Antenatal care measures increase iron and

folic acid receipt and consumption among

pregnant women in Bihar, India

Amanda Wendt, Melissa Young, Rob Stephenson, Amy

Webb Girard, Usha Ramakrishnan, Reynaldo Martorell

Nutrition and Health Sciences, Laney Graduate School, Emory University

Institute of Public Health, University of Heidelberg, Germany*

Amanda Wendt, PhD MS

Alexander von Humboldt

Postdoctoral Research Fellow

Institute of Public Health

University of Heidelberg

Germany

Anemia: A global health issue

Iron deficiency is one of the most common causes of anemia

Almost one third of the world’s population suffers from anemia

– 1.62 billion

9 out of 10 live in developing countries

World India Bihar

Women

(15-49y)30% 48% 60%

Pregnant

Women38% 54% 58%

WHO: Global anemia prevalence and trends; NFHS-4; Stevens et al, 2013

Government of India Policy

100mg/day for at least 100 days starting after first

trimester & 100 days post-partum

Iron and Folic Acid (IFA)

Guidelines

Kapil, 2014

Maternal Health Indicators Bihar

Maternal Mortality Ratio 274

Any Antenatal Check-Up 85%

3+ Antenatal Care (ANC) visits 37%

At least one Tetanus Toxoid

(TT)85%

Consumed IFA for 100+ days 13%

Maternal Health in Bihar

Annual Health Survey, 2012-3

Rationale

Previous studies focus on individual factors of IFA adherence

IFA distribution and counseling are implemented through

government facilities and health workers

Facilities may impact success of IFA receipt and consumption

among women

Study Objective

Objective: To examine individual and facility-level

determinants of IFA receipt and consumption among pregnant

women in rural Bihar

Hypothesis: IFA receipt and IFA consumption will be associated

with facility level factors as well as individual factors

Survey & Sample

District Level Household Survey Round 3, 2007-08

Cross-Sectional

Representative at national, state, district level

Multi-stage stratified probability proportional to size

sampling design

Final Study Populations: Flowchart of Exclusions of

Ever-Married Women

46,840 ever-married women (DLHS-3: Bihar)

21,633 women who had an index pregnancy in the study period

12,609 women who received antenatal care

12,420 women with complete data

Women covered by one Health Sub-

Center (HSC) (n=7,765)

21,331 women who had live birth in the study period

Women who received IFA

(n=2,905)

Sample A

Outcome: IFA Receipt

Sample B

Outcome: Adequate

IFA Consumption

Key Variables

Outcomes:

IFA Receipt

IFA Consumption

Individual Level Variables:

Age

Age of Marriage

Maternal Education

Gender Composition of Living

Children

Birth Order of Index Pregnancy

Caste

Religion

Wealth Index Quintiles

ANC Timing & Frequency

ANC Factors

ANC Practice

ANC Counseling

Facility Level Variables:

Health Sub-Center(HSC) Factors

Village Health Day &

Monitoring

Personnel

Infrastructure

IFA Supply

Distance to Nearest HSC

Factor Analysis

Purpose: to reduce the number of variables

Antenatal Care: 13

Heath Sub-Center: 9

Principal components analysis

Orthogonal rotation – varimax method

Polychoric correlation matrices

More accurate correlations between categorical variables

Characteristics and Rotated Factor Loadings for Antenatal Care Factorsa

ANCb Practices ANCb Counseling

Eigenvalue 7.16 1.60

Proportion Variance Explained 0.73 0.16

Rotated Factor Loadings

Blood Tested 0.87 -

Weight Measured 0.81 -

Abdomen Examined 0.81 -

Sonogram/Ultrasound Taken 0.80 -

Breast Examined 0.76 -

Height Measured 0.76 -

Delivery Date Given 0.72 -

Importance of Cleanliness at Delivery - 0.83

Better Nutrition for Mother and Child -0.80

Family Planning for Spacing - 0.75

Breastfeeding - 0.75

Importance of Institutional Delivery - 0.73

Nutrition Advice - 0.59

a Factor loadings ≤|0.5|

are not shown.

b ANC: Antenatal Care

Characteristics and Rotated Factor Loadings for Health Sub-Center Factorsa

Village Health

Day: Support &

Monitoring

Personnel

Characteristics

Sub Center

Infrastructure

Eigenvalue 1.10 0.63 0.54

Proportion Variance Explained 0.42 0.24 0.21

Rotated Factor Loadings

Observation of any Village Health Day 0.56 - -

Written feedback from PHCc 0.50 - -

VHSCd present in some villages in HSCb area 0.42 - -

Medical Officer visited HSCb in previous month 0.36 - -

Received and utilized untied funds from previous financial

year -

0.37

-

HSCb Personnel - 0.46 -

HSCb Training - 0.44 -

Present Condition of Existing Building - - 0.44

Water Available at Sub Center - - 0.43

a Factor loadings ≤|0.35| are not shown; b Health Sub-Center; c Primary Health Center; d VHSC: Village Health and

Sanitation Committee

Individual Level Factors of Study Population by Prevalence of Iron and

Folic Acid Receipt and ConsumptionReceived Any IFA

(Sample A)Consumed IFA for 90+ Days

(Sample B)

Chi-Square Chi-Square

N % p-value N % p-value

Overall 7765 37.4 2905 23.8

Age

<20 y 1237 37.8 <0.0001 1148 22.4 0.2352

20-24 y 3071 42.0 468 23.1

>24 y 3457 33.2 1289 25.2

Mother's Education

None / Don't Know 4812 28.7 <0.0001 1382 14.6 <0.0001

1-4 y 610 36.7 224 18.3

5-8 y 1202 46.5 559 25.9

≥9 y 1141 64.9 740 40.8

Wealth Index Quintiles

Poorest 2154 26.5 <0.0001 571 13.0 <0.0001

Second 3061 33.4 1023 19.3

Middle 1510 44.4 671 23.3

Fourth 835 59.0 493 37.8

Richest 205 71.7 147 52.3

Individual Level Factors of Study Population by Prevalence of Iron and Folic Acid Receipt

and ConsumptionReceived Any IFA

(Sample A)Consumed IFA for 90+ Days

(Sample B)

Chi-Square Chi-Square

N % p-value N % p-value

Overall 7765 37.4 2905 23.8

ANC Timing & Frequency <0.0001 <0.0001

Early Enrollment & ≥4 ANC Visits 837 71.7 600 49.5

Late Enrollment & ≥4 ANC Visits 215 63.7 137 40.9

Early Enrollment & <4 ANC Visits 2232 40.3 900 17.6

Late Enrollment & <4 ANC Visits 4481 28.3 1268 14.1

ANC Practices <0.0001 <0.0001

None 3903 19.2 749 9.2

1-3 2630 48.9 1287 19.6

4-7 1232 70.5 869 42.4

ANC Counseling <0.0001 <0.0001

None 2841 21.3 605 15.4

1-3 3376 41.0 1385 20.4

4-6 1548 59.1 915 34.3

IFA: Iron and Folic Acid; ANC: Antenatal Care; Early Enrollment: 1st trimester; Late Enrollment: 2nd – 3rd trimester

Multilevel Modeling of Any IFA Receipt During Last Pregnancy (Sample A)

ParameterAll Factors +

InteractionsParameter

All Factors +

Interactions

OR (95% CI) OR (95% CI)

Age of Marriage Antenatal Care Timing and Frequency

<18 y ref Early enrollment and ≥4 visits 3.53 (2.44, 5.11)

≥18 y 1.21 (1.05, 1.39) Late enrollment and ≥4 visits 2.44 (1.69, 3.51)

Mother's Education Early enrollment and <4 visits 1.36 (1.19, 1.55)

None / Don't Know ref Late enrollment and <4 visits ref

1-4 y 0.98 (0.79, 1.21) Antenatal Care Factors

5-8 y 1.26 (1.06, 1.49) Practice 13.12 (9.51, 18.09)

≥9 y 1.67 (1.35, 2.06) Counseling 2.61 (2.12, 3.21)

Religion Practice*Counseling Interaction 0.37 (0.25, 0.56)

Hindu ref Timing & Frequency * Practice 0.68 (0.56, 0.82)

Muslim & Others 0.79 (0.66, 0.94) Community Level Random Effect (SE) 0.6259 (0.0473)

Wealth Index Quintiles

Poorest ref

Second 1.12 (0.97, 1.29)

Middle 1.16 (0.97, 1.39)

Fourth 1.30 (1.03, 1.64)

Richest 1.31 (0.87, 1.97)

ANC: Antenatal Care; HSC: Health Sub-Center; PHC: Primary Health Center; SE: Standard Error; AIC: Akaike Information Criterion; Factor ORs reflect a 1-unit change

Also adjusted for: age, gender composition of living children, birth order, caste, husband’s education, HSC Factors, IFA stock out, Distance to nearest HSC

Multilevel Modeling of IFA Consumption for 90+ Days During Last Pregnancy (Sample B)Parameter All Factors Parameter All Factors

OR (95% CI) OR (95% CI)

Mother's Education Antenatal Care Timing and Frequency

None / Don't Know ref Early enrollment and ≥4 visits 3.4 (2.52, 4.59)

1-4 y 1.05 (0.68, 1.62) Late enrollment and ≥4 visits 3.19 (2.03, 5.01)

5-8 y 1.31 (0.95, 1.80) Early enrollment and <4 visits 1.05 (0.81, 1.37)

≥9 y 1.75 (1.24, 2.48) Late enrollment and <4 visits ref

Caste Antenatal Care Factors

Scheduled Castes & Tribes 0.71 (0.53, 0.97) Practice 2.62 (1.86, 3.71)

Others ref Counseling 1.08 (0.83, 1.4)

Wealth Index Quintiles IFA Available on Day of Survey

Poorest ref Yes 1.37 (1.04, 1.82)

Second 1.18 (0.84, 1.66) No ref

Middle 0.96 (0.65, 1.4)

Fourth 1.31 (0.86, 2.00) Community Level Random Effect 0.781 (0.108)

Richest 2.05 (1.17, 3.56)

ANC: Antenatal Care; HSC: Health Sub-Center; PHC: Primary Health Center; AIC: Akaike Information Criterion; Factor ORs reflect a 1-unit change;

Also adjusted for: age, age of marriage, gender composition of living children, birth order, religion, husband’s education, HSC Factors, Distance to nearest HSC

Summary

IFA Receipt

All ANC components were significantly, positively associated

with odds of IFA receipt

No health sub-center factors were significant

Significant variation remained among facility sites

IFA Consumption

ANC practice and frequency were significantly, positively

associated with odds of IFA consumption

Women were more likely to consume IFA for 90+ days when

their HSC had IFA in stock

Significant variation remained among facility sites

Limitations & Strengths

Limitations

Variables based on self-report

IFA supply based on availability on the day of survey

High use of private sector for ANC and IFA purchasing

All women attended ANC

Did not have data on IFA counseling

Strengths

State representative dataset

ANC quality factors

Health Sub-Center capacity factors

Health Sub-Centers were linked to the women they serve

Multilevel modeling

Conclusions and Implications

Measuring ANC quality can reveal insight on how ANC is associated with

outcomes more than timing & frequency alone

ANC quality should be measured when assessing any singular ANC service

ANC interactions may show a quality threshold in terms of IFA receipt –

highlighting the importance of services and counseling during initial

appointments

IFA supply does play a role in successful IFA consumption

A consistent supply must be ensured to support adequate IFA consumption

Variation between sites may indicate facilities and/or communities do impact

IFA receipt and consumption beyond the individual level