QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jasti, S. Articles by Bentley, M. E. Search for Related Content PubMed PubMed Citation Articles by Jasti, S. Articles by Bentley, M. E.

---

Community and International Nutrition

Pill Count Adherence to Prenatal Multivitamin/Mineral Supplement Use among Low-Income Women¹

Sunitha Jasti^*,2, Anna Maria Siega-Riz, Mary E. Cogswell^**, Abraham G. Hartzema and Margaret E. Bentley

^* Department of Family, Nutrition, and Exercise Sciences, Queens College of City University of New York, Flushing, NY 11367; Department of Nutrition, School of Public Health, University of North Carolina at Chapel Hill and Carolina Population Center, Chapel Hill, NC 27516; ^** Maternal and Child Nutrition Branch, Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, Atlanta, GA 30341; and Pharmacy Health Care Administration, University of Florida, Gainesville, FL 32610

²To whom correspondence should be addressed. E-mail: sunitha_jasti{at}qc.edu.

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION LITERATURE CITED

 
In the United States, the prevalence of third trimester anemia among low-income pregnant women is 29% and has not improved since the 1980s. Although low adherence has been linked to the ineffectiveness of iron supplementation programs, data regarding adherence to supplementation in low-income women are currently lacking. Hence this study was conducted to better understand the factors associated with adherence to the use of iron-containing prenatal multivitamin/mineral supplements among low-income pregnant women. Adherence to supplement use was assessed by pill counts among 244 pregnant women of 867 women who were initially randomized to receive 1 of 3 prenatal supplements. All women received care at a public prenatal clinic. Maternal characteristics associated with adherence were identified using predictive modeling. Women took 74% of supplements as prescribed. Adherence was higher among non-Hispanic white women than among non-Hispanic black women (79% vs. 72%, P 0.01). Interactions of ethnicity with age group, smoking status, and prior supplement use were significant. Multivariate regression analysis stratified by ethnicity revealed that among the white women education beyond high school, unmarried status, nulligravidity, and smoking were positively associated with adherence. In contrast, among the black women, supplement use 3 mo prior to current pregnancy and no loss of appetite were positively associated with adherence. Further research investigating the influence of cultural factors is necessary to better understand adherence to supplement use and the differences in adherence among ethnic groups.

KEY WORDS: • adherence • prenatal supplements • pregnant women • anemia • iron deficiency

One of the national health objectives designated by the U.S. Department of Health and Human Services is to reduce third trimester anemia among low-income women to 20% by the year 2010 (1). Maternal anemia during pregnancy is associated with preterm delivery, low birth weight, and increased perinatal mortality and morbidity (2–5). Because many women enter pregnancy with little to no iron stores (6), the recommendation of daily oral iron supplementation has been a routine practice in prenatal clinics in the United States (7,8). Oral iron supplementation has been clinically shown to prevent iron deficiency (9–15). However, in public health settings the effectiveness of iron supplementation programs in reducing maternal anemia has been less than ideal (16–18).

The effectiveness of supplementation has been linked to the adherence of usage (16,18–21), but limited information is available about the adherence to supplementation during pregnancy among low-income women in the United States. Furthermore, most available information is self-reported, thereby potentially overestimating adherence (18,21–23), and the dosage and duration of supplementation are generally unknown. Data from the 3rd National Health and Nutrition Examination Survey 1988–1994 (NHANES III)³ indicate that 63% of low-income pregnant women who were surveyed (n = 132) reportedly took iron-containing supplements during the previous month (24). Among a culturally diverse group of 344 low-income pregnant women who were routinely prescribed prenatal supplements, 86% reportedly took the supplements at least 4 times/wk (25).

Previous research indicated that low-income women are less likely to consume supplements than higher income women (24,26), but determinants of prenatal supplement use among low-income women are not well described. In fact, determinants of supplement use were not reported separately for low-income pregnant women who participated in NHANES III (24). Qualitative studies in developing countries (27,28) reported reasons for nonadherence to iron supplements such as inadequate supply, poor access to and utilization of prenatal care services, inadequate counseling, and certain beliefs, including fear of iron supplements causing too much blood or a big baby, making delivery more difficult. Forgetting to take the pills and lack of motivation were also reported (28). Little is known about reasons for nonadherence in the United States, although side effects are often cited as a reason (29) and have become the basis for several daily vs. weekly and special formulation iron supplementation trials (20,29–32). Although a definite causal relation between high iron doses and gastrointestinal effects has been shown (33,34), the relation between side effects and adherence has been disputed (19,20). Very few studies included measures of both side effects and adherence to iron supplements to enable a quantitative examination of the association between the two (20,30,35).

Knowledge of iron-containing supplement use determinants would help in designing and targeting interventions that improve adherence to groups at higher risk of nonuse, subsequently preventing iron deficiency anemia. The objectives of this analysis were to estimate adherence to supplement use among low-income pregnant women in a public health setting and to evaluate factors associated with adherence. Factors examined included maternal sociodemographic, clinical, and health behavior characteristics as well as side effects. Differences in iron dose in the supplements were controlled for in analyses.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION LITERATURE CITED

 
    Study population and data collection. Data were collected as part of the Iron Supplementation Study, a randomized, double-blind, clinical control study of pregnant women. The study recruited women who were receiving care at the Wake County Human Resources prenatal clinic in Raleigh, North Carolina, between 1997 and 1999. Women were recruited into the study at their first prenatal visit after they provided written informed consent and if they met the following criteria: <20 wk gestation at the time of enrollment, spoke English, had not used supplements containing iron in the month prior to enrollment, had a singleton pregnancy, and had no known history of blood disorders. All women who received a positive pregnancy test at the clinic were given a prescription for folic acid supplements along with an appointment date for the first prenatal visit. This date usually occurred 8–12 wk after the pregnancy test. Thus, most women took only folic acid supplements and not iron-containing supplements prior to enrollment in the study.

All women with serum ferritin 40 µg/L (serum ferritin assessed at the time of enrollment) were randomized to receive daily supplements containing 30 or 60 mg of elemental iron. Women with serum ferritin > 40 µg/L who were nonanemic were randomized to receive daily supplements containing 0 or 30 mg of elemental iron. Women with serum ferritin > 40 µg/L who were anemic were excluded from the study and referred to the clinic physician. Supplements were provided free of charge to all participants. Baseline information was abstracted from medical records for sociodemographic and clinical characteristics and health behaviors. Self-administered questionnaires provided information on the women’s supplement use 3 mo prior to current pregnancy, adherence to supplement use during the study period, and side effects experienced. Adherence was also measured using pill counts. The study was approved by the Institutional Review Board of the University of North Carolina, School of Medicine; the Institutional Review Committee of Wake Medical Center; and the Human Subjects Committee at CDC.

    Recruitment and participation. During the recruitment period, a total of 4413 women were approached to enter the study, of which 2779 were ineligible to participate. Major reasons for ineligibility were inability to speak English and gestational age 20 wk. Of 1634 eligible women, 967 women agreed to participate in the study (59.2% participation rate), and 867 of them were successfully randomized to study treatments. The remaining 100 women were excluded primarily due to gestational age being 20 wk at the time of randomization, which occurred a few days after the enrollment.

Of the 867 women who were randomized to receive iron treatments, only 549 retrieved their study supplements from the clinic pharmacy. Questionnaire and pill count data were available for 244 of the women, who constituted the sample for this study. The recruitment as well as losses and exclusions from the original study sample of 867 randomized women is shown in Figure 1.

View larger version (25K): [in this window] [in a new window]   FIGURE 1 (A) Recruitment statistics and (B) sample losses at various stages of study participation. *Inclusion criteria: women were <20 wk gestation at the time of enrollment, spoke English, had not used supplements containing iron in the month prior to enrollment, had a singleton pregnancy, and had no known history of blood disorders.

Among a subsample of women for which all 3 measures were available (n = 51), pill counts correlated better than self-reports with the MEMS data (Pearson correlation 0.62 vs. 0.35). Hence adherence was based on pill counts in all analyses. Pill count adherence was defined as the proportion of supplements taken as prescribed and was calculated using the following formula.

Pill count adherence was treated as a continuous variable, with values ranging from 0 to 100%, with 0% indicating no adherence and 100% indicating full adherence. Additionally, a dichotomous categorization defined women as adherent if they took 65% or more of the prescribed supplements, which translates to taking the supplements at least 4 d/wk. A similar definition of adherence was used in previous studies based on self-reported adherence data (25,26).

The number of pill counts that we were able to obtain from each participant varied based on the following factors: length of gestation at the time of recruitment, number of visits to the prenatal clinic, and whether the participant remembered to return the pill bottle at each return visit to the pharmacy. More than half of the women (56%) with pill count data returned only 1 pill bottle during their participation period. Among the women who returned 2 or more pill bottles, within-person variability was relatively low when adherence was categorized into 4 levels of 25% increments (0–25, 26–50, 51–75, 76–100). Values did not change in 60% of the women and changed only by 1 category in 32.4% of the women (adherence decreased by 1 category in 15.7% and increased in 16.7%). Hence only the pill count 1 was used as the dependent variable in all analyses while controlling for the length of gestation at the time of adherence assessment.

    Potential determinants of adherence. Maternal age was categorized into 3 groups: <19, 19–29, and 30 y old. Ethnicity was categorized as non-Hispanic black, non-Hispanic white, and other. Ethnicities other than non-Hispanic black and non-Hispanic white were grouped together and excluded from regression analysis because they represented only a small proportion (7%) of the study sample. Marital status was grouped into 2 categories: unmarried and married. The unmarried category included women who were single, separated, divorced, or widowed. Education was initially categorized into <12, 12, and >12 y; however, because adherence between women with <12 y or 12 y of education did not differ, we combined these 2 groups in further analyses.

The number of self-reported previous pregnancies was categorized into 0 (nulligravid) and 1 (primigravid and multigravid). Prepregnancy BMI, defined as weight in kilograms divided by height in square meters, was based on self-reported prepregnancy weight and height measured at the first prenatal visit. BMI was used as a dichotomous variable reflecting prepregnancy obesity status (BMI > 29) (36).

The period of gestation at the time the women entered prenatal care was assessed based on the women’s self-reported estimates of last menstrual period (LMP) or an ultrasound conducted early in the pregnancy if the LMP was unknown. Gestational age was treated as a continuous variable throughout the study. Current smoking status and supplement use 3 mo prior to current pregnancy, which were self-reported at recruitment, were dichotomized as "yes" or "no." General adherence to prescribed medications (other than study supplements), which was also self-reported, was categorized as "always," "mostly," or "never/sometimes."

Hemoglobin concentration at entry to prenatal care was included to test the association between iron status (women are informed of their hemoglobin values) and adherence. Venous blood samples drawn at the first prenatal visit were routinely analyzed for hemoglobin at the clinic using a HemoCue (HemoCue). Hemoglobin concentration was then abstracted from the medical record and used as a continuous variable.

    Side effects. Women were asked whether they experienced a series of symptoms including stomach discomfort, stomach pain, vomiting, gas, constipation, diarrhea, loss of appetite, abdominal cramps, an unpleasant taste in the mouth, and heartburn. These symptoms are common side effects associated with iron supplement use (37). Dichotomous variables for each symptom reflected any occurrence and no occurrence. Women were also asked to indicate whether they thought the symptoms were caused by supplements or by the pregnancy itself.

    Statistical analysis. Comparisons of the proportion of women with selected sociodemographic characteristics between the study sample (n = 244) and the excluded groups were done using ² tests. Comparisons included the following: study participants vs. the larger group of women who were randomized, but did not participate (n = 867–244); study participants vs. women who retrieved their supplements at the pharmacy, but did not have adherence assessment (n = 549–244); and study participants vs. women who answered questions regarding supplement use, but did not return a pill bottle (n = 437–244). We used t tests to examine the differences in means and ² tests for the differences in proportions of independent variables by ethnicity. We examined the proportion of women who reported each gastrointestinal symptom by dose of iron in the prenatal supplement. Linear regression was used to test the associations between each of the selected determinants and adherence. All linear regressions were controlled for variables that were unique to our study design, including iron dose in supplements, MEMS participation, and period of gestation at the time of adherence assessment (pill count measurement), because these variables were associated with adherence. Ethnicity (non-Hispanic black vs. non-Hispanic white) interacted (P < 0.1) with several independent variables including age group, supplement use 3 mo prior to current pregnancy, and smoking status. Hence we specified separate regression models for the 2 ethnic groups.

All variables showing a modest crude association (P < 0.2) with adherence were entered into a multivariate linear regression full model. Collinearity was assessed before these variables were entered into the model. No 2 variables had a correlation > 0.8, and thus all were entered. We used the F test between pairs of regression models to obtain the best predictive model by backward selection. Significance for independent variables to remain in the final model was set at P < 0.05. Because the distribution of pill count adherence was skewed to the right, we examined plots of residuals against fitted values; variance was constant for the final models.

Two approaches were used to test whether missing data on pill counts biased our results. Heckman’s selectivity model, which generates a test statistic to indicate the direction and significance of selection bias when data are missing on a dependent variable in a linear regression (38), was used. The model specifically addresses the bias that results from using nonrandomly selected samples to estimate behavioral relations and involves 2 steps. An equation for response status (return of pill bottle) is first estimated, using the known characteristics of both respondents and nonrespondents. From this equation, an inverse Mill’s ratio is generated (), which is a monotone decreasing function of the probability that an observation is selected into the sample. is then added along with other independent variables to the second equation, a standard linear regression predicting the outcome, adherence. In the 2-step estimator, serves as the test statistic for measuring selectivity. If the estimate for appears to be significant, then there is selectivity in the sample and therefore results from this model should be used. Otherwise we can assume that the estimates from the uncorrected model are unbiased. The direction and significance of the selection bias can be assessed from the sign and significance of the coefficient in the corrected outcome regression model (38). The second approach was multiple imputation, an advanced technique for estimating missing data, that consists of 3 steps: 1) creating multiple data sets based on an imputation model, 2) running regression analysis on all imputed data sets, and 3) summarizing results from all of those data sets using SAS (39–42).

All analyses except multiple imputation were performed using statistical software STATA (version 7, Stata); multiple imputation was done using SAS (version 8.1, SAS Institute).

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION LITERATURE CITED

 
    Sample characteristics. Sixty-one percent of the women in our study sample (n = 244) were non-Hispanic blacks, 32% were non-Hispanic whites, and the remaing 7% were of other race/ethnicity. Maternal age ranged from 13 to 41 y old, with a mean of 23 y. Fewer than 20% of participants were married and 43% were pregnant for the first time. More than a third of the women had not graduated from high school (Table 1). Gestational age at entry to prenatal care was 12.9 ± 2.8 wk (mean ± SD) and ranged from 7 to 19 wk (data not shown). Compared with women lost from the study, participants were more likely to be nulligravid and older than 30 y (Table 1). There were no other differences in the distribution of sociodemographic characteristics among the sample of women studied and the excluded women (Table 1).

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION LITERATURE CITED

 
Despite efforts to reduce iron deficiency and anemia during pregnancy, information on adherence to iron-containing prenatal supplements in high-risk groups such as low-income women is currently lacking. Detailed knowledge of supplement adherence in this population might contribute to targeting efforts to those subgroups that are at the highest risk. The majority of low-income women who returned at least 1 pill bottle for ascertainment of adherence adhered to supplementation during pregnancy. However, low-income non-Hispanic black women were less likely to adhere to supplementation than low-income non-Hispanic white women. More importantly, factors associated with adherence differed among low-income black and white women. Although maternal characteristics such as higher education, unmarried status, nulligravidity, and current smoking status were positively associated with adherence among low-income white women, few factors predicted adherence among low-income black women.

Demographic characteristics were important among white women. For example, education beyond high school was positively associated with adherence, and being married and having 1 or more previous pregnancies were negatively associated with adherence. Smoking was a positive predictor of adherence. Similarly, in a previous study based on the National Maternal and Infant Health Survey (NMIHS) of a representative group of pregnant women in the United States, higher education and smoking were positively associated with adherence to supplement use (26). Better knowledge of the benefits of supplements in women with higher education and increased concern about pregnancy outcome in smokers may explain these findings. In contrast to our study, being married was positively associated with adherence and higher gravidity was not associated with adherence (26). In our study, married women made up the minority, whereas among a nationally representative sample of U.S. pregnant women in 1988 (26) married women constituted the majority. Marriage may be a proxy for other factors that influence adherence to supplement use, such as social support, and these factors may vary over time and by income. Within our low-income sample, decreased adherence among white women with 1 or more previous pregnancies is of concern, because low-income and parous women are at higher risk of iron deficiency (43).

Few maternal characteristics were predictive of adherence among low-income black women. Other than gestational age at the time of entry into prenatal care, supplement use 3 mo prior to pregnancy and loss of appetite were the only 2 factors significantly associated with adherence among black women. It is possible that health behavior and belief variables are more important than sociodemographic variables in predicting adherence in this group of black women. It is also likely that low-income black women differ from low-income white women in ways that our study variables did not capture, which might explain the lack of other important predictors or the fact that low-income black women were more homogenous as a group. For example, additional subanalysis indicated that there was more variation in education beyond high school among white women compared to black women in our study (data not shown). Health beliefs related to pregnancy, anemia, and iron supplementation may better explain adherence behavior of black women and help identify groups at high risk of nonadherence. For example, qualitative studies in developing countries explored the barriers to supplement adherence through focus groups and in-depth interviews (27,28). Several perceptions related to iron supplements were reported as potential barriers to taking them, including fear of gaining too much weight, fear of difficult labor due to large babies, fear of developing too much blood, and fear of causing deformities in babies. Cost of supplements was sometimes a barrier, as well as poor shelf-life of supplements that are distributed free of charge in some countries (27,28). Further investigation focusing on such beliefs, interactions between health providers and pregnant women, and other factors such as social support using qualitative methods is necessary to understand the difference in adherence levels between black and white women and determinants of adherence among black women.

Adherence in our study, as determined by a pill count measure, was slightly greater than that reported by low-income women participating in NHANES (24), but was slightly less than that reported by low-income women attending health centers in Massachusetts before 1990 (25). Differences in measures and definitions of adherence as well as sample selection may partly explain the observed difference. The disparity in adherence to supplement use by race, as indicated in our study, was similar to the NHANES analysis for pregnant women of all income groups (24), the NMIHS analysis (26), and a previous study among low-income pregnant women (25). Although the difference in adherence between non-Hispanic black and non-Hispanic white women may not appear to be large at 7%, it is of concern due to the higher risk of iron deficiency and anemia among black women (43). Interestingly, supplements were free in this study, suggesting that racial disparities in adherence remain when cost is not a factor. Gestational age at the time of enrollment in our study was the only studied factor that was significantly associated with adherence to supplement use in both groups, and the gestational age at enrollment did not differ between the black and white women. Our study, however, was limited to women who enrolled in prenatal care before 20 wk gestation: thus it is possible that timing of entry and access to prenatal care may be a factor in the ethnic disparity in adherence in other populations.

The positive association between gestational age at the time of entry into the study and supplement adherence may reflect the difficulty of taking supplements during the first trimester because of a higher incidence of pregnancy-induced nausea and/or vomiting. We observed that women entering prenatal care in the first trimester reported more side effects from supplements than women who entered care in the second trimester (data not shown). However, only loss of appetite was significantly associated with adherence among our non-Hispanic black participants, and all other side effects had no association with adherence. This finding is similar to other studies, in which side effects explained only a small part of poor adherence (20,27,28,35,44). Interestingly, none of the women in our study attributed loss of appetite to use of supplements. It is possible that certain symptoms during pregnancy may reduce adherence whether or not they are perceived to be caused by supplements. Another possible explanation for the association between gestational age at entry to prenatal care and adherence is that women entering care late may be considered high-risk patients and may receive more intense counseling from providers, which encourages their adherence to prenatal supplements. Although all women in our study entered prenatal care before 20 wk, we did not have data on counseling received by the women to evaluate this scenario. However, previous data demonstrate similar results. Pregnant women in their second and third trimesters who participated in NHANES also reported higher adherence to supplements than women in their first trimester (24).

Our assessment of supplement adherence using pill counts was more accurate than self-reports, which have been used in other studies (24–26). However, compared to the gold standard MEMS measure, both self-reports and pill counts overestimated adherence in our validation sample (data not shown), a finding that other researchers have reported as well (45,46). Women may have saved their unused pills for later use or tossed them before returning the bottle due to social desirability or other reasons. Furthermore, we only used 1 measure of adherence during the entire length of pregnancy. Among women who returned for a refill more than once and provided multiple observations, variability across the measures was found to be low. However, women who did not return for refills more than once may be different from women who provided multiple measures. Hence our estimates of adherence may overestimate or underestimate actual adherence during the entire pregnancy. The large proportion of missing data on pill counts was another problem in our study. Although it is not possible to rule out selection bias due to missing data in our results, the similarity in results from complete case analysis and multiple imputation, as well as Heckman selectivity analysis, suggests that this potential bias was probably minimal. Finally, our findings may not be generalizable to women in other prenatal clinics who do not have free access to supplements or to women in ethnic groups other than those studied, such as Hispanic women or women who enter prenatal care after 20 wk gestation.

Our study indicates that in good circumstances, i.e., enrollment in prenatal care within the first half of pregnancy and free access to supplements, the majority of low-income women take supplements on a regular basis (as defined by supplement intake 65% of the time); only about 1 in 4 women do not. Adherence may be lower, however, among women who were lost to follow-up or those who forgot to return pill bottles. More concerning is the disparity in adherence by ethnicity among low-income women, because black women are at greater risk of iron deficiency and anemia than white women (43). Further examination of cultural factors related to pregnancy and supplementation is required to understand differences in adherence levels between ethnic groups and to identify risk factors for nonadherence among black women.

	ACKNOWLEDGMENTS

 
The authors thank Chirayath Suchindran and Kim Chantala for statistical assistance.

	FOOTNOTES

 
¹ Supported by a Cooperative Agreement with ASPH/CDC (No. S454).

³ Abbreviations used: LMP, last menstrual period; MEMS, Medication Event Monitoring System; NHANES III, National Health and Nutrition Examination Survey; NMIHS, National Maternal Infant Health Survey.

Manuscript received 14 November 2004. Initial review completed 7 December 2004. Revision accepted 6 February 2005.

	LITERATURE CITED

TOP ABSTRACT METHODS RESULTS DISCUSSION LITERATURE CITED

 

1. U.S. Department of Health and Human Services (2000) With understanding and improving health and objectives for improving health. Healthy People 2010 2nd ed. 2000:19-37-19-39 U.S. Government Printing Office Washington, DC.

2. Scholl, T. O., Hediger, M. L., Fischer, R. L. & Shearer, J. W. (1992) Anemia vs iron deficiency: Increased risk of preterm delivery in a prospective study. Am. J. Clin. Nutr. 55:985-988.[Abstract/Free Full Text]

3. Garn, S. M., Ridella, S. A., Petzold, A. S. & Falkner, F. (1981) Maternal hematologic levels and pregnancy outcomes. Semin. Perinatol. 5:155-162.[Medline]

4. Siega-Riz, A. M., Adair, L. S. & Hobel, C. J. (1998) Maternal hematologic changes during pregnancy and the effect of iron status on preterm delivery in a West Los Angeles population. Am. J. Perinatol. 15:515-522.[Medline]

5. World Health Organization (1992) The prevalence of anaemia in women: A tabulation of available information 1992 World Health Organization Geneva, Switzerland.

6. Beard, J. L. (1998) Iron deficiency in pregnancy. Perinat. Nutr. Rep. 5:1-4.

7. Academy of Science (1968) Iron deficiency in the United States. J. Am. Med. Assoc. 203:407-412.[Medline]

8. Centers for Disease Control and Prevention (1998) Pregnancy nutrition surveillance, 1996 full report 1998 Centers for Disease Control and Prevention, U.S. Department of Health and Human Services Atlanta, GA.

9. Agarwal, K. N., Agarwal, D. K. & Mishra, K. P. (1991) Impact of anaemia prophylaxis in pregnancy on maternal haemoglobin, serum ferritin & birth weight. Indian J. Med. Res. 94:277-280.[Medline]

10. Svanberg, B., Arvidsson, B., Norrby, A., Rybo, G. & Solvell, L. (1975) Absorption of supplemental iron during pregnancy—a longitudinal study with repeated bone-marrow studies and absorption measurements. Acta Obstet. Gynecol. Scand. (suppl.) 48:87-108.

11. Taylor, D. J., Mallen, C., McDougall, N. & Lind, T. (1982) Effect of iron supplementation on serum ferritin levels during and after pregnancy. Br. J. Obstet. Gynaecol. 89:1011-1017.[Medline]

12. Fleming, A. F., Martin, J. D., Hahnel, R. & Westlake, A. J. (1974) Effects of iron and folic acid antenatal supplements on maternal haematology and fetal wellbeing. Med. J. Aust. 2:429-436.[Medline]

13. Milman, N., Agger, A. O. & Nielsen, O. J. (1991) Iron supplementation during pregnancy. Effect on iron status markers, serum erythropoietin and human placental lactogen. A placebo controlled study in 207 Danish women. Dan. Med. Bull. 38:471-476.[Medline]

14. Milman, N., Agger, A. O. & Nielsen, O. J. (1994) Iron status markers and serum erythropoietin in 120 mothers and newborn infants. Effect of iron supplementation in normal pregnancy. Acta. Obstet. Gynecol. Scand. 73:200-204.[Medline]

15. Milman, N., Byg, K. E. & Agger, A. O. (2000) Hemoglobin and erythrocyte indices during normal pregnancy and postpartum in 206 women with and without iron supplementation. Acta. Obstet. Gynecol. Scand. 79:89-98.[Medline]

16. Vijayaraghavan, K., Brahmam, G. N., Nair, K. M., Akbar, D. & Rao, N. P. (1990) Evaluation of national nutritional anemia prophylaxis programme. Indian J. Pediatr. 57:183-190.[Medline]

17. Sloan, N. L., Jordan, E. A. & Winikoff, B. (1992) Does iron supplementation make a difference? 1992 John Snow, Inc MotherCare Project, Arlington, VA.

18. Schultink, W., van der Ree, M., Matulessi, P. & Gross, R. (1993) Low compliance with an iron-supplementation program: A study among pregnant women in Jakarta, Indonesia. Am. J. Clin. Nutr. 57:135-139.[Abstract/Free Full Text]

19. Galloway, R. & McGuire, J. (1994) Determinants of compliance with iron supplementation: supplies, side effects, or psychology?. Soc. Sci. Med. 39:381-390.

20. Ekstrom, E. C., Kavishe, F. P., Habicht, J. P., Frongillo, E. A., Jr, Rasmussen, K. M. & Hemed, L. (1996) Adherence to iron supplementation during pregnancy in Tanzania: determinants and hematologic consequences. Am. J. Clin. Nutr. 64:368-374.[Abstract/Free Full Text]

21. Bondarianzadeh, D., Siassi, F., Omidvar, N., Golestan, B. & Keighobadi, K. (1998) Low compliance with iron supplementation program among pregnant women in the rural areas of Kerman district, I.R. Iran. Nutr. Res. 18:945-952.

22. Norell, S. E. (1981) Accuracy of patient interviews and estimates by clinical staff in determining medication compliance. Soc. Sci. Med. [E] 15:57-61.

23. Bonnar, J., Goldberg, A. & Smith, J. A. (1969) Do pregnant women take their iron?. Lancet 1:457-458.[Medline]

24. Cogswell, M. E., Kettel-Khan, L. & Ramakrishnan, U. (2003) Iron supplement use among women in the United States: Science, policy and practice. J. Nutr. 133:1974S-1977S.[Abstract/Free Full Text]

25. Suitor, C. W. & Gardner, J. D. (1990) Supplement use among a culturally diverse group of low-income pregnant women. J. Am. Diet. Assoc. 90:268-271.[Medline]

26. Yu, S. M., Keppel, K. G., Singh, G. K. & Kessel, W. (1996) Preconceptional and prenatal multivitamin-mineral supplement use in the 1988 National Maternal and Infant Health Survey. Am. J. Public Health 86:240-242.[Abstract/Free Full Text]

27. Galloway, R., Dusch, E., Elder, L., Achadi, E., Grajeda, R., Hurtado, E., Favin, M., Kanani, S., Marsaban, J., Meda, N., Moore, K. M., Morison, L., Raina, N., Rajaratnam, J., Rodriquez, J. & Stephen, C. (2002) Women’s perceptions of iron deficiency and anemia prevention and control in eight developing countries. Soc. Sci. Med. 55:529-544.

28. Ghanekar, J., Kanani, S. & Patel, S. (2002) Toward better compliance with iron-folic acid supplements: Understanding the behavior of poor urban pregnant women through ethnographic decision models in Vadodara, India. Food Nutr. Bull. 23:65-72.[Medline]

29. Viteri, F. E., Ali, F. & Tujague, J. (1999) Long-term weekly iron supplementation improves and sustains nonpregnant women’s iron status as well or better than currently recommended short-term daily supplementation. J. Nutr. 129:2013-2020.[Abstract/Free Full Text]

30. Simmons, W. K., Cook, J. D., Bingham, K. C., Thomas, M., Jackson, J., Jackson, M., Ahluwalia, N., Kahn, S. G. & Patterson, A. W. (1993) Evaluation of a gastric delivery system for iron supplementation in pregnancy. Am. J. Clin. Nutr. 58:622-626.[Abstract/Free Full Text]

31. Mumtaz, Z., Shahab, S., Butt, N., Rab, M. A. & DeMuynck, A. (2000) Daily iron supplementation is more effective than twice weekly iron supplementation in pregnant women in Pakistan in a randomized double-blind clinical trial. J. Nutr. 130:2697-2702.[Abstract/Free Full Text]

32. McKenna, D., Spence, D., Haggan, S. E., McCrum, E., Dornan, J. C. & Lappin, T. R. (2003) A randomized trial investigating an iron-rich natural mineral water as a prophylaxis against iron deficiency in pregnancy. Clin. Lab. Haematol. 25:99-103.[Medline]

33. Brock, C., Curry, H., Hanna, C., Knipfer, M. & Taylor, L. (1985) Adverse effects of iron supplementation: A comparative trial of a wax-matrix iron preparation and conventional ferrous sulfate tablets. Clin. Ther. 7:568-573.[Medline]

34. Frykman, E., Bystrom, M., Jansson, U., Edberg, A. & Hansen, T. (1994) Side effects of iron supplements in blood donors: superior tolerance of heme iron. J. Lab. Clin. Med. 123:561-564.[Medline]

35. Hyder, S. M., Persson, L. A., Chowdhury, A. M. & Ekstrom, E. C. (2002) Do side-effects reduce compliance to iron supplementation? A study of daily- and weekly-dose regimens in pregnancy. J. Health Popul. Nutr. 20:175-179.[Medline]

36. Institute of Medicine, Food and Nutrition Board (1993) Iron deficiency anemia: guidelines for prevention, detection and management among U.S. children and women of childbearing age 1993 National Academy Press Washington, DC .

37. USP DI Drug Information for the Health Care Professional 2003 Thomson Micromedex Greenwood Village, CO.

38. Heckman, J. J. (1979) Sample selection bias as a specification error. Econometrica 47:153-161.

39. Rubin, D. B. (1987) Multiple Imputation for Nonresponse in Surveys 1987 Wiley New York, NY.

40. Schafer, J. L. (1997) Analysis of Incomplete Multivariate Data 1997 Chapman & Hall London, UK.

41. Horton, J. N. & Lipsitz, S. R. (2001) Multiple imputation in practice: Comparison of software packages for regression models with missing variables. Am. Stat. 55:244-254.

42. SAS Institute (2000) SAS STAT 2000 SAS Institute Cary, NC.

43. Looker, A. C., Dallman, P. R., Carroll, M. D., Gunter, E. W. & Johnson, C. L. (1997) Prevalence of iron deficiency in the United States. J. Am. Med. Assoc. 277:973-976.[Abstract]

44. Winichagoon, P. (2002) Prevention and control of anemia: Thailand experiences. J. Nutr. 132:862S-866S.[Abstract/Free Full Text]

45. Matsui, D., Hermann, C., Klein, J., Berkovitch, M., Olivieri, N. & Koren, G. (1994) Critical comparison of novel and existing methods of compliance assessment during a clinical trial of an oral iron chelator. J. Clin. Pharmacol. 34:944-949.[Abstract]

46. Cromer, B. A., Steinberg, K., Gardner, L., Thornton, D. & Shannon, B. (1989) Psychosocial determinants of compliance in adolescents with iron deficiency. Am. J. Dis. Child 143:55-58.[Abstract]

This article has been cited by other articles:

				S. Jasti, A. M. Siega-Riz, M. E. Cogswell, and A. G. Hartzema Correction for Errors in Measuring Adherence to Prenatal Multivitamin/Mineral Supplement Use among Low-Income Women J. Nutr., February 1, 2006; 136(2): 479 - 483. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jasti, S. Articles by Bentley, M. E. Search for Related Content PubMed PubMed Citation Articles by Jasti, S. Articles by Bentley, M. E.