![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Department of International Health, Rollins School of Public Health, Emory University, Atlanta, GA 30322;
*
Division of Nutritional Sciences, Cornell University, Ithaca, NY 14853; and
National Center for Chronic Disease Prevention and Physical Activity, Centers for Disease Control and Prevention, Atlanta, GA
3To whom correspondence and reprint requests should be addressed. E-mail: uramakr{at}sph.emory.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • iron deficiency • dietary supplements • anemia • diet assessment • women
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Inadequate intake combined with poor availability and increased loss due to infections such as helminthes and malaria have been described as the leading causes of this nutrient deficiency. Young children and women of reproductive age, especially pregnant women, are most vulnerable and at increased risk of iron deficiency because of their increased demands and/or losses (2
). Recent work has shown that among women of reproductive age in the United States, subpopulations such as Mexican Americans (MA)4
are two to three times more likely to be iron deficient than are non-Hispanic whites (NHW) (3
,4
). The most recent national estimates, which used data from the third National Health and Nutrition Examination Survey, 1988–1994 (NHANES III) and the multi-indicator approach (i.e., abnormal values for any two of three iron status indicators, namely, serum ferritin, transferrin saturation and free erythrocyte protoporphyrin) for the prevalence of iron deficiency were 19% among MA, 10% among NHW, and 15% among non-Hispanic Black girls and women aged 12–49 y (5
). The reasons for these differences are unclear. Previous work by our group found that the increased risk of iron deficiency anemia among MA women compared with NHW women could be described in part by lower socioeconomic status and lower iron intake from supplements (4
). The intake of dietary iron based on the 24-h recalls was similar in both groups, which raises concerns of dietary quality and bioavailability of iron in the diet. In particular, bioavailability studies show that heme iron (from animal foods) is absorbed more efficiently than nonheme iron (from plant foods). In addition, the amount of iron absorbed is affected by the presence of enhancers such as vitamin C and inhibitors such as phytates and tannins (6
–9
). Few studies have examined the importance of these dietary factors in explaining iron status at a population level (10
). We examined the relationship between dietary patterns related to iron intake and bioavailability with low iron stores among MA and NHW women of reproductive age with the use of nationally representative data from NHANES III. Low serum ferritin was used as the main outcome because it represents iron depletion, the earliest stage of iron deficiency, and because it is directly correlated with iron stores. |
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Participants in NHANES III represent the civilian noninstitutionalized population 
2 mo of age in the United States. NHANES III was conducted by the National Center for Health Statistics; the procedures for data collection and analysis have been published (11
). A stratified multistage probability design was used to select participants, and data were collected via household interviews and physical examinations in mobile examination centers. Ethical approval was obtained and written consent was received from all study participants (11
).
Our study sample was restricted to MA and NHW nonpregnant girls and women 12–39 y of age (n = 3138). We included adolescent girls (12–19 y) because they are also at risk for iron deficiency. Subjects were excluded if they had missing data on serum ferritin, dietary intake, supplement use, ethnicity and age or had received treatment for anemia in the past 3 mo (n = 59). NHANES III participants lacking serum ferritin values did not differ significantly from our final sample with regard to age, ethnicity, income or education. Participants who had missing data on any of the selected variables of interest or who were excluded because of potential modifiers of serum ferritin also did not differ significantly in age, income, education, parity, oral contraceptive use, daily iron intake or supplement use. The final sample consisted of 1368 MA and 1473 NHW adolescent girls and women of childbearing age.
Serum ferritin level is influenced by causes other than iron deficiency (12
). To control for potential biases that may result from misclassification of subjects as iron replete (i.e., not deficient), the analysis was repeated on a subsample that did not include women (n = 484) who had signs of infection as indicated by elevated C-reactive protein (>6 mg/L) or abnormal white blood cell count (>11.0 or <3.5 x 109/L); possible liver disease as determined by abnormal levels of alanine aminotransferase (>74 µmol/L or >1.23 µkat/L), aspartate aminotransferase (>68 µmol/L or 1.13 µkat/L) or alkaline phosphatase (>154.5 µmol/L or >2.58 µkat/L); or serum ferritin level 20 µg/L with a low hemoglobin concentration (13
).
We defined low iron stores as serum ferritin <12 µg/L (14
,15
), measured using the Bio-Rad Laboratories Quantimune Ferritin IRMA kit (Hercules, CA) (16
).
Data collected using a qualitative food-frequency questionnaire (FFQ) for the consumption of a variety of foods and food groups in the past month were used to classify individuals as low, medium or high with respect to intake of heme iron, nonheme iron, enhancers of iron absorption and inhibitors of iron absorption. For each food group, low consumption was <25th percentile, medium consumption >25th to 75th percentile and high consumption >75th percentile; these percentiles were based on values from NHW women.
The FFQ used in NHANES III was designed to provide typical or qualitative data for ranking persons by intake of specific foods and food groups and not to produce population nutrient estimates. The food list contained 17 categories of specific foods or food groups; it was developed to be comparable to food lists used in past NHANES and expanded to capture detailed intake of foods containing specific nutrients such as vitamins A and C (11
). The period of recall was the past month and information on portion sizes was not collected. The FFQ was incorporated into the household interview and was administered by trained interviewers either in English or Spanish based on the respondent’s language preference. The instrument was pretested and modified to be culturally appropriate especially for use in population subgroups, namely, non-Hispanic White, Black and Mexican Americans using information from previous NHANES (NHANES II and HHANES).
For intake of heme iron, only foods of animal origin (i.e., dietary heme iron) were included. For nonheme iron, two main sources, namely, plant foods and iron-containing supplements were considered. Only plant foods with >0.35 mg iron/serving as determined by the method of Pennington and colleagues (17
) were included in the nonheme group, because plant foods generally contain <0.10 mg iron (a negligible amount) or >0.35 mg iron/serving. Similarly, enhancers of iron absorption included dietary heme iron, dietary vitamin C and vitamin C supplements. For similar reasons as for plant foods, only foods containing >24 mg vitamin C/serving (17
) were counted toward dietary vitamin C; all foods containing any heme iron were included as enhancers. Foods rich in phytate or tannin comprised the iron absorption inhibitors and were identified on the basis of published values (18
).
We converted the individual food frequencies to daily intakes by multiplying the number of servings by an ethnic-specific weighting value. The weights were based on the average serving size and the content of iron (for heme and nonheme iron), vitamin C (for enhancers), and phytates and tannins (for inhibitors) derived from the 24-h recall data that was collected by an automated dietary interview (11
). Weighting factors were used because the individual food sources, and therefore the amounts of dietary iron, vitamin C and inhibitors, were different for MA and NHW. For mixed dishes, we used absolute amounts of iron and vitamin C and used all foods listed in the FFQ, regardless of preparation.
For each subject, we then took the weighted serving average of all foods containing heme iron and summed them to obtain that subject’s total dietary intake of heme iron. This step was performed to determine total dietary intakes of nonheme iron and vitamin C as well. For the entire study cohort, heme and nonheme iron values were split three ways, i.e., low, medium or high. Enhancer values were split three ways according to the combination of heme iron and vitamin C consumption: low consumption of both heme iron and vitamin C = low enhancers; low of one and medium of the other = low enhancers; low plus high = medium enhancers; medium plus medium = medium enhancers; medium plus high = high enhancers; and high plus high = high enhancers.
Intake of supplements containing iron or vitamin C was determined from a series of questions about the frequency, dosage and type of supplement(s) that the subject had taken within the last month. Because of small sample sizes, we treated supplemental iron and vitamin C intakes as dichotomous variables (0 = No, 1 = Yes). We also incorporated the average daily intake of supplemental iron and vitamin C into the estimates of total daily intake of nonheme iron and absorption enhancers, respectively, as described earlier.
We included the sociodemographic variables of ethnicity, income and education. Ethnicity was based on self-reported data. Poverty income ratio (PIR) is the total household income divided by the poverty threshold for the year of the interview (19
), which is determined annually by the U.S. Bureau of the Census, taking into account geographic location, rate of inflation and family size (20
). Head of the household’s education was used for girls 12–18 y, and the last year of school completed was used for women 19–39 y.
Physiologic variables such as body mass index (BMI; kg/m2), parity, and oral contraceptive use were also included. BMI 95th percentile for age was used to categorize women aged 12–17 y as obese (21
), and BMI 30 was used to categorize women aged 18–39 y as obese (22
). Parity was classified as 0, 1 or 2 live births. Oral contraceptive use was self-reported and was coded as current user or nonuser.
Other nondietary factors were considered, but we did not include them in the final model because we found that they were not associated with either ethnicity or iron status. These factors were subject’s country of origin, origin of household’s head, language of interview, food insecurity, smoking, alcohol consumption, currently trying to lose weight, time of venipuncture, fasting status before blood sampling, menstruation within the last 6 mo, level of C-reactive protein and serum lead concentration.
Stratified analysis by ethnicity examined the age-adjusted relationship between various dietary factors and low serum ferritin. Multivariate logistic regression analysis was used to determine the association between ethnicity and low iron stores while adjusting for the effects of dietary and sociodemographic covariates. Factors that were not significant were excluded from the model, such as obesity and oral contraceptive use. The final models included income and parity. Two-way interactions between ethnicity and diet were also tested. All statistical analyses were done using SUDAAN (version 7.5; Research Triangle Institute, Research Triangle Park, NC) to account for the complex sample design, and statistical significance was set at P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). They were also less likely to use oral contraceptives and have more children.
|
). Even with the lower intake of supplements among MA, total intakes of nonheme iron and enhancers were higher for MA than for NHW. Despite the differences in patterns of intake based on the FFQ, total iron intake based on the 24-h recall was similar (13 mg/d) in both ethnic groups, as was the average contribution of heme iron to total iron intake based on the 24-h recall, which was 12.7% for MA and 15.5% for NHW.
|
). None of the other dietary factors were associated independently with low iron stores among MA. Among NHW, consumption of heme iron, nonheme iron and nonheme iron plus iron supplement were significantly associated with low iron stores: the prevalance of low iron stores was greatest among those with medium intake.
|
The results of multivariate analysis indicated that the intakes of dietary heme iron and of dietary nonheme iron plus iron supplement were significantly associated with low iron stores (Table 4
). Women who consumed vitamin C supplements were half as likely to have low iron stores even after adjusting for age, ethnicity and patterns of intake. This association was attenuated and was no longer statistically significant after controlling for income and parity. Oral contraceptive use and overweight were not included in the final models because they were not significantly associated with low iron stores after adjusting for other factors. Compared with women with high heme iron intake, those with medium heme iron intake were more likely to have low iron stores; the opposite was seen for nonheme iron intake. The OR for MA women having low iron stores compared with NHW women, however, did not change and remained statistically significant even after adjustment for dietary factors.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Earlier work suggested that income is a key explanatory factor for differences in iron status (4
) and in our analysis, adjusting for other reproductive health related factors (such as parity and oral contraceptive use, which are known to be associated with iron status in women of reproductive age), failed to reduce the excess risk of low iron stores. Differences in blood loss, which were not measured in NHANES III, may provide an alternate explanation. For example, Harlow et al. (23
) found that American girls of European descent were less likely (OR = 0.48) to have an episode of heavy bleeding compared with African-American girls. MA women may have greater menstrual blood loss, parasitic infections that lead to blood loss and other infections that increase iron loss, and thereby reduce iron stores.
A limitation of our analyses is the lack of data on meal-specific consumption of iron intakes, enhancers and inhibitors that may influence iron bioavailability. We used information from the FFQ because it provides a good estimate of usual long-term intake (24
) and assumed portion sizes based on the single 24-h dietary recall to estimate usual iron, inhibitor and enhancer intakes. However, we were unable to account for the timing of consumption of iron enhancers and inhibitors, whose effects on iron absorption may be decreased if they are consumed separately from the iron-containing meal. A recent study showed that the composition of a meal influences iron absorption, and that the effects of inhibitors and enhancers of iron absorption are additive across meals (25
). Another concern is that our ability to detect effects may have been limited by the fact that the FFQ in NHANES III was not designed specifically to estimate iron intakes, but rather to be representative of overall dietary patterns and therefore useful to examine the relationship between patterns of intake and iron status. Although the FFQ is a reasonable measure of habitual intakes and ranks individuals appropriately (26
), it tends to reflect recent intakes and does not take into account patterns of intake during previous critical periods in the life cycle. Differences in iron status may begin earlier than age 12 y and MA women may have had inadequate dietary intakes and poorer iron status compared with NHW women from early childhood through adolescence, thereby increasing their risk of poor status throughout their reproductive years.
In conclusion, the unexplained high prevalence of iron deficiency among MA women indicates the need to devise well-designed prospective and intervention studies to help identify strategies to improve their iron status. Although the prevalence of iron deficiency is low in the United States compared with other parts of the world, at least one of every six MA women of reproductive age remains at risk of iron deficiency (1
,3
) and its associated consequences such as reduced productivity and adverse reproductive outcomes (27
,28
). This warrants appropriate public health action, and interventions that promote supplement consumption and/or reduce iron losses require further investigation.
|
FOOTNOTES |
|---|
2 Supported by grants from the National Institutes of Health, HD-34531 and the ASPH/CDC/ATSDR internship program.
4 Abbreviations used: CI, confidence interval; FFQ, food-frequency questionnaire; MA, Mexican American; NHANES, National Health and Nutrition Examination Survey; NHW, non-Hispanic white; OR, odds ratio; PIR, poverty income ratio.
Manuscript received 16 July 2001. Initial review completed 27 September 2001. Revision accepted 11 February 2002.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. ACC/SCN (2000) Fourth Report on the World Nutrition Situation 2000 ACC/SCN, WHO (Geneva) in collaboration with IFPRI Washington, DC. .
2. Allen, L. & Casterline-Sabel, J. E. (2001) Prevalence and causes of nutritional anemias. Ramakrishnan, U. eds. Nutritional Anemias 2001 CRC Press Boca Raton, FL. .
3.
Looker, A. C., Johnson, C. L., McDowell, M. A. & Yetley, E. A. (1989) Iron status: prevalence of impairment in three Hispanic groups in the United States. Am. J. Clin. Nutr. 49:553-558.
4.
Frith-Terhune, A. L., Cogswell, M. E., Kettel Khan, L., Will, J. C. & Ramakrishnan, U. (2000) Iron deficiency anemia: higher prevalence in Mexican American than in non-Hispanic white females in the third National Health and Nutrition Examination Survey, 1988–1994. Am. J. Clin. Nutr. 72:963-968.
5. U.S. Department of Health and Human Services (2000) Healthy People 2010. 2nd ed. "Understanding and Improving Health" and "Objectives for Improving Health," 1 and 2 U.S. Government Printing Office Washington, DC .
6.
Reddy, M. B., Hurrell, R. F. & Cook, J. D. (2000) Estimation of nonheme-iron bioavailability from meal composition. Am. J. Clin. Nutr. 71:937-943.
7. Hulthen, L., Gramartkovski, E., Gleerup, A. & Hallberg, L. (1995) Iron absorption from the whole diet. Relation to meal composition iron requirements and iron stores. Eur. J. Clin. Nutr. 49:794-808.[Medline]
8.
Monsen, E. R., Hallberg, L., Layrisse, M., Hegsted, D. M., Cook, J. D., Mertz, W. & Finch, C. A. (1978) Estimation of available dietary iron. Am. J. Clin. Nutr. 31:134-141.
9. Monsen, E. R. & Balintfy, J. L. (1982) Calculating dietary iron bioavailability, refinement and computerization. J. Am. Diet. Assoc. 80:3073-3011.
10. Murphy, S. P., Beaton, G. H. & Calloway, D. H. (1992) Estimated mineral intakes of toddlers: predicted prevalence of inadequacy in village populations in Egypt, Kenya and Mexico. Am. J. Clin. Nutr. 56:5625-5672.
11. National Center for Health Statistics (1994) Plan and operation of the Third National Health and Nutrition Examination Survey, 1988–94. Vital Health Stat. 1 1994:no. 32 Hyattsville, MD. .
12.
Yip, R. & Dallman, P. R. (1988) The roles of inflammation and iron deficiency as causes of anemia. Am. J. Clin. Nutr. 48:1295-1300.
13. Fleming, D. J., Jacques, P. F., Dallal, G. E., Tucker, K. L., Wilson, P. W. F. & Wood, R. J. (1998) Dietary determinants of iron stores in a free-living elderly population: the Framingham Heart Study. Am. J. Clin. Nutr. 67:722-733.[Abstract]
14.
Cook, J. D., Finch, C. A. & Smith, N. J. (1976) Evaluation of the iron status of a population. Blood 48:449-455.
15. Centers for Disease Control and Prevention (1998) Recommendations to prevent and control iron deficiency in the United States. Morb. Mortal. Wkly. Rep. 47:1-29.[Medline]
16. Gunter, E., Lewis, B. & Koncikowski, S. (1996) Laboratory Procedures Used for the Third National Health and Nutrition Examination Survey (NHANES III) 1988–1994 1996 Centers for Disease Control and Prevention Hyattsville, MD .
17. Pennington, J.A.T., Bowes, A.D.P. & Church, H. N. (1997) Bowes & Church’s Food Values of Portions Commonly Used 17th ed. 1997 Lippincott Williams & Wilkins Baltimore, MD. .
18. Harland, B. F. & Oberleas, D. (1987) Phytate in foods. World Rev. Nutr. Diet. 52:235-259.[Medline]
19. U.S. Department of Agriculture (1987) Food and Nutrition Service Financial Management and Program Information Division, Annual Historical Review of FNS Programs: Fiscal Year 1987 1987 U.S. Department of Agriculture Washington, DC .
20. U.S. Bureau of the Census (1991) Poverty in the United States, 1990 1991 U.S. Bureau of the Census Washington, DC. Current Population Reports, p. 175.
21. Rosner, B., Prineas, R., Loggi, J. & Daniels, S. (1998) Percentiles for body mass index in U.S. children 5 to 17 years of age. J. Pediatr. 132:211-222.[Medline]
22. National Heart, Lung, and Blood Institute (1998) Clinical Guidelines on the Identification, Evaluation and Treatment of Overweight and Obesity in Adults: The Evidence Report 1998 National Institutes of Health National Heart, Lung, and Blood Institute, Bethesda, MD. .
23.
Harlow, S. D., Crawford, S. L., Sommer, B. & Greendale, G. A. (1996) Ethnic differences in the duration and amount of menstrual bleeding during the post-menarcheal period. Am. J. Epidemiol. 144:980-988.
24. Willet, W. (1990) Nutritional Epidemiology 1990:252 Oxford University Press New York, NY. .
25.
Hallberg, L. & Hulthen, L. (2000) Prediction of dietary iron absorption: an algorithm for calculating absorption and bioavailability of dietary iron. Am. J. Clin. Nutr. 71:1147-1160.
26.
Willet, W. C., Sampson, L., Stampfer, M. J., Rosner, B., Bain, C, Witschi, J., Henneken, C. H. & Speizer, F. E. (1985) Reproducibility and validity of a semi-quantitative food frequency questionnaire. Am. J. Epidemiol. 122:51-65.
27. Beard, J. L. (2001) Functional consequences of nutritional anemia in adults. Ramakrishnan, U. eds. Nutritional Anemias 2001 CRC Press Boca Raton, FL. .
28. Ramakrishnan, U. (2001) Functional consequences of nutritional anemia during pregnancy and early childhood. Ramakrishnan, U. eds. Nutritional Anemias 2001 CRC Press Boca Raton, FL. .
This article has been cited by other articles:
|
|
 |
|
  S. Peneau, L. Dauchet, A.-C. Vergnaud, C. Estaquio, E. Kesse-Guyot, S. Bertrais, P. Latino-Martel, S. Hercberg, and P. Galan Relationship between iron status and dietary fruit and vegetables based on their vitamin C and fiber content Am. J. Clinical Nutrition, May 1, 2008; 87(5): 1298 - 1305. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. McClung, L. J. Marchitelli, K. E. Friedl, and A. J. Young Prevalence of Iron Deficiency and Iron Deficiency Anemia among Three Populations of Female Military Personnel in the US Army. J. Am. Coll. Nutr., February 1, 2006; 25(1): 64 - 69. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E Mazur, G. S Marquis, and H. H Jensen Diet and food insufficiency among Hispanic youths: acculturation and socioeconomic factors in the third National Health and Nutrition Examination Survey Am. J. Clinical Nutrition, December 1, 2003; 78(6): 1120 - 1127. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
