅ ̅ is the mean of 𝑥𝑘, Ck is the concentration index for 𝑥𝑘, and 𝐺𝐶𝜀is
the generalized concentration index for the error term (휀𝑖). The equation shows that C is equal to
a weighted sum of concentration indices of the k regressors, where the weight for 𝑥𝑘 is the
elasticity of 𝑦 with respect to 𝑥𝑘 (𝜂𝑘 = 𝛽𝑘𝑥𝑘̅̅̅̅
𝜇). Elasticites reflect the income-related inequality in
34
health not explained by systematic variation in the regressors by income. Thus, decomposition
looks at the contribution of each control variable to inequality in an oral health outcome.
3.4.5 Weighting of Data
To account for complex survey design and probability sampling, survey weights were reported for
each dataset and are included in all analyses.
3.4.6 Significance Testing
Significance testing is performed in order to test for differences: (1) between concentration indices
and equality (concentration index value of 0); (2) between observed and expected concentration
indices; (3) between observed concentration indices within the same dataset; and (4) between
observed concentration indices for the same outcome between different datasets. Appendix E
provides the results of significance testing.
3.5 Ethical Considerations
This research involves analysis of secondary data from publicly accessible datasets for HANES I
and NHANES 2007-2008 available on the Centers for Disease Control and Prevention website,
and the NCNS 1970-72 available on the Survey Documentation and Analysis website. The CHMS
2007-2009 data set was accessed through a Statistics Canada RDC. For all datasets, personally
identifiable information from survey respondents were not obtained or used for analysis purposes.
Further, the results of the study conformed to Statistics Canada’s confidentiality policies.
According to the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans,
“REB review is not required for research that relies exclusively on secondary use of anonymous
information, or anonymous human biological materials, so long as the process of data linkage or
recording or dissemination of results does not generate identifiable information” Article 2.4.
Permission to access the CHMS dataset was obtained through formal application and signed
contract found in Appendix A.
35
3.6 Data Limitations
Individual survey design and methodology provided some limitations to our analysis. Although
each survey collected demographic, socioeconomic and oral health information from nationally
representative samples of their populations, the degree and availability of comparable outcome and
explanatory variables were limited. Outcome variables, such as access to dental care and treatment
needs were not collected by all surveys. As well, the denominator for measures of decayed, filled,
and missing teeth differed between surveys. Earlier surveys collected information on all adult
teeth, including third molar counts to provide a range of 00-32 tooth counts, whereas the 2007-
2009 CHMS excluded third molar counts (range 00-28). Fortunately, our analyses found no
significant difference between CIs at different tooth counts (Appendix C). In addition, the 2007-
2008 NHANES performed a basic screening examination on each participant, providing
information on the presence of one or more decayed, sealed, or restored tooth. Since this survey
did not collect information on the number of teeth decayed, sealed, or restored, we could not
measure the severity of dental disease.
Due to variation reporting educational attainment across surveys, our analyses were limited to adult
populations. The 1970-1972 NCNS only reported education based on participant interviewed
rather than head of household. As such, we did not include individuals who were not eligible, due
to age, to obtain high school education. Education was based on high school graduation rather than
categorical levels of education, such as secondary and post-secondary education; this was
primarily due to the method of data collection in each dataset and inability to transform variables
to incorporate higher levels of education into our analysis. In addition, explanatory variables such
as dental insurance, type of dental insurance, rural/remote location, occupational status, oral
hygiene behaviors, could not be used.
Our analyses are limited to binary outcome variables measuring the presence of oral health
outcomes. Future studies would benefit from examining not only the presence of oral health but
also the severity of oral health outcomes across a population, such as number of decayed, filled,
and missing teeth.
36
Results
4.1 Sample characteristics
Descriptive statistics for the analysis samples are presented in Tables 6 and 7. The majority of the
participants in all surveys are of young to middle adult age (20-59), with equal representation
among sexes, and across income quintiles. There was greater representation of participants with
higher educational attainment (high school graduate) in both Canada and the United States in the
2000s compared to earlier surveys. There was no statistically significant difference between
analysis and full survey samples.
Table 6. Analysis sample characteristics (%).
Canada
1970-1972
Canada
2007-2009
United States
1971-1974
United States
2007-2008
N=9,483 N=3,313 N=12,608 N=5,003
Age
20-39yr 57.7 39.3 45.8 40.8
40-59yr 34.9 42.8 36.9 42.1
60-74yr 7.4 17.9 17.3 17.1
Sex
Female 58.6 49.4 52.7 51.2
Male 41.4 50.6 47.3 48.8
Age/Sex
Male 20-39yr 14.9 19.7 21.7 20.3
Male 40-59yr 13.2 21.2 18.0 20.6
Male 60-74yr 5.9 8.4 7.6 7.9
Income
Lowest 28.9 23.8 15.8 13.9
Lower middle 18.7 18.9 10.1 28.2
Middle 27.6 16.8 28.0 14.9
Upper middle 15.5 12.5 24.0 19.7
Highest 4.4 28.0 22.1 20.2
Education
No high school 64.6 8.9 36.7 18.7
High school graduate 35.6 91.1 63.3 81.3 1Weighted proportions expressed as percentage. 2Based on full adult population analysis sample.
37
Table 7 outlines the descriptive statistics for three clinical oral health outcomes by survey and
income quintile. Oral disease outcomes, such as the presence of decayed teeth and edentulism,
have declined in both countries over time, with greater reductions exhibited in Canada than the
United States. There have been no considerable changes in the presence of one or more filled teeth
over the past 40 years in both countries.
Income gradients for oral health outcomes were well defined in all four surveys. For the presence
of one or more decayed teeth, income gradients were more pronounced in the United States, with
greater inequality between the lowest and highest income quintiles over time. The absolute
[relative] percentage difference between highest and lowest income quintiles for decayed teeth in
Canada were 9.30% [1.16] and 19.5% [2.21] in the 1970s and 2000s, respectively, and 17.4%
[1.50] and 32.3% [4.08] in the United States in the 1970s and 2000s.
Over time, income gradients for the presence of one or more filled teeth have reduced in both
countries. In the United States at both time periods, the absolute differences between the lowest
and highest income quintiles (30% and 21.5%) were greater compared to Canadian surveys (5.6%
and 10.7%). The relative differences in Canada were 0.93 and 0.92 from 1970-2009, and 0.68 and
0.77 in the United States from 1971-2008.
The difference between the lowest and highest income groups for the presence of edentulism has
narrowed over time. The absolute difference in Canada has increased for the presence of
edentulism from 7.6% to 9.7% (relative difference 1.42 and 7.46, respectively), and decreased in
the United States from 21.3% to 7.0% (relative difference 3.69 and 3.69, respectively).
38
Table 7. Oral health outcomes of sample population.
Canada
1970-1972
Canada
2007-2009
United States
1971-1974
United States
2007-2008
Presence of one or more decayed teeth 62.3 (59.2, 65.3) 21.6 (19.3, 23.7) 46.8 (45.4, 48.1) 21.3 (19.9, 22.8)
Income
Lowest 68.1 (61.3, 74.0) 35.5 (30.8, 40.4) 52.1 (48.6, 55.3) 42.8 (38.8, 46.9)
Lower middle 68.1 (60.6, 74.8) 28.7 (23.5, 34.6) 54.4 (50.3, 58.4) 26.8 (24.0, 29.7)
Middle 66.9 (60.1, 73.1) 20.5 (15.7, 26.3) 54.1 (51.7, 56.5) 17.0 (13.8, 20.7)
Upper middle 65.1 (54.3, 74.6) 23.1 (17.8, 29.5) 44.7 (42.0, 47.3) 15.3 (12.4, 18.7)
Highest 58.8 (41.8, 73.9) 16.0 (12.8, 19.8) 34.7 (31.9, 37.4) 10.5 (8.0, 13.6)
Presence of one or more filled teeth 72.9 (70.2, 75.3) 92.9 (90.2, 96.5) 82.1 (81.1, 83.0) 83.5 (82.1, 84.7)
Income
Lowest 69.1 (63.1, 74.6) 89.0 (84.5, 92.3) 63.9 (60.7, 66.9) 70.5 (66.6, 74.2)
Lower middle 71.4 (65.0, 77.1) 91.7 (87.8, 94.5) 70.2 (66.2, 73.9) 80.2 (77.6, 82.6)
Middle 72.9 (66.9, 78.1) 95.0 (91.1, 97.2) 79.0 (76.9, 80.9) 84.6 (80.9, 87.7)
Upper middle 66.0 (56.2, 74.6) 93.1 (86.1, 96.8) 88.3 (86.5, 89.9) 87.1 (83.8, 89.9)
Highest 74.7 (61.4, 84.5) 95.7 (93.0, 97.4) 93.9 (92.4, 95.0) 92.0 (89.3, 93.9)
Presence of edentulism 23.0 (21.2, 25.1) 5.6 (4.7, 6.3) 15.5 (14.7, 16.4) 4.9 (4.2; 5.6)
Income
Lowest 25.6 (21.9, 29.5) 11.2 (9.7, 14.6) 29.2 (26.9, 31.6) 9.6 (7.8, 1.6)
Lower middle 25.2 (20.4, 30.7) 6.2 (4.5, 8.6) 22.4 (19.6, 25.5) 6.7 (5.4, 8.3)
Middle 21.6 (17.9, 25.9) 4.4 (2.5, 7.7) 15.5 (13.9, 17.2) 3.8 (2.3, 6.0)
Upper middle 19.3 (14.7, 24.8) 3.6 (1.5, 8.1) 10.7 (9.3, 12.3) 2.5 (1.4, 4.5)
Highest 18.0 (11.1, 27.8) 1.5 (0.7, 2.9) 7.9 (6.4, 9.5) 2.6 (1.4, 4.6) 1Weighted proportions and 95% confidence intervals. 2Decayed and filled teeth outcomes based on dentate population.
39
4.2 Concentration Indices
Table 8 presents the CIs for three oral health outcomes in each of the four surveys. The columns
list the observed and expected CIs. The difference between observed and expected CIs quantifies
the avoidable inequality in each outcome and is provided in Table 9. Appendix E provides
significance-testing results for comparisons between CI outcomes.
Table 8. Observed and Expected Concentration Indices.1,2
Presence of >1 Decayed
Teeth
Presence of >1 Filled
Teeth
Prevalence of
Edentulism
Observed Expected Observed Expected Observed Expected
Canada
1970-1972 -0.141 -0.014 0.076 0.031 -0.160 0.002 (0.037) (0.031) (0.030) (0.023) (0.016) (0.005) [0.001] [0.644] [0.011] [0.168] [<0.001] [0.971]
2007-2009 -0.167 -0.003 0.051 -0.001 -0.082 0.003 (0.024) (0.015) (0.017) 0.011 (0.011) (0.007) [<0.001] [0.809] [0.003] [0.009] [<0.001] [0.725]
United States
1971 -1974 -0.163 -0.003 0.230 0.003 -0.161 -0.002 (0.015) (0.009) (0.011) (0.005) (0.010) (0.006) [<0.001] [0.740] [0.000] [0.594] [<0.001] [0.797]
2007-2008 -0.224 -0.016 0.115 0.008 -0.087 -0.004 (0.018) (0.011) (0.014) (0.009) (0.013) (0.010) [<0.000] [0.138] [<0.000] [0.348] [<0.001] [0.707] 1 CI reported, (SE), [p-value] 2 p-value<0.05 indicates significant difference from equality.
Table 9. Concentration Indices.
Presence of >1
Decayed Teeth
Presence of >1
Filled Teeth
Prevalence of
Edentulism
Canada 1970-1972 -0.156 0.045 -0.162
Canada 2007-2009 -0.164 0.053 -0.085
United States 1971 -1974 -0.160 0.227 -0.159
United States 2007-2008 -0.209 0.106 -0.083
40
In all four surveys, the presence of one or more decayed teeth is concentrated among the poor,
with greater inequalities reported in both countries over time. This pro-poor finding is similar for
edentulism; however, there has been considerable reduction in the magnitude of inequality in
edentulism in both countries over time. The presence of one or more filled teeth is concentrated
among the better off in all four surveys, with a decline in the magnitude of inequality in this
outcome over time in the United States and increases in inequality in Canada. When comparing
between countries, greater oral health inequalities were exhibited in both measures of one or more
decayed and filled teeth in the United States compared to Canada.
4.2.1 Redistribution of the Concentration Index
Table 10 outlines the approximate percentage redistribution from least deprived to most deprived
required to eliminate inequality for each outcome. Negative values indicate the outcome would
have to be redistributed from low-income groups to high-income groups and vice versa.
Table 10. Percentage redistribution requirements (%).
Presence of >1
Decayed Teeth
Presence of >1
Filled Teeth
Prevalence of
Edentulism
Canada 1970-1972 -11.7 3.4 -12.2
Canada 2007-2009 -12.3 3.9 -6.4
United States 1971 -1974 -12.0 17.0 -11.9
United States 2007-2008 -15.7 8.0 -6.2 Note: Percentages indicate how much of the outcome would need to be redistributed among the population in order
to achieve equality.
For the presence of edentulism, the percentage redistribution required to eliminate these
inequalities has decreased in both countries over time. Percentage redistribution for inequalities in
the presence of one or more filled teeth differ over time and between countries. Greater
redistribution requirements are exhibited in the United States over time and less for the Canada.
The greatest redistribution requirements are exhibited in more recent surveys for the presence of
one or more decayed teeth. The amount of redistribution from least to most deprived for the
presence of decayed teeth has increased in both countries over time.
41
4.3 Decomposition Analysis
Tables 11 to 14 show the decomposition results for each of the four surveys. The contribution of
each explanatory variable on the CI is a product of that variable’s elasticity and CI. Elasticities
show how a change in an explanatory variable impacts the oral health outcome. It is interpreted as
a unit-less measure where larger values indicate greater sensitivity of the outcome variable to
changes in the explanatory variable. The CI examines the distribution of an explanatory variable
along the income gradient. Its interpretation has been mentioned previously. Overall, the
contribution identifies the extent to which the explanatory variable contributes to the oral health
outcome’s CI.
4.3.1 Canada 1970-1972
Table 11 shows the decomposition results for all three oral health outcomes in Canada from 1970-
1972. The first column for each outcome identifies the effect of explanatory variables on reporting
oral health outcomes. As shown in Table 11, the largest elasticities for the presence of one or more
decayed teeth were observed in family size, middle-aged adults (40-59years), and education. For
the presence of one or more filled teeth, the largest elasticities were in sex and education. That is,
the presence of one or more filled teeth is most sensitive to variations in sex and education
compared to other explanatory variables. The negative elasticity indicated for males (-0.100)
implies that more men decrease the presence of one or more filled teeth, whereas the positive
elasticity noted in education (0.111) implies that as education increases, so does the presence of
one or more filled teeth. As such, the contribution of sex and education on income-related
inequalities in filled teeth may not only be explained by changes in their distribution across
income(CI), but also due to their large impact on the presence of one or more filled teeth. For the
presence of edentulism, age, sex, and education had the greatest elasticities, suggesting that
increases in age and the number of men, and decreases in educational attainment increases the
presence of edentulism in this population.
42
Table 11. Decomposition Results, Canada 1970-1972.
Presence of one or more decayed
tooth
Presence of one or more filled
tooth Presence of edentulism
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d 𝜂𝑘a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.076 0.005 0.000 -0.025 0.005 0.000 0.173 0.024 0.004
60-74yr -0.033 -0.045 0.001 0.001 -0.009 -0.045 0.000 0.000 0.358 -0.126 -0.045 -0.041
Sex Female e
Male 0.032 -0.007 0.000 0.000 -0.100 -0.007 0.001 0.000 0.921 -0.093 -0.086 -0.086
Age/Sex
Male 20-39yr e
Male 40-59yr 0.015 -0.009 -0.009 0.024 -0.009 0.000 -0.040 0.061 -0.002
Male 60-74yr 0.010 -0.104 -0.104 -0.001 -0.006 -0.104 0.001 0.001 -0.018 -0.109 0.002 -0.001
Income
Lowest e
Lower Middle -0.011 -0.443 0.005 0.000 -0.443 0.000 -0.005 -0.114 0.001
Middle -0.028 -0.105 0.003 -0.002 -0.105 0.000 -0.021 0.347 -0.007
Upper middle -0.018 0.216 -0.004 -0.012 0.216 -0.003 -0.014 0.763 -0.011
Highest -0.015 0.588 -0.009 -0.005 -0.004 0.588 -0.002 0.059 -0.005 0.956 -0.005 -0.022
Education < High school e
High school
Grad -0.073 0.031 -0.002 -0.002 0.111 0.031 0.003 0.003 -0.245 0.024 -0.006 -0.006
Family
Size Family Size 0.094 0.049 0.005 -0.003 0.049 0.049 0.002 0.032 -0.078 0.063 -0.005 -0.005 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
43
The second column for each outcome identifies the concentration indices, known as the degree of
income-related inequality, for each explanatory variable; this is also displayed in Figure 6 (below).
In the Canada 1970-1972 survey, older adults aged 60-74 years were concentrated among the poor
(C=-0.126), specifically older males ages 60-74 years (C=-0.109). In general, males were slightly
more concentrated among the poor (C=-0.093).
Figure 6. Concentration indices by explanatory variable, Canada 1970-1972
Figure 7 provides information on the contribution of each explanatory variable to inequality for
the Canada 1970-1972 survey. Values to the left of origin indicate pro-poor contribution to
inequalities, whereas values to the right of the origin indicate pro-rich contributions. From the
decomposition analysis, sex and age were large contributors to inequalities in edentulism.
Socioeconomic factors, such as income, household size, and education were more predominant
contributors to measures of filled and decayed teeth, compared to other explanatory variables.
There were no large contributors to inequalities in one or more decayed teeth from the explanatory
variables used in our analysis.
44
Figure 7. Aggregate contribution to income-related inequality, Canada 1970-1972
4.3.2 Canada 2007-2009
The largest elasticities for the presence of decayed teeth were observed in education, household
size, income, and sex (Table 12). For the presence of more than one filled teeth, there were no
large elasticities in any variables. For the presence of edentulism, the greatest elasticities were
observed in age, sex, education, and income.
In other words, an increase in the number of men would increase the presence of decayed
teeth(0.199) and decrease the presence of edentulism(-0.335). Increases in education decreases the
presence of decayed teeth(-0.381) and edentulism(-0.146), and slightly increases the presence of
filled teeth(0.012). Increases in income would result in a rise in the presence of filled teeth, and
decrease the presence of decayed teeth and edentulism. As well, as age increases, so does the
presence of edentulism.
-0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20
Decayed
Filled
Edentulism
Age Sex Age/Sex Income Education Household Size
45
Table 12. Decomposition Results Canada 2007-2009.
Presence of one or more decayed
tooth
Presence of one or more filled
tooth Presence of edentulism
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d 𝜂𝑘a CIb Contrc Agg d
Age
20-39yr
40-59yr 0.065 0.095 0.006 0.029 0.095 0.003 0.290 0.101 0.029
60-74yr 0.058 -0.202 -0.012 -0.006 0.010 -0.202 -0.002 0.001 0.542 -0.232 -0.126 -0.096
Sex Female e
Male 0.199 0.066 0.013 0.013 -0.008 0.066 -0.001 -0.001 -0.335 0.064 -0.021 -0.021
Age/Sex
Male 20-39yr
Male 40-59yr -0.016 0.145 -0.002 -0.013 0.145 -0.002 0.259 0.156 0.040
Male 60-74yr 0.002 -0.107 0.000 -0.002 -0.001 -0.107 0.000 -0.002 0.156 -0.171 -0.027 0.014
Income
Lowest e
Lower Middle -0.049 -0.368 0.018 0.003 -0.368 -0.001 -0.028 -0.335 0.009
Middle -0.091 -0.010 0.001 0.005 -0.010 0.000 -0.028 0.022 -0.001
Upper middle -0.061 0.287 -0.017 0.003 0.287 0.001 -0.021 0.315 -0.007
Highest -0.230 0.708 -0.163 -0.161 0.012 0.708 0.008 0.008 -0.120 0.720 -0.086 -0.084
Education < High school e
High school Grad -0.381 0.031 -0.012 -0.012 0.012 0.031 0.000 -0.000 -0.146 0.035 -0.005 -0.005
Household
Size Household Size 0.376 0.078 0.029 0.029 -0.020 0.078 -0.002 -0.002 0.005 0.084 0.000 0.000 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
46
In Canada 2007-2009 older adults aged 60-74 years were concentrated among the poor (-0.102)
(Figure 8). Males and high school graduates were represented among the better off (0.062 and
0.032). In contrast, compared to younger adults, older males tended to be concentrated among the
poor (-0.089).
Figure 8. Concentration indices by explanatory variable, Canada 2007-2009
Income-related inequalities in oral health outcomes were largely explained by income in Canada
in 2007-2009 (Figure 9). Notably, the contribution of income to inequalities was greater in
measures of disease, decay and edentulism, than measures of more than one filled teeth. Education
contributed to inequalities in the presence of one or more decayed teeth, but less significantly for
other measures. Age and sex appear to have a contributing role to inequalities in edentulism.
47
Figure 9. Aggregate contribution to income-related inequality, Canada 2007-2009
4.3.3 United States 1971-1974
In the United States from 1971-1974, elasticities in the presence of one or more decayed teeth were
highest for family size, income, and education (Table 13). The trends in elasticities in these
variables are similar to that of other surveys. There was no large impact of variation in age or sex
on the presence of one or more decayed teeth or one or more filled teeth. Education and family
size exhibited the greatest elasticities in the presence of one or more filled teeth. Elasticities in age
and education were highest for the presence of edentulism, with no large impact on this outcome
with other explanatory variables.
-0.25 -0.15 -0.05 0.05 0.15 0.25
Decayed
Filled
Edentulism
Age Sex Age/Sex Income Education Household Size
48
Table 13. Decomposition Results United States 1971-1974.
Presence of one or more decayed
tooth
Presence of one or more filled
tooth Presence of edentulism
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d 𝜂𝑘a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.090 0.105 -0.009 -0.001 0.105 0.000 0.450 0.111 0.050
60-74yr -0.048 -0.253 0.012 0.003 0.001 -0.253 0.000 0.000 0.494 -0.317 -0.157 -0.107
Sex Female e
Male 0.073 0.055 0.004 0.004 -0.023 0.055 -0.001 -0.001 -0.027 0.059 -0.002 -0.002
Age/Sex
Male 20-39yr e
Male 40-59yr -0.003 0.171 -0.001 0.003 0.171 0.001 0.001 0.172 0.000
Male 60-74yr -0.007 -0.157 0.001 0.001 -0.002 -0.157 0.000 0.001 0.000 -0.213 0.000 0.000
Income
Lowest e
Lower Middle -0.001 -0.642 0.001 0.005 -0.642 -0.003 -0.001 -0.583 0.000
Middle -0.012 -0.269 0.003 0.030 -0.269 -0.008 -0.036 -0.202 0.007
Upper middle -0.061 0.264 -0.016 0.041 0.264 0.011 -0.053 0.318 -0.017
Highest -0.104 0.759 -0.079 -0.092 0.050 0.759 0.038 0.038 -0.078 0.779 -0.061 -0.070
Education < High school e
High school
Grad -0.103 0.132 -0.014 -0.014 0.129 0.132 0.017 0.017 -0.267 0.163 -0.043 -0.043 Family
Size Family Size 0.198 0.055 0.011 0.011 -0.110 0.055 -0.006 -0.006 -0.036 0.072 -0.003 -0.003 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution; e Reference group.
49
As shown in Table 13 and Figure 10, older adults are concentrated among the poor (C=-0.317),
specifically males (C=-0.213), younger males are more likely to be concentrated among the better
off (C=0.172), high school graduates were among the better off (C=0.163). Males generally were
better off (C=0.059) in the United States in the 1970s.
Figure 10. Concentration indices by explanatory variable, United States 1971-1974.
Figure 11 displays the contribution of each explanatory variable to inequality in oral health
outcomes in the United States in 1971-1974. As noted, income and education strongly contributed
to inequalities in all oral health outcomes, with older age strongly contributing to inequalities in
the presence of edentulism.
50
Figure 11. Aggregate contributions to income-related inequality, United States 1971-1974
4.3.4 United States 2007-2008
Table 14 provides the results of the decomposition analysis for the United States from 2007-2009.
The impact of explanatory variables on oral health outcomes is synonymous with the other three
survey results. Elasticities in age and sex were low for the presence of one or more decayed teeth
and one or more filled teeth. Household size, income, and education exhibited greater elasticities
on the presence of one or more decayed teeth than the one or more filled teeth. For the presence of
edentulism, the highest elasticities were education, household size, age, and income.
-0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25
Decayed
Filled
Edentulism
Age Sex Age/Sex Income Education Household Size
51
Table 14. Decomposition Results United States 2007-2008.
Presence of one or more decayed
tooth
Presence of one or more filled
tooth Presence of edentulism
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d 𝜂𝑘a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.017 0.099 -0.002 0.055 0.099 0.005 0.111 0.098 0.011
60-74yr -0.045 -0.088 0.004 0.002 0.014 -0.088 -0.001 0.004 0.125 0.050 0.006 0.017
Sex Female e
Male 0.091 0.026 0.002 0.002 -0.043 0.026 -0.001 -0.001 -0.090 0.029 -0.003 -0.003
Age/Sex
Male 20-39yr e
Male 40-59yr 0.017 0.121 0.002 0.003 0.121 0.000 0.013 0.112 0.002
Male 60-74yr 0.002 -0.016 0.000 0.002 0.001 -0.016 0.000 0.000 0.010 0.111 0.001 0.003
Income
Lowest e
Lower Middle -0.077 -0.455 0.035 0.015 -0.455 -0.007 -0.030 -0.405 0.012
Middle -0.090 -0.027 0.002 0.011 -0.027 0.000 -0.075 0.053 -0.004
Upper middle -0.133 0.325 -0.043 0.018 0.325 0.006 -0.108 0.326 -0.035
Highest -0.167 0.733 -0.122 -0.128 0.025 0.733 0.019 0.017 -0.124 0.603 -0.075 -0.102
Education < High school e
High school
Grad -0.314 0.073 -0.023 -0.023 0.037 0.073 0.003 0.003 -0.211 0.089 -0.019 -0.019 Household
Size Household Size 0.258 0.022 0.006 0.006 -0.027 0.022 -0.001 -0.001 0.191 -0.092 -0.018 -0.018 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
52
Figure 12 shows the concentration indices of each explanatory variable. Older adults were
concentrated among the better off (C=0.050); however, when excluding edentulous individuals,
older adults appear to be concentrated among the poor (C=-0.088) (Table 14). Similar findings
were noted with older males in the full sample (dentate/edentate) concentrated among the poor
(C=-0.016). In both samples, males (C=0.029) and high school graduates (C=0.089) tended to be
among the better off.
Figure 12. Concentration indices by explanatory variable, United States 2007-2008
Figure 13 provides information on the contribution of explanatory variables to income-related
inequalities for the United States 2007-2008 survey. As shown, education and income were strong
contributors to income-related inequality in the measures of edentulism and presence of one or
more filled teeth. There was very little contribution of sex and age to inequalities in these measures.
53
Figure 13. Contribution to income-related inequality, United States 2007-2008
4.4 Summary Points
Overall, our analyses revealed pro-poor inequalities in the presence of one or more decayed teeth
and edentulism, and pro-rich inequalities in the presence of one or more filled teeth in all four
surveys.
In the 1970s, inequalities in oral health outcomes between Canada and the United States were
relatively equal in magnitude. The prevalence of one or more decayed teeth and edentulism was
higher in Canada compared to the United States, with the prevalence of reporting of one or more
filled teeth higher in the United States in the 1970s (Table 7). Income gradients for all oral health
measures were greater in the United States compared to Canada (Table 7); concentration indices
revealed greater income-related inequalities in the presence of one or more decayed and filled teeth
-0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20
Decayed
Filled
Edentulism
Age Sex Age/Sex Income Education Household Size
54
in the United States, whereas income-related inequalities in the presence of edentulism was greater
in Canada.
Comparing Canada from 1970 to 2009, income gradients for all oral health outcomes are steeper
in the more recent Canadian survey than previously reported (Table 7). Moreover, there have been
increases in income-related inequalities in the presence of one or more decayed and filled teeth
with higher percentage redistribution requirements in 2007-2009 (12.3% and 3.9%) compared to
1970-1972 (11.7% and 3.4%) (Table 10). The prevalence of edentulism along with income-related
inequalities in this measure have declined in Canada over time (Table 7).
From the 1970s to 2000s, the prevalence of one or more decayed teeth and edentulism have
declined in the United States, with the prevalence of one or more filled teeth remaining the same
over time (Table 6). Income gradients for all oral health outcomes have persisted, with lower
income-related inequalities reported in the presence of one or more filled teeth and edentulism
since the 1970s (Tables 7 and 8). Income-related inequalities in the presence of one or more
decayed teeth increased in the United States with redistribution requirements increasing from
12.0% to 15.7% over time (Table 10).
In more recent years, the prevalence of all three oral health outcomes is similar in Canada and the
United States (Table 7); however, income-related inequalities in measures of one or more decayed
or filled teeth appear to be greater in size in the United States. Income-related inequalities in the
presence of edentulism are of relatively equal size in both countries. The greatest difference in
percentage redistribution requirements was exhibited in the presence of one or more filled teeth in
the United States (8.0%) compared to Canada (3.9%).
In terms of the decomposition analysis, age contributed most to inequalities in edentulism
compared to measures of decayed and filled teeth. The contribution of age on inequalities in
edentulism decreased in both countries over time, which might be explained by the overall
reduction in inequalities in this outcome (Table 7).
The contribution of sex to oral health inequalities was greatest for edentulism in Canada at both
time periods, with no large effect on inequalities in the United States or in other oral health
outcomes. Our findings indicate being male had a strong positive association with reporting
55
edentulism in Canada in the 1970s, but had a negative association in Canada in the 2000s. For
measures of decayed and filled teeth, there was no large effect of age or sex on inequalities.
Income contributed to inequalities in all oral health outcomes. Notably, in Canada in the 1970s,
income did not strongly contribute to inequalities The presence of decayed teeth and edentulism
were less likely to be reported as income increased, with opposite trends for the presence of filled
teeth. Similar trends were found with education, as attainment of high school education was
negatively associated with reporting tooth decay and edentulism in all surveys.
56
Discussion
To our knowledge, this is the first study to estimate and compare the magnitude of income-related
oral health inequalities in Canada and the United States from 1970-2009. The results conform to
the proposed the objectives:
i. To compare the magnitude of oral health inequalities in Canada in the 1970s and 2000s.
ii. To compare the magnitude of oral health inequalities in the United States in the 1970s
and 2000s.
iii. To compare the magnitude of oral health inequalities between Canada and the United
States over time.
iv. To determine the contributors to income-related oral health inequalities in Canada and the
United States.
v. To determine changes in the contributors to income-related oral health inequalities in
Canada and the United States.
5.1 Key findings
This study identified the persistence and magnitude of oral health inequalities in Canada and the
United States over the past 40 years. It revealed the greatest inequalities are exhibited in the
presence of untreated disease, characterized as one or more decayed teeth, and demonstrates
inequalities in oral disease are greater in the United States than in Canada over time.
We find income-related inequalities have decreased over time for measures of filled teeth and
edentulism, but have increased for measures of decayed teeth in both countries. We were further
able to identify contributors to inequalities and changes to these contributors over time. Generally
though, since the 1970s, income and education have contributed the greatest to inequalities in oral
health in both countries. Our results also provide information on contributions unique to different
oral health outcomes, in different countries, and at different times.
57
5.1.1 Oral Health Inequalities in Canada and the United States
Decreases in income-related inequalities in measures of one or more filled teeth were identified
over time. The overall presence of one or more filled teeth in both countries has plateaued between
the 1970s and 2000s. Pro-rich income gradients for reporting the presence of one or more filled
teeth in both countries at both time periods were less than for measures of decayed teeth and
edentulism. These findings suggest that access to dental care to receive restorative treatment
continues to be slightly concentrated among the better off. Measures of one or more filled teeth
serve as an indicator of previous treatment for dental disease. Given the cross-sectional nature of
these surveys, it is difficult to determine the time in which dental treatment for teeth was provided.
As well, filled tooth counts do not often distinguish between types of restorative care, such as
preventive or cosmetic restorations. However, declines in inequalities in one or more filled teeth
provide valuable insight on the potential societal influences on oral health over the life course.
Measures of filled teeth provide information on previous utilization of dental care as well as history
of disease. Overall, our findings identify lowest income-related inequality in this measure
suggesting the presence of one or more filled teeth is concentrated among the better off, but is the
most equal distribution in comparison to other measures.
Our results indicate decreases in income-related inequalities in edentulism over time in both
countries. This trend may due to the overall decline in the prevalence of edentulism in both over
the past 40 years. Edentulism is a measure of the cumulative effects of oral disease in an
individual’s lifetime. This measure identifies the overall burden of oral diseases and serves as a
true indicator of oral disease. The main reasons for the decline in the prevalence and income-
related inequality in edentulism may be due increases in tooth retention over the past three decades;
this trend has been attributed to improved conservative dental philosophies, such as increased
focus on prevention as well as positive health-seeking behaviours and attitudes exhibited by the
general population (U.S. DHHS, 2000).
For measures of one or more decayed teeth, our results reveal that despite the decline in the level
of untreated decay in both countries there have been increases in income-related inequalities over
time. From our analyses, of all the untreated dental diseases within a population, it is
disproportionately represented among the poor. These findings are consistent with existing
international literature on inequalities in oral health outcomes. Inequalities are often greater in
58
disease than in treatment. Australia reported greater social gradients in missing and untreated
decayed outcomes with less inequality in filled teeth in an adult Australian population (Mejia,
Jamieson, Ha, & Spencer, 2014). In addition, Meija and colleagues (2014) found as the prevalence
of decayed teeth declines in a population, groups of higher socioeconomic status often experience
the sharpest decline compared to other groups. Our findings corroborate with this claim that
although dental decay rates have declined over time, inequalities across the income gradient show
that the poor have a disproportionately higher share of dental decay (Tables 7 and 8).
The social, political, and economic environments, as well as the oral health care systems may
explain the differences in income-related inequalities across countries. In the 1970s, the high pro-
poor inequalities in the presence of decayed teeth and edentulism could be attributed to the lack of
dental coverage for low and middle-income populations in Canada. In the 1970s, only 3.4% of the
Canadian population were eligible for some sort of dental program (Stamm, Health, Canada, &
Directorate, 1986). Prepayment dental plans were introduced around the 1960s, enabling a
reduction in cost at the point of care for those eligible, thus making those insured more likely to
afford care. Private dental plans were often linked to employment-based benefits, suggesting those
of middle to higher income would better be able to pay for care. This may have contributed to pro-
rich inequalities in the presence of one or more filled teeth. In addition, public dental plans
covering recipients on social assistance favored extraction services over conservative dental care,
which may have led to high pro-poor inequalities in edentulism (Stamm et al., 1986).
Further, inter-regional variations in expenditures of dental care may explain inequalities in oral
health outcomes in Canada in the 1970s. Leake (1984) described potential inequity in the share of
dental care expenditures across Canada, identifying a lower percent of overall dental expenditures
incurred in maritime provinces compared to the Prairies, Quebec, Ontario, and British Columbia
(Leake, 1984). This along with variations in unemployment rates across Canada may also
indirectly contribute to inequalities in oral health due to an individual’s ability to pay dental care
when living with low or no income (Leake, 1984; Statistics Canada, 2012). However, due to data
limitations preventing analysis at the provincial/territorial level, these inferences cannot be
confirmed.
Oral health inequalities in the United States in the 1970s may be explained by similar
characteristics to the social and economic environments as Canada as well as the oral health care
59
system at the time. The greater magnitude of inequalities in the United States compared to Canada
in the 1970s, may be explained by the shortage of dentists and variation and scarcity of dental
coverage across the country (Waldman, 1980). The likelihood of utilizing dental care was shown
to increase with increasing income, which corroborates our finding of pro-rich inequality in the
presence of one or more filled teeth (Douglass & Cole, 1979).
Compared to the 1970s, income-related inequalities in the United States have decreased for the
presence of one or more filled teeth and edentulism (Table 8). This may be explained by decreases
in the prevalence of edentulism since the 1970s in the United States (Table 7). As stated previously,
the timing of restorative care for the presence of one or more filled teeth cannot be determined
without longitudinal analysis or additional measures; therefore, the reduction in pro-rich
inequalities in this measure is difficult to ascertain. The increases in income-related inequalities in
decayed teeth may be attributed to rise in unemployment and shifts to part-time employment
beginning in the early 2000s; this would likely create to economic barriers to affording dental
treatment for lower income populations (Borbely, 2009). In addition, the persistence of income-
related inequalities in decayed teeth and edentulism may be explained by the lack of dental
insurance coverage and quality of dental insurance for low-income adult populations in the United
States (McGinn-Shapiro, 2008). As well, trends in dental care utilization for adult populations in
the United States began to decline in the early 2000s with stable rates of utilization for higher
income individuals (Wall, Vujicic, & Nasseh, 2012); this may explain increases in pro-poor
inequalities in the measures of untreated dental disease (decay) and pro-rich inequalities in
measures of filled teeth. The ability to finance dental care through insurance coverage may also
have influenced inequalities in oral health care outcomes. It has been suggested that health benefits,
including dental care, offered through employment dropped 10.6% between 2000 and 2010
(Gould, 2012).
Similar outcomes of income-related inequalities were exhibited in Canada in the 2000s as in the
United States (Table 7). Smaller income-related inequalities in the presence of one or more
decayed teeth in Canada compared to the United States may be explained by better access to
preventive dental care through use of dental hygienists in recent years as well as better coverage
of dental insurance for low-income adults and social assistance recipients (FPTDWG, 2013).
Similar pro-poor inequalities in the presence of edentulism may be explained by availability and
quality of dental insurance for lower income populations, such that extraction services may be
60
favored over more conservative treatments. In addition, pro-rich inequalities in the presence of one
or more filled teeth could be explained by increased costs of dental services over time and the
potential inability of lower income populations to afford dental treatment.
5.1.2 Age and Oral Health
Age contributions to inequalities in treated and untreated dental diseases were identified in our
decomposition analysis. Specifically, older adults in both countries tended to contribute more to
inequalities in edentulism, with a lesser effect of age on inequalities in recent surveys. In addition,
older adults tended to be more representative of the worse off. These findings coincide with
existing research identifying greater rates of missing teeth with age and steeper income gradients
in these populations (Mejia et al., 2014; Sanders & Spencer, 2004).
The contribution of age to inequalities in edentulism may be explained by the inverse relationship
between retention of teeth and increasing age (Shen et al., 2013). As well, Manski identified that
tooth retention may be inversely related to diminishing income (Manski et al., 2010; Manski et al.,
2009). This supports our findings that the older populations had a higher likelihood of reporting
dental disease – dental decay and edentulism. This finding is important as the number of dental
care needs increase with age, the ability to finance this care may also decrease.
Several studies have identified decreases in utilization of dental services of older adult populations
after the age of 65. Declines in utilization and oral health may be explained by the loss of
employment-based dental insurance as well as the reduction in income after retirement (Manski et
al., 2010; Manski et al., 2009). The ability to afford dental care treatment after retirement has been
cited as a cost-barrier for older adult populations (Kiyak & Reichmuth, 2005). The probability of
dental insurance coverage tends to drop off for individuals over the age of 65 (Bhatti, Rana, &
Grootendorst, 2007). A Canadian report identified that the highest level of dental disease and
highest rate of no insurance were among adults 60-79 years old (Health Canada, 2010). In the
United States and Canada dental insurance coverage for older adults is estimated to range from
14.5 to 46.8 per cent (Kiyak & Reichmuth, 2005). The effect of income and dental insurance
coverage on dental care utilization cannot be determined from our analysis. In general older adults
were concentrated among poor groups, which may imply that this population group may
experience potential cost-associated barriers to accessing oral health care and challenges in
maintaining optimal oral health.
61
5.1.3 Sex and Oral Health
The contribution of sex to inequalities in edenutlism and the number of decayed teeth were
exhibited in both Canadian surveys. Over time, it appears that sex differences in clinical oral health
outcomes are diminishing. This finding coincides with existing work by Ravaghi and colleagues
(2013b) who reported sex differences in clinical oral health outcomes in the Canadian population.
Interestingly, our analyses revealed that men were more likely to report the presence of one or
more decayed teeth, whereas Ravaghi et al. (2013b) found that women had a greater mean number
of decayed teeth and greater presence of edentulism. This contrast raises interesting questions as
to the difference between the prevalence and severity measures of oral disease. It has been
discussed previously that different measures of oral health may reveal different outcomes; however
comparisons between the prevalence and mean level of dental disease have not frequently been
compared across income groups or by sex.
Sex differences are rarely explored or identified in the dental literature (Wamala, Merlo, &
Boström, 2006). Due to the scarcity of reporting and identifying sex differences for oral health
outcomes, the mechanism for which these differences occur is unknown. Our findings may be a
result of access and lifestyle differences between sexes. For example, Ravaghi et al (2013b) stated
that lower income women in Canada might have more limited access to oral health care than their
male counterparts, leading to worse oral health outcomes. However, Tapp (2009) identified that
men were more likely to be self-employed, and have higher unemployment rates than female
counterparts, which may suggest affordability as an issue to accessing care for this population
group. In summary, further investigation into the potential contributors of reporting oral health
outcomes and inequalities between sexes should be conducted.
5.1.4 Income and Oral Health
Income was a strong contributor to oral health inequalities in all four surveys. The positive effect
of increasing income on oral health outcomes, such as one or more filled teeth is consistent with
existing literature. Often higher income individuals report less need for dental treatment, as well
as better access to dental care. Conversely, the lower the income level, the greater the likelihood
of reporting decayed teeth and edentulism in all four surveys. As such, lower income groups are
less likely to receive care, although they have higher treatment needs (Ramraj et al., 2012). This
concept is referred to as the inverse care law (Hart, 1971).
62
Our findings coincide with existing literature on income gradients and the effect of income on oral
health outcomes. Cost is often cited as a predominant barrier to accessing dental care (Manski et
al., 2012; Mejia et al., 2014; Thompson et al., 2014). Those of lower or middle income are more
likely to express treatment needs or difficulty accessing care (Ramraj et al., 2013). This may be
explained by the increased likelihood of reporting dental insurance with increased income (Health
Canada, 2010). The rise in non-standard, temporary, part-time employment diminishes the
availability of employment-based dental insurance for many low- and middle-income Canadians
and Americans, and may hinder the availability of dental insurance coverage for these populations.
Our findings support this statement as untreated dental disease appear to be concentrated among
the worse-off who may be unable to afford dental care in an environment of insurance scarcity.
Recent work by Bernabé and Marcene (2011) suggest the degree of state income inequality may
be attributed to inequalities in tooth loss in the United States, which may imply if income were
more equally distributed across a population, inequalities in oral health would be reduced. It also
implies that inequalities in oral health may be explained by factors beyond individual-level factors
(Bernabé & Marcenes, 2011).
International literature also suggests the degree of national income inequality may affect utilization
of dental care services. Bhandari et al. (2015) found that for adults in 66 countries, every 10%
increase in Gini coefficient, a measure of income inequality, was associated with a 15% lower
odds of using dental services (Bhandari et al., 2015). They also found total health expenditure,
public expenditure on health, health system responsiveness, or type of dental health system, as
well as income inequality, explained the association between income inequality and use of dental
services. They reported more equal countries have greater use of dental services (Bhandari et al.,
2015).
Nation-level inequalities in Canada and the United States have reportedly risen since the 1980s.
The Organisation for Economic Co-operation and Development [OECD] reported income
inequality in the United States is considerably higher than other developed countries, and has
increased substantially since the 1980s (OECD, 2014). Canada ranked 21st in terms of income
inequality compared to other OECD countries in the late 2000s. For both before- and after-tax
measures of family income, inequality - as expressed by the Gini coefficient - has increased in
Canada since 1976 (Rajotte, 2013). These findings may explain the increased contribution of
63
income to oral health inequalities over the past 40 years; however, the mechanisms as to how
income inequality relate to health inequalities are not fully understood (Pickett & Wilkinson,
2015).
5.1.5 Education and Oral Health
From our analyses, educational attainment contributed to oral health inequalities. High school
graduation contributed greatly to inequalities in the presence of one or more filled teeth, and had
a negative effect on reporting one or more decayed teeth and the presence of edentulism. These
findings match with existing literature indicating greater caries experience in individuals with
lower educational backgrounds (Schwendicke et al., 2015). In addition, the contribution of
education to oral health outcomes has been reported in different countries, regardless of the type
of social policy around dental care (Guarnizo-Herreno, Watt, et al., 2013).
It may be argued the knowledge and skills gained through education affect cognitive function,
receptiveness to health education message, or better oral health literacy. People from low
socioeconomic background, including education as a determinant, are reportedly more likely to
engage in unhealthy behaviours (Galobardes, Lynch, & Smith, 2007). Further, these lifestyle
choices may include consumption of sugars and foods high in refined carbohydrates, which
considerably influence an individuals’ susceptibility to dental decay (Sisson, 2007). Therefore, it
is logical to assume those with high school attainment or more would be less likely to report oral
disease outcomes and more likely to report oral health or treated disease.
5.2 Recommendations
In 2012, the International Association of Dental Research Global Oral Health Inequalities [IADR-
GOHIRA] Steering and Task Group outlined a research agenda to generate evidence to aid in
reducing oral health inequalities (Sgan-Cohen et al., 2013). Of the four research aims proposed by
the IADR-GOHIRA, our study meets the first two of “better understanding the full range of oral
health determinants that include biological and environmental factors as well as behaviours and
social determinants of health and well-being” and “research on social and physical environments,
across the social gradient…” (Sgan-Cohen et al., 2013). Our findings provide insight on the
potential effects that social and political environments as well as oral health care systems may have
64
on oral health inequalities. In addition, in this context, we have been able to identify contributing
factors to income-related oral health inequalities.
Our results emphasize the degree to which income and socioeconomic status contribute to income-
related oral health inequalities. The increase in contribution of income to these inequalities in more
recent years identifies the need to further investigate its role in accessing and obtaining optimal
oral health. As dental care in Canada and the United States is predominately financed through out-
of-pocket or private insurance payments, with cost being reported as a common barrier to accessing
dental care, it is critical to explore mechanisms to improve affordability of care, which may
ultimately lead to better access to care for low- and middle-income populations; these mechanisms
may include increasing dental insurance coverage, increasing salaries and wages, and increasing
availability to a broader range of providers.
Low- and middle-income individuals are often ineligible for employment-based or public dental
insurance due to their type of employment or level of income (Ramraj et al., 2013). Income
eligibility for public dental programs is often low or tied to social assistance programs (Quiñonez
et al., 2007). With increases in non-standard employment, such as temporary and part-time work,
individuals are often not eligible for private or employment-based insurance. Therefore,
consideration for changes in enrollment criteria and eligibility standards for public programs, as
well as costs of private insurance plans should be sought.
In addition to dental insurance coverage across populations, out-of-pocket payments for services
not covered under insurance plans may pose an additional financial burden. Therefore, there is
need to mitigate the additional costs of some dental care, which may be performed either through
improving the quality of dental insurance, or increasing wages and salaries for low- and middle-
income populations. Improvements in oral health have been reported with expanded coverage of
dental insurance (Bailit et al., 1985). Increasing wages and salaries for low- and middle income
populations would work to provide individuals and families with enough earnings to afford basic
dental care without hindering daily costs of living. The costs of dental care may impinge on the
affordability of meeting basic lifestyle needs (Snow & McNally, 2009). The Commission on Social
Determinants of Health reinforces the need for employment policies to incorporate living wages
that consider costs of sustaining healthy living rather than meeting basic living requirements
(Marmot et al., 2008).
65
Alternate approaches to improving oral health aside from reducing the cost, may be improving
availability of access to preventive dental care providers. Dental care in Canada and the United
States predominates through traditional private practice delivery. Alternate care settings, such as
access to community-based dental visits, mobile dental clinics, or alternative dental providers,
such as dental hygienists and dental therapists, may prove promising to reducing the costs to care,
increasing access to preventive care, and potentially improving oral health (Shaefer & Miller,
2011).
Of important note is that merely reducing cost-barriers to accessing dental care may not eliminate
income-related inequalities in oral health. Oral health outcomes are a result of the complex
interaction between societal and individual level factors, not only inclusive to accessing oral health
care. Therefore, a reduction or elimination of oral health inequalities within a population requires
an approach that addresses the social determinants of health within individuals and society itself.
At present, effective approaches to reducing income-related inequalities in health and oral health
are not well understood. This may be due to potential costs required to perform and evaluate these
interventions, as well as the lack of consistent measurement and surveillance systems for oral
health outcomes to evaluate changes over time. In the United States, the NHANES has provided a
means to monitor the oral health status of Americans over the past 60 years. However, consistency
in reporting outcomes, as well as ability to provide state-level comparisons is limited. In Canada,
there are only two nationally representative surveys on clinical oral health outcomes, which were
reported nearly 40 years apart.
It is imperative that federal and provincial/state governments commit to performing ongoing oral
health surveillance on a regular basis. The United States has made progress on ensuring periodicity
of oral health surveillance through the Affordable Care Act, which indicates,
“NATIONAL HEALTH AND NUTRITION EXAMINATION SURVEY.- The Secretary shall develop
oral healthcare components that shall include tooth-level surveillance for inclusion in the National Health
and Nutrition Examination Survey. Such components shall be updated by the Secretary at least every 6
years. For purposes of this paragraph, the term ‘‘tooth-level surveillance’’ means a clinical examination
where an examiner looks at each dental surface, on each tooth in the mouth and as expanded by the Division
of Oral Health of the Centers for Disease Control and Prevention.” (Affordable Care Act, 2010)
Efforts to ensure consistent and periodic surveillance in Canada have been requested from oral
health associations and working groups (Federal, Provincial, Territorial Dental Working Group,
66
2013); however, the commitment to these recommendations is not clear. In addition, the degree of
heterogeneity in collecting oral health data within and between countries and jurisdictions
impinges on the ability determine what improves oral health, much less oral health inequalities. It
is therefore critical for national and jurisdictional oral health representatives to agree upon
standardized measures for oral health.
5.3 Limitations
Given the heterogeneity between the four surveys used in our analyses, it is important to note the
shortcomings of our results. The availability of consistent data between surveys prohibited the
number and type of comparisons of outcomes. This was attributed to differences in data collection
methodologies such as questionnaire design and method of clinical examination. The exclusion of
certain population groups, as identified in Appendix B, indicates that our findings may not be
generalizable to the entire Canadian or American population.
Further, as there have only been two nationally representative surveys in Canada that contain
clinical oral health outcomes that were administered approximately 40 years apart from each other,
it is difficult to understand and compare changes to inequalities in oral health in each decade. Due
to the cross-sectional nature of surveys used in this study, our results cannot confirm causal
association between any of the explanatory variables on income-related oral health inequalities.
Instead, it provides a means to further investigate these potential associations through longitudinal
analyses.
5.4 Concluding Remarks
Our findings indicate inequalities in oral health outcomes have persisted and, to some extent, have
increased over the past 40 years in Canada and the United States; they also identify how
socioeconomic status, sex, and the life course may explain inequalities in oral health. Age and sex
differences offer insights as to how oral health outcomes may operate throughout the life course
and between sexes. They provide a means for generating hypotheses regarding sex differences and
how structural factors like social and economic conditions may influence oral health outcomes
67
between sexes and for the entire population. As such, our findings suggest that changes in the
social and economic environments within Canada and the United States have influenced oral health
inequalities over time. For example, the sustained rise of income inequality experienced in both
countries may explain the increase in contribution of income to oral health inequalities since the
1970s. This phenomenon might also be explained by shifts in labor markets, as described by the
loss of employment-based dental insurance with changes to non-standard employment; this would
ultimately lead to greater cost barriers to care for individuals of low- to middle income that may
explain the rise in income-related inequalities in decayed teeth outcomes. As well, the attenuation
of education’s contribution to inequalities over time may be a result of how societies as a whole
have become more cognizant of healthy oral hygiene behaviours since the 1970s. Likewise, the
decline in prevalence of dental disease could also be explained by the uptake in conservative
approaches to dental disease management by providers and increased knowledge of healthy
behaviours by individuals. In addition, the differences in inequalities between Canada and the
United States may be ascribed to structural- and individual-level characteristics that cannot be
reduced to measurement or are not presently quantifiable.
In conclusion, our findings provide a benchmark for comparison of oral health inequalities in
Canada and the United States. Ongoing monitoring of oral health outcomes within populations
will enable researchers and policy-makers to evaluate changes and uncover contributors to
inequalities in their populations.
68
References
Affordable Care Act, 42 U.S.C. § 18001 (2010).
Allin, S. (2008). Does equity in healthcare use vary across Canadian provinces? Healthcare
Policy, 3(4), 83.
American Dental Hygienists' Association. (2014). Direct Access States. Retrieved from
http://www.adha.org/resources-docs/7513_Direct_Access_to_Care_from_DH.pdf
Andersen, R. M. (1995). Revisiting the behavioral model and access to medical care: does it
matter? J Health Soc Behav, 36(1), 1-10.
Andersen, R. M. (2008). National health surveys and the behavioral model of health services use.
Medical care, 46(7), 647-653.
Bailit, H., Newhouse, J., Brook, R., Duan, N., Goldberg, G., Hanley, J., . . . Lohr, K. (1985).
Does more generous dental insurance coverage improve oral health? J Am Dent Assoc,
110(5), 701-707.
Baker, S. R. (2009). Applying Andersen's behavioural model to oral health: what are the
contextual factors shaping perceived oral health outcomes? Community Dent Oral
Epidemiol, 37(6), 485-494. doi: 10.1111/j.1600-0528.2009.00495.x
Baldota, K. K., & Leake, J. L. (2004). A macroeconomic review of dentistry in Canada in the
1990s. J Can Dent Assoc, 70(9), 604-609.
Bernabe, E., Delgado-Angulo, E. K., Murasko, J. E., & Marcenes, W. (2012). Family income
and tooth decay in US children: does the association change with age? Caries Res, 46(3),
221-227. doi: 10.1159/000337389
Bernabe, E., & Marcenes, W. (2011). Income inequality and tooth loss in the United States. J
Dent Res, 90(6), 724-729. doi: 10.1177/0022034511400081
Bernabe, E., & Sheiham, A. (2014). Extent of differences in dental caries in permanent teeth
between childhood and adulthood in 26 countries. Int Dent J, 64(5), 241-245. doi:
10.1111/idj.12113
Bhandari, B., Newton, J. T., & Bernabe, E. (2015). Income Inequality and Use of Dental
Services in 66 Countries. J Dent Res. doi: 10.1177/0022034515586960
Bhatti, T., Rana, Z., & Grootendorst, P. (2007). Dental insurance, income and the use of dental
care in Canada. J Can Dent Assoc, 73(1), 57.
69
Bloom, B., Simile, C., Adams, P., & Cohen, R. (2012). Oral health status and access to oral
health care for US adults aged 18-64: National Health Interview Survey, 2008. Vital and
health statistics. Series 10, Data from the National Health Survey(253), 1-22.
Borbely, J. M. (2009). US labor market in 2008: economy in recession. Monthly Lab. Rev., 132,
3.
Burau, V., & Blank, R. H. (2006). Comparing health policy: an assessment of typologies of
health systems. Journal of Comparative Policy Analysis, 8(01), 63-76.
Burt, B. A., & Eklund, S. A. (2005). Dentistry, dental practice, and the community: Elsevier
Health Sciences.
Canadian Dental Association. (2010). Dental health services in Canada: Facts and figures 2010.
Retrieved September, 17, 2011.
Canadian Dental Therapists Association. (2012). A brief history of dental therapy in Canada.
Retrieved from http://dental-therapists.com/his.htm
Centers for Medicare and Medicaid Service. National health expen- diture tables, 1960-2011.
http://www.cms.gov/Research-Statistics-Da- ta-and-Systems/Statistics-Trends-and-
Reports/NationalHealthExpend- Data/Downloads/tables.pdf. Accessed May 1th, 2013.
Cosgrove, J. C. (2008). Medicaid: Extent of Dental Disease in Children Has Not Decreased, and
Millions Are Estimated to Have Untreated Tooth Decay. Report to Congressional
Requesters. GAO-08-1121. US Government Accountability Office.
Cunha-Cruz, J., Hujoel, P. P., & Nadanovsky, P. (2007). Secular Trends in Socio-economic
Disparities in Edentulism: USA, 1972-2001. Journal of Dental Research, 86(2), 131-136.
doi: 10.1177/154405910708600205
Department of Heatlh and Human Services. (2000). Oral Health in America: A Report of the
Surgeon General (pp. 332). Rockville, MD: U.S. Department of Health and Human
Services, National Institute of Dental and Craniofacial Research, National Institutes of
Health.
Douglass, C. W., & Cole, K. O. (1979). Utilization of dental services in the United States.
Journal of Dental Education, 43(4), 223-238.
Dye, B. A., Tan, S., Smith, V., Lewis, B., Barker, L., Thornton-Evans, G., . . . Li, C. (2007).
Trends in oral health status: United States, 1988-1994 and 1999-2004. Vital and health
statistics. Series 11, Data from the national health survey(248), 1-92.
70
Edelstein, B. (2010). The dental safety net, its workforce, and policy recommendations for its
enhancement. Journal of Public Health Dentistry, 70, S32-S39. doi: 10.1111/j.1752-
7325.2010.00176.x
Elani, H. W., Harper, S., Allison, P. J., Bedos, C., & Kaufman, J. S. (2012). Socio-economic
inequalities and oral health in Canada and the United States. J Dent Res, 91(9), 865-870.
doi: 10.1177/0022034512455062
Federal, Provincial, and Territorial Dental Working Group [FPTDWG]. (2013). Reducing dental
disease: A Canadian oral health framework. Retrieved from:
http://www.caphd.ca/sites/default/files/FrameworkOctober%202014%20-
%20FINAL%20English.pdf
Galobardes, B., Lynch, J., & Smith, G. D. (2007). Measuring socioeconomic position in health
research. British Medical Bulletin, 81(1), 21-37.
Gould E. A decade of declines in employer-sponsored health insurance coverage. Economic
Policy Institute Briefing Paper #337, 2012. Retrieved from: www.epi.org/publication/
bp337-employer-sponsored-health-insurance/
Grignon, M., Hurley, J., Wang, L., & Allin, S. (2010). Inequity in a market-based health system:
Evidence from Canada's dental sector. Health Policy, 98(1), 81-90. doi:
10.1016/j.healthpol.2010.05.018
Guarnizo-Herreno, C. C., Tsakos, G., Sheiham, A., & Watt, R. G. (2013). Oral health and
welfare state regimes: a cross-national analysis of European countries. Eur J Oral Sci,
121(3 Pt 1), 169-175. doi: 10.1111/eos.12049
Guarnizo-Herreno, C. C., Watt, R. G., Pikhart, H., Sheiham, A., & Tsakos, G. (2013).
Socioeconomic inequalities in oral health in different European welfare state regimes. J
Epidemiol Community Health, 67(9), 728-735. doi: 10.1136/jech-2013-202714
Guarnizo-Herreno, C. C., Watt, R. G., Pikhart, H., Sheiham, A., & Tsakos, G. (2014).
Inequalities in oral impacts and welfare regimes: analysis of 21 European countries.
Community Dent Oral Epidemiol, 42(6), 517-525. doi: 10.1111/cdoe.12119
Hart, J. T. (1971). The inverse care law. Lancet, 1(7696), 405-412.
Health Canada Report Health Canada (2010) Report on the Findings of the Oral Health
Component of the Canadian Health Measures Survey 2007–2009. Ottawa: Health Canada
71
Jagger, D. C., Sherriff, A., & Macpherson, L. M. (2013). Measuring socio-economic inequalities
in edentate Scottish adults--cross-sectional analyses using Scottish Health Surveys 1995-
2008/09. Community Dent Oral Epidemiol, 41(6), 499-508. doi: 10.1111/cdoe.12040
Kiyak, H. A., & Reichmuth, M. (2005). Barriers to and enablers of older adults' use of dental
services. J Dent Educ, 69(9), 975-986.
Konings, P., Harper, S., Lynch, J., Hosseinpoor, A. R., Berkvens, D., Lorant, V., . . .
Speybroeck, N. (2010). Analysis of socioeconomic health inequalities using the
concentration index. Int J Public Health, 55(1), 71-74. doi: 10.1007/s00038-009-0078-y
Koolman, X., & van Doorslaer, E. (2004). On the interpretation of a concentration index of
inequality. Health Econ, 13(7), 649-656. doi: 10.1002/hec.884
Leake, J. (1984). Expenditures on dental services in Canada, Canadian provinces and territories
1960-1980. Journal (Canadian Dental Association), 50(5), 362.
Listl, S. (2015). Countries with public dental care coverage have lower social inequalities in the
use of dental services than countries without such coverage. J Evid Based Dent Pract,
15(1), 41-42. doi: 10.1016/j.jebdp.2014.12.001
Mackenbach, J. P. (2003). An analysis of the role of health care in reducing socioeconomic
inequalities in health: the case of the Netherlands. Int J Health Serv, 33(3), 523-541.
Mackenbach, J. P., & Kunst, A. E. (1997). Measuring the magnitude of socio-economic
inequalities in health: an overview of available measures illustrated with two examples
from Europe. Social science & medicine, 44(6), 757-771.
Mandal, M., Edelstein, B. L., Ma, S., & Minkovitz, C. S. (2014). Changes in state policies
related to oral health in the United States, 2002‐ 2009. Journal of Public Health
Dentistry, 74(4), 266-275.
Manski, R., Moeller, J., Chen, H., Widstrom, E., Lee, J., & Listl, S. (2015). Disparity in dental
coverage among older adult populations: a comparative analysis across selected European
countries and the USA. Int Dent J, 65(2), 77-88. doi: 10.1111/idj.12139
Manski, R. J., Moeller, J., Chen, H., St Clair, P. A., Schimmel, J., Magder, L., & Pepper, J. V.
(2010). Dental care expenditures and retirement. J Public Health Dent, 70(2), 148-155.
doi: 10.1111/j.1752-7325.2009.00156.x
Manski, R. J., Moeller, J. F., Chen, H., St Clair, P. A., Schimmel, J., Magder, L. S., & Pepper, J.
V. (2009). Dental care coverage transitions. Am J Manag Care, 15(10), 729-735.
72
Manski, R. J., Moeller, J. F., Chen, H., St Clair, P. A., Schimmel, J., & Pepper, J. V. (2012).
Wealth effect and dental care utilization in the United States. J Public Health Dent,
72(3), 179-189. doi: 10.1111/j.1752-7325.2012.00312.x
Marmot, M., Friel, S., Bell, R., Houweling, T. A., Taylor, S., & Health, C. o. S. D. o. (2008).
Closing the gap in a generation: health equity through action on the social determinants
of health. The Lancet, 372(9650), 1661-1669.
McGinn-Shapiro, M. (2008). Medicaid coverage of adult dental services. State health Policy
monitor, 2(2), 1-6.
McGrail, K. M., Van Doorslaer, E., Ross, N. A., & Sanmartin, C. (2009). Income-related health
inequalities in Canada and the United States: a decomposition analysis. American Journal
of Public Health, 99(10), 1856.
Mejia, G., Jamieson, L. M., Ha, D., & Spencer, A. J. (2014). Greater inequalities in dental
treatment than in disease experience. J Dent Res, 93(10), 966-971. doi:
10.1177/0022034514545516
Mertz, E. (2008). Registered Dental Hygienists in Alternative Practice: increasing access to
dental care in California: Center for the Health Professions, University of California, San
Francisco.
O'Donnell Owen, v. D. E., Adam, W., & Lindelow, M. (2007). Analyzing Health equity using
household survey data. Washington DC: USA: The World Bank.
OECD.StatExtracts database. Paris: Organisation for Economic Co-operation and Development;
n.d. [Accessed 2013 Jan 28] Available: http://stats.oecd.org/#
OECD. (2014). United States: Tackling high inequalities creating opportunities for all. Retrieved
from http://www.oecd.org/unitedstates/Tackling-high-inequalities.pdf
Pickett, K. E., & Wilkinson, R. G. (2015). Income inequality and health: a causal review. Soc Sci
Med, 128, 316-326. doi: 10.1016/j.socscimed.2014.12.031
Prus, S. G. (2011). Comparing social determinants of self-rated health across the United States
and Canada. Soc Sci Med, 73(1), 50-59. doi: 10.1016/j.socscimed.2011.04.010
Quiñonez, C., & Grootendorst, P. (2011). Equity in dental care among Canadian households. Int
J Equity Health, 10, 14. doi: 10.1186/1475-9276-10-14
Quiñonez, C., Sherret, L., Grootendorst, P., Shim, M., Azarpazhooh, A., & Locker, D. (2005).
An environmental scan of publicly financed dental care in Canada. Community Dental
Health Services Research Unit and Office of the Chief Dental Officer, Health Canada.
73
Quiñonez, C., Locker, D., Grootendorst, P., Sherret, L., Azarpazhooh, A., & Figueiredo, R.
(2007). An environmental scan of public dental programs in Canada. Community Dental
Health Services Research Unit, Dental Research Institute, Faculty of Dentistry,
University of Toronto.
Rajotte, J. (2013). Income Inequality in Canada: An overview. Report of the Standing
Committee on Finance. Retrieved from
www.parl.gc.ca/content/hoc/Committee/412/FINA/.../finarp03-e.pdf
Ramraj, C., Azarpazhooh, A., Dempster, L., Ravaghi, V., & Quiñonez, C. (2012). Dental
treatment needs in the Canadian population: analysis of a nationwide cross-sectional
survey. BMC Oral Health, 12, 46. doi: 10.1186/1472-6831-12-46
Ramraj, C., Sadeghi, L., Lawrence, H. P., Dempster, L., & Quiñonez, C. (2013). Is accessing
dental care becoming more difficult? Evidence from Canada's middle-income population.
PLoS One, 8(2), e57377. doi: 10.1371/journal.pone.0057377
Ramraj, C., Weitzner, E., Figueiredo, R., & Quiñonez, C. (2014). A macroeconomic review of
dentistry in Canada in the 2000s. J Can Dent Assoc, 80, e55.
Ravaghi, V., Quiñonez, C., & Allison, P. J. (2013a). Comparing inequalities in oral and general
health: findings of the Canadian Health Measures Survey. Can J Public Health, 104(7),
e466-471.
Ravaghi, V., Quiñonez, C., & Allison, P. J. (2013b). The magnitude of oral health inequalities in
Canada: findings of the Canadian health measures survey. Community Dent Oral
Epidemiol, 41(6), 490-498. doi: 10.1111/cdoe.12043
Regidor, E. (2004a). Measures of health inequalities: part 1. J Epidemiol Community Health,
58(10), 858-861. doi: 10.1136/jech.2003.015347
Regidor, E. (2004b). Measures of health inequalities: part 2. J Epidemiol Community Health,
58(11), 900-903. doi: 10.1136/jech.2004.023036
Sanders, A. E., Slade, G. D., John, M. T., Steele, J. G., Suominen-Taipale, A. L., Lahti, S., . . .
Allen, P. F. (2009). A cross-national comparison of income gradients in oral health
quality of life in four welfare states: application of the Korpi and Palme typology. J
Epidemiol Community Health, 63(7), 569-574. doi: 10.1136/jech.2008.083238
74
Sanders, A. E., & Spencer, A. J. (2004). Social Inequality: Social inequality in perceived oral
health among adults in Australia. Australian and New Zealand journal of public health,
28(2), 159-166.
Schwendicke, F., Dorfer, C. E., Schlattmann, P., Page, L. F., Thomson, W. M., & Paris, S.
(2015). Socioeconomic inequality and caries: a systematic review and meta-analysis. J
Dent Res, 94(1), 10-18. doi: 10.1177/0022034514557546
Sgan-Cohen, H. D., Evans, R. W., Whelton, H., Villena, R. S., MacDougall, M., Williams, D.
M., . . . Task, G. (2013). IADR Global Oral Health Inequalities Research Agenda (IADR-
GOHIRA(R)): a call to action. J Dent Res, 92(3), 209-211. doi:
10.1177/0022034512475214
Shaefer, H. L., & Miller, M. (2011). Improving access to oral health care services among
underserved populations in the US: is there a role for mid-level dental providers? Journal
of health care for the poor and underserved, 22(3), 740-744.
Sheiham, A., Conway, D., & Chestnutt, I. (2015). 1.1 Impact of oral diseases and oral health
inequalities. Social inequalities in oral health: from evidence to action, 4.
Shen, J., Wildman, J., & Steele, J. (2013). Measuring and decomposing oral health inequalities in
an UK population. Community Dent Oral Epidemiol, 41(6), 481-489. doi:
10.1111/cdoe.12071
Siddiqi, A., & Hertzman, C. (2007). Towards an epidemiological understanding of the effects of
long-term institutional changes on population health: a case study of Canada versus the
USA. Soc Sci Med, 64(3), 589-603. doi: 10.1016/j.socscimed.2006.09.034
Siddiqi, A., Kawachi, I., Keating, D. P., & Hertzman, C. (2013). A comparative study of
population health in the United States and Canada during the neoliberal era, 1980-2008.
Int J Health Serv, 43(2), 193-216.
Siddiqi, A., Ornelas, I. J., Quinn, K., Zuberi, D., & Nguyen, Q. C. (2013). Societal context and
the production of immigrant status-based health inequalities: a comparative study of the
United States and Canada. J Public Health Policy, 34(2), 330-344. doi:
10.1057/jphp.2013.7
Sisson, K. L. (2007). Theoretical explanations for social inequalities in oral health. Community
Dent Oral Epidemiol, 35(2), 81-88. doi: 10.1111/j.1600-0528.2007.00354.x
Snow, P., & McNally, M. E. (2009). Examining the implications of dental treatment costs for
low-income families. Journal (Canadian Dental Association), 76, a28-a28.
75
Somkotra, T., & Detsomboonrat, P. (2009). Is there equity in oral healthcare utilization:
experience after achieving Universal Coverage. Community Dent Oral Epidemiol, 37(1),
85-96. doi: 10.1111/j.1600-0528.2008.00449.x
Stamm, J. W., Health, C., Canada, W., & Directorate, C. H. S. (1986). Dental Care Programs in
Canada: Historical Development, Current Status and Future Directions: Health and
Welfare Canada.
Statistics Canada. Labour force survey estimates (LFS), supplementary unemployment rates by
sex and age group, annual (CANSIM Table 282-0086). Ottawa: Statistics Canada, 2012.
Tapp, S. (2009, July 6). Canadian labour market developments: recesiioon impacts, recent trends
and future outlook. Ottawa: Office of the parliamentary budget officer. Retirieved from
www.parl.gc.ca/pbo- dpb
Thompson, B., Cooney, P., Lawrence, H., Ravaghi, V., & Quiñonez, C. (2014). Cost as a barrier
to accessing dental care: findings from a Canadian population-based study. J Public
Health Dent, 74(3), 210-218. doi: 10.1111/jphd.12048
Uswak, G., & Keller-Kurysh, E. (2012). Influence of private practice employment of dental
therapists in Saskatchewan on the future supply of dental therapists in Canada. J Dent
Educ, 76(8), 1092-1101.
Wagstaff, A. (2005). The bounds of the concentration index when the variable of interest is
binary, with an application to immunization inequality. Health Econ, 14(4), 429-432. doi:
10.1002/hec.953
Wagstaff, A., Paci, P., & Van Doorslaer, E. (1991). On the measurement of inequalities in
health. Social science & medicine, 33(5), 545-557.
Waldman, H. B. (1980). The reaction of the dental profession to changes in the 1970s. American
Journal of Public Health, 70(6), 619-624.
Waldman, H. B. (1980). The reaction of the dental profession to changes in the 1970s. American
Journal of Public Health, 70(6), 619-624.
Wall, T. (2012). Recent trends in dental emergency department visits in the United
States:1997/1998 to 2007/2008. J Public Health Dent, 72(3), 216-220. doi:
10.1111/j.1752-7325.2012.00339.x
Wall, T. P., Vujicic, M., & Nasseh, K. (2012). Recent trends in the utilization of dental care in
the United States. Journal of Dental Education, 76(8), 1020-1027.
76
Wamala, S., Merlo, J., & Boström, G. (2006). Inequity in access to dental care services explains
current socioeconomic disparities in oral health: the Swedish National Surveys of Public
Health 2004–2005. Journal of epidemiology and community health, 60(12), 1027-1033.
77
Appendix A: Statistics Canada Microdata Research Contract
78
Appendix B: Characteristics of surveys used in analyses.
Survey Year Design Sample
size
Sample characteristics Sampling
method
Data collection
method
NCNS 1970-1972 Cross-sectional 19,590 Canadian population
Ages 0-100
Excluded Indians in bands and
persons living in institutions and
military camps from the ten
provinces.
Stratified,
multi-stage
Household
interview;
Clinical
examination
CHMS 2007-2009 Cross-sectional 5,600 Canadian population
Ages 6-79
Excluded persons living on Indian
Reserves or Crown lands, residents
of institutions, full-time members of
the Canadian Armed Forces and
residents of certain remote regions.
Stratified,
multi-stage
Personal household
and individual
interviews;
mobile examination
centre
NHANES I 1971-1974 Cross-sectional 32,000 Civilian, non-institutionalized U.S.
population
Ages 1-74
Excluded persons residing upon
reservation lands.
Stratified,
multi-stage
Household
interview;
Medical and dental
examination
NHANES 2007-2008 Cross-sectional 10,149 Civilian, non-institutionalized U.S.
population
Ages 0-80+
Stratified,
multi-stage
Household
interview;
Health examination
in mobile
examination center
79
Appendix C: Changes in outcome based on tooth counts (NCNS 1970-1972).
Table A1. Oral health outcomes of sample populations by tooth count (weighted proportions)
Presence of one or more decayed teeth Presence of one or more filled teeth
Tooth count 00-32 Tooth count 00-28 Tooth count 00-32 Tooth count 00-32
Overall 62.3 (59.2, 65.3) 59.5 (56.3, 62.6) 72.9 (70.2, 75.3) 72.7 (70.0, 75.2)
Income
Lowest 68.1 (61.3, 74.0) 64.7 (57.8, 71.0) 69.1 (63.1, 74.6) 68.7 (62.7, 74.2)
Lower Middle 68.1 (60.6, 74.8) 64.9 (57.3, 71.8) 71.4 (65.0, 77.1) 71.4 (64.9, 77.0)
Middle 66.9 (60.1, 73.1) 64.3 (57.3, 70.7) 72.9 (66.9, 78.1) 72.6 (66.7, 77.8)
Upper Middle 65.1 (54.3, 74.6) 64.9 (54.1, 74.4) 66.0 (56.2, 74.6) 65.7 (55.9, 74.3)
Highest 58.8 (51.8, 73.9) 53.5 (37.4, 68.9) 74.7 (61.4, 84.5) 74.7 (61.4, 84.5)
Table A2. Concentration Indices by tooth count (NCNS 1970-1972).
Presence of one or more
decayed teeth
(Tooth count 00-32)
Presence of one or more
decayed teeth
(Tooth count 00-28)
Presence of one or more
filled teeth
(Tooth count 00-32)
Presence of one or more
filled teeth
(Tooth count 00-28)
Observed Expected Observed Expected Observed Expected Observed Expected
Concentration
Index -0.142 0.014 -0.134 0.018 0.076 0.032 0.079 0.031
Standard
Error 0.037 0.031 0.038 0.032 0.030 0.023 0.030 0.023
p-value 0.000 0.645 0.000 0.565 0.011 0.169 0.009 0.171
Concentration
index
(Observed-
Expected)
-0.156 -0.152 0.045 0.047
80
Table A3. Decomposition Results for the presence of one or more decayed teeth by tooth count (NCNS 1970-1972).
Based on tooth count 00-32 Based on tooth count 00-28
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.076 0.005 0.000 -0.070 0.005 0.000
60-74yr -0.033 -0.045 0.001 0.001 -0.031 -0.045 0.001 -0.001
Sex Female e
Male 0.032 -0.007 0.000 0.000 0.014 -0.007 0.000 0.000
Age/Sex
Male 20-39yr e
Male 40-59yr 0.015 -0.009 0.000 0.024 -0.009 0.000
Male 60-74yr 0.010 -0.104 -0.001 -0.001 0.005 -0.104 -0.001 -0.001
Income
Lowest e
Lower Middle -0.011 -0.443 0.005 -0.014 -0.443 0.006
Middle -0.028 -0.105 0.003 -0.033 -0.105 0.003
Upper middle -0.018 0.216 -0.004 -0.017 0.216 -0.004
Highest -0.015 0.588 -0.009 -0.005 -0.020 0.588 -0.012 -.007
Education < High school e
High school Grad -0.073 0.031 -0.002 -0.002 -0.085 0.031 -0.003 -0.003
Family Size 0.094 0.049 0.005 0.005 0.128 0.049 0.006 0.006 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
81
Table A4. Decomposition Results for the presence of one or more filled teeth by tooth count (NCNS 1970-1972).
Based on tooth count 00-32 Based on tooth count 00-28
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.025 0.005 0.000 -0.025 0.005 0.000
60-74yr -0.009 -0.045 0.000 0.000 -0.009 -0.045 0.000 0.000
Sex Female e
Male -0.100 -0.007 0.001 0.001 -0.103 -0.007 0.001 0.001
Age/Sex
Male 20-39yr e
Male 40-59yr 0.024 -0.009 0.000 0.023 -0.009 0.000
Male 60-74yr -0.006 -0.104 0.001 0.001 -0.006 -0.104 0.001 0.001
Income
Lowest e
Lower Middle 0.000 -0.443 0.000 0.001 -0.443 0.000
Middle -0.002 -0.105 0.000 -0.001 -0.105 0.000
Upper middle -0.012 0.216 -0.003 -0.012 0.216 -0.002
Highest -0.004 0.588 -0.002 -0.050 -0.003 0.588 -0.002 -0.004
Education < High school e
High school Grad 0.111 0.031 0.003 0.003 0.112 0.031 0.003 0.003
Family Size 0.094 0.049 0.049 0.049 0.032 0.045 0.049 0.049 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
82
Appendix D: Changes in outcome based on income variable (NHANES 2007-2008).
Table A5 Oral health outcomes of sample populations by income variable (weighted proportions and 95% confidence intervals).
Presence of one or more
decayed teeth
Presence of one or more filled
teeth Presence of Edentulism
Household
Income
Family
Income
Household
Income
Family
Income
Household
Income
Family
Income
Overall 21.3 (19.9, 22.8) 83.5 (82.1, 84.7) 4.9 (4.2, 5.6)
Income
Lowest 42.8 (38.8, 46.9) 41.2 (36.5, 46.0) 70.5 (66.6, 74.2) 70.3 (66.7, 73.8) 9.6 (7.8, 1.6) 9.9 (7.7, 12.4)
Lower Middle 26.8 (24.0, 29.7) 31.0 (27.3, 34.9) 80.2 (77.6, 82.6) 82.9 (80.7, 85.0) 6.7 (5.4, 8.3) 6.9 (5.3, 8.9)
Middle 17.0 (13.8, 20.7) 26.8 (23.3, 30.6) 84.6 (80.9, 87.7) 85.5 (81.8, 88.5) 3.8 (2.3, 6.0) 5.9 (4.4, 7.8)
Upper Middle 15.3 (12.4, 18.7) 16.8 (13.9, 20.1) 87.1 (83.8, 89.9) 88.2 (81.9, 92.4) 2.5 (1.4, 4.5) 3.4 (2.2, 5.3)
Highest 10.5 (8.0, 13.6) 10.4 (8.3, 12.8) 70.5 (66.6, 74.2) 92.3 (89.6, 94.4) 2.6 (1.4, 4.6) 2.6 (1.7, 4.2)
Table A6. Concentration Indices(CI) by income variable (NHANES 2007-2008) Presence of one or more decayed
teeth Presence of one or more filled teeth Presence of Edentulism
Household
Income Family Income
Household
Income Family Income
Household
Income Family Income
Observed Expected Observed Expected Observed Expected Observed Expected Observed Expected Observed Expected
CI -0.231 -0.020 -0.230 -0.018 0.113 0.007 0.125 0.008 -0.083 0.003 -0.078 0.001
Standard
Error 0.018 0.011 0.018 0.011 0.014 0.009 0.014 0.009 0.014 0.010 0.014 0.010
p-value 0.000 0.060 0.000 0.097 0.000 0.434 0.000 0.374 0.000 0.731 0.000 0.911
CI
(Observed-
Expected) -0.212 -0.212 0.106 0.117 -0.086 -0.079
83
Table A7. Decomposition Results for the presence of one or more decayed teeth by income (NHANES 2007-2008).
Based on Household Income Based on Family Income
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d
Age
20-39yr e
40-59yr -0.017 0.101 -0.002 -0.011 0.115 -0.001
60-74yr -0.045 -0.085 0.004 0.002 -0.045 -0.054 0.002 0.002
Sex Female e
Male 0.091 0.025 0.002 0.002 0.085 0.020 0.002 0.002
Age/Sex
Male 20-39yr e
Male 40-59yr 0.017 0.120 0.002 0.017 0.122 0.002
Male 60-74yr 0.002 -0.017 0.000 0.002 0.003 0.013 0.000 0.002
Income
Lowest e
Lower Middle -0.077 -0.455 0.035 -0.018 -0.629 0.011
Middle -0.090 -0.027 0.002 -0.048 -0.291 0.014
Upper middle -0.133 0.325 -0.043 -0.128 0.097 -0.012
Highest -0.167 0.733 -0.122 -0.128 -0.259 0.621 -0.161 -0.148
Education < High school e
High school Grad -0.314 0.073 -0.023 -0.023 -0.320 0.074 -0.024 -0.024
Household/Family Size 0.094 0.258 0.022 0.022 -0.024 0.232 0.049 0.049 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
84
Table A8. Decomposition Results for the presence of one or more filled teeth by income (NHANES 2007-2008).
Based on Household Income Based on Family Income
𝜂𝑘a CIb Contrc Agg d 𝜂𝑘
a CIb Contrc Agg d
Age
20-39yr e
40-59yr 0.055 0.101 0.006 0.054 0.115 0.006
60-74yr 0.014 -0.085 -0.001 0.005 0.014 -0.054 -0.001 0.005
Sex Female e
Male -0.043 0.025 -0.001 -0.001 -0.042 0.020 -0.001 -0.001
Age/Sex
Male 20-39yr e
Male 40-59yr 0.003 0.120 0.000 0.003 0.122 0.000
Male 60-74yr 0.001 -0.017 0.000 0.000 0.001 0.013 0.000 0.000
Income
Lowest e
Lower Middle 0.015 -0.455 -0.007 0.007 -0.629 -0.004
Middle 0.011 -0.027 0.000 0.011 -0.291 -0.003
Upper middle 0.018 0.325 0.006 0.019 0.097 0.002
Highest 0.025 0.733 0.019 0.018 0.042 0.621 0.026 0.021
Education
< High school e
High school
Grad 0.037 0.073 0.003 0.003 0.037 0.074 0.003 0.003
Household/Family Size 0.094 -0.027 0.022 0.022 -0.005 -0.016 0.049 0.049 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
85
Table A9. Decomposition Results for the presence of edentulism by income (NHANES 2007-2008).
Based on Household Income Based on Family Income
𝜂𝑘a CI Contr % 𝜂𝑘
a CI Contr %
Age
20-39yr e
40-59yr 0.111 0.098 0.011 0.119 0.115 0.014
60-74yr 0.125 -0.112 -0.014 -0.003 0.133 -0.077 -0.010 0.003
Sex Female e
Male -0.090 0.029 -0.003 -0.003 -0.092 0.023 -0.002 -0.002
Age/Sex
Male 20-39yr e
Male 40-59yr 0.013 0.119 0.002 0.011 0.124 0.001
Male 60-74yr 0.010 -0.038 0.000 0.002 0.010 -0.008 0.000 0.001
Income
Lowest e
Lower Middle -0.030 -0.440 0.013 0.003 -0.617 -0.002
Middle -0.075 -0.008 0.001 -0.048 -0.274 0.013
Upper middle -0.108 0.338 -0.037 -0.118 0.113 -0.013
Highest -0.124 0.737 -0.091 -0.114 -0.186 0.627 -0.117 -0.119
Education < High school e
High school Grad -0.211 0.079 -0.017 -0.017 -0.218 0.079 -0.017 -0.017
Household/Family Size 0.094 0.190 0.025 0.025 -0.058 0.200 0.051 0.051 aElasticity; b Concentration Index; c Contribution; d Aggregate contribution e Reference group.
86
Appendix E: Significance Testing
Table A10. Test of significance between observed and expected concentration indices for each outcome (t-values).
Presence of one or more decayed
teeth
Presence of one or more filled teeth Presence of edentulism
Canada 1970-1972 -256.2 115.9 -941.1
Canada 2007-2009 -320.8 142.2 -375.2
United States 1971-1974 -1027.0 317.6 -1530.9
United States 2007-2009 -697.4 594.6 -357.9
Table A11. Test of significance between concentration indices for oral health outcomes within surveys (t-values).
Presence of one or more filled teeth Prevalence of edentulism
Canada
1970-1972
Presence of >1 decayed teeth -444.8 47.0
Presence of >1 filled teeth - 658.8
Prevalence of edentulism - -
Canada 2007-2009
Presence of >1 decayed teeth -405.0 -173.2
Presence of >1 filled teeth - 367.9
Prevalence of edentulism - -
United States 1971-
1974
Presence of >1 decayed teeth -293.4 -1.5
Presence of >1 filled teeth - 2953.3
Prevalence of edentulism - -
United States 2007-
2008
Presence of >1 decayed teeth -1023.4 -436.4
Presence of >1 filled teeth - 747.9
Prevalence of edentulism - -
87
Table A12. Test of significance between concentration indices for presence of one or more decayed teeth (t-values).
Canada 2007-2009 United States 1971 -1974 United States 2007-2009
Canada 1970-1972 45.1 56.0 184.9
Canada 2007-2009 - -8.8 113.4
United States 1971 -1974 - - 212.2
Table A13. Test of significance between concentration indices for presence of one or more filled teeth (t-values).
Canada 2007-2009 United States 1971 -1974 United States 2007-2009
Canada 1970-1972 58.1 -476.4 -106.5
Canada 2007-2009 - -566.9 -177.8
United States 1971 -1974 - - 520.7
Table A14. Test of significance between concentration indices for presence of edentulism (t-values).
Canada 2007-2009 United States 1971 -1974 United States 2007-2009
Canada 1970-1972 -309.5 5.4 -296.1
Canada 2007-2009 - -374.7 18.9
United States 1971 -1974 - - -362.3