[show abstract][hide abstract] ABSTRACT: Results of previous studies on diet and gallbladder disease (GBD), defined as having gallstones or having had surgery for gallstones, have been inconsistent. This research examined patterns of food intake in Mexican Americans and their associations with GBD.
The study population included 4641 Mexican Americans aged 20-74 years who participated in the 1988-94 third National Health and Nutrition Examination Survey (NHANES III). GBD was diagnosed by ultrasound. Food intake patterns were identified by principal components analysis based on food frequency questionnaire responses. Component scores representing the level of intake of each pattern were categorized into quartiles, and prevalence odds ratios (POR) were estimated relative to the lowest quartile along with 95% confidence intervals (CI).
There were four distinct patterns in women (vegetable, high calorie, traditional, fruit) and three in men (vegetable, high calorie, traditional). After age adjustment, none were associated with GBD in women. However, men in the third (POR = 0.42, 95%CI 0.21-0.85) and fourth (POR = 0.53, 95%CI 0.28-1.01) quartiles of the traditional intake pattern were half as likely to have GBD as those in the lowest quartile.
These findings add to a growing literature suggesting dietary intake patterns can provide potentially useful and relevant information on diet-disease associations. Nevertheless, methods to do so require further development and validation.
Public Health Nutrition 07/2000; 3(2):233-43. · 2.25 Impact Factor
[show abstract][hide abstract] ABSTRACT: Objective Results of previous studies on diet and gallbladder disease (GBD), defined as having gallstones or having had surgery for gallstones, have been inconsistent. This research examined patterns of food intake in Mexican Americans and their associations with GBD.Design Cross-sectional.Subjects The study population included 4641 Mexican Americans aged 20–74 years who participated in the 1988–94 third National Health and Nutrition Examination Survey (NHANES III). GBD was diagnosed by ultrasound. Food intake patterns were identified by principal components analysis based on food frequency questionnaire responses. Component scores representing the level of intake of each pattern were categorized into quartiles, and prevalence odds ratios (POR) were estimated relative to the lowest quartile along with 95% confidence intervals (CI).Results There were four distinct patterns in women (vegetable, high calorie, traditional, fruit) and three in men (vegetable, high calorie, traditional). After age adjustment, none were associated with GBD in women. However, men in the third (POR = 0.42, 95%CI 0.21–0.85) and fourth (POR = 0.53, 95%CI 0.28–1.01) quartiles of the traditional intake pattern were half as likely to have GBD as those in the lowest quartile.Conclusions These findings add to a growing literature suggesting dietary intake patterns can provide potentially useful and relevant information on diet–disease associations. Nevertheless, methods to do so require further development and validation.
Public Health Nutrition 05/2000; 3(02):233 - 243. · 2.25 Impact Factor
[show abstract][hide abstract] ABSTRACT: Mean serum total cholesterol levels appear to be higher in the Federal Republic of Germany (FRG) than in the United States (US) while coronary heart disease death rates are lower. The study examined possible factors for the difference including possible differences in laboratory methodology. Cross-sectional data from the first two waves of the German National Health Surveys (1984-1986 and 1987-1989; n = 9709) and from the Second National Health and Nutrition Examination Survey (1976-1980; n = 7832) were compared for men and women 25-69 years of age. The influence of age, body mass index, diet, cigarette smoking, education, income, use of oral contraceptives or antihypertensive agents, alcohol consumption and potential differences in laboratory measurement were explored using multiple regression techniques separately for men and women for ages 25-39, 40-59 and 60-69 years of age. Overall ages, unadjusted mean total cholesterol levels were higher in German than US men (6.02 vs. 5.64 mmol/l) and in German than US women (6.04 vs. 5.80 mmol/l) as were HDL cholesterol levels (men: 1.30 vs 1.14 mmol/l; women: 1.65 vs. 1.38 mmol/l). Adjusting for lifestyle factors explained, on the average, 40% of the differences in mean total cholesterol of which half or 20% was accounted for by adjusting for alcohol intake. Adjusting for possible laboratory differences explained, on the average, an additional 30% of the differences. Frequency of alcohol intake was the most important factor in explaining differences in mean HDL cholesterol levels. Adjustment for differences in alcohol intake had negligible effects on reducing the differences in mean non-HDL cholesterol.
European Journal of Epidemiology 02/2000; 16(7):669-75. · 5.12 Impact Factor
[show abstract][hide abstract] ABSTRACT: Omega-3 fatty acids have potential antiatherogenic, antithrombotic, and antiarrhythmic properties, but their role in coronary heart disease remains controversial. To evaluate the association of omega-3 fatty acids in adipose tissue with the risk of myocardial infarction in men, a case-control study was conducted in eight European countries and Israel. Cases (n639) included patients with a first myocardial infarction admitted to coronary care units within 24 hours from the onset of symptoms. Controls (n700) were selected to represent the populations originating the cases. Adipose tissue levels of fatty acids were determined by capillary gas chromatography. The mean (SD) proportion of -linolenic acid was 0.77% (0.19) of fatty acids in cases and 0.80% (0.19) of fatty acids in controls (P0.01). The relative risk for the highest quintile of -linolenic acid compared with the lowest was 0.42 (95% confidence interval [CI] 0.22 to 0.81, P-trend0.02). After adjusting for classical risk factors, the relative risk for the highest quintile was 0.68 (95% CI 0.31 to 1.49, P-trend0.38). The mean proportion of docosahexaenoic acid was 0.24% (0.13) of fatty acids in cases and 0.25% (0.13) of fatty acids in controls (P0.14), with no evidence of association with risk of myocardial infarction. In this large case-control study we could not detect a protective effect of docosahexaenoic acid on the risk of myocardial infarction. The protective effect of -linolenic acid was attenuated after adjusting for classical risk factors (mainly smoking), but it deserves further research. (Arterioscler Thromb Vasc Biol. 1999;19:1111-1118.) Key Words: myocardial infarction -linolenic acid docosahexaenoic acid case-control studies adipose tissue.
[show abstract][hide abstract] ABSTRACT: Biomarkers of dietary exposure or nutritional status are sought actively to overcome limitations of traditional dietary methodology. We compared plasma and adipose tissue biomarkers for carotenoids and tocopherols. The data consisted of samples from 91 men and 122 women, ages 45-70 years, from the control group of the European Community Multicentre Study on Antioxidants, Myocardial Infarction, and Cancer of the Breast (EURAMIC) Study. Pearson correlations between plasma and adipose tissue measurements for beta-carotene, lycopene, and alpha-tocopherol adjusted for smoking status displayed low, although significant, correlations of 0.39, 0.24, and 0.39, respectively. The correlation was further stratified by sex. After being corrected for measurement error using deattenuation factors obtained from a reproducibility study, the stratified correlation coefficients were as high as 0.80 for beta-carotene in men, 0.62 for lycopene in women, and 0.52 for alpha-tocopherol in women. In addition, plasma and adipose tissue measurements from the myocardial infarction (MI) subset of the EURAMIC study population were used to evaluate the odds of MI, adjusting for confounders. We found that the concentration of lycopene in plasma was not positively associated significantly with MI (odds ratio, 1.78; P = 0.26). Adipose tissue lycopene, in contrast to reports elsewhere on the total population, showed an inverse association with MI (odds ratio, 0.62; P = 0.15). These results suggest that plasma and adipose carotenoids represent different markers for nutritional status and cannot be used interchangeably in epidemiological and dietary validation studies.
[show abstract][hide abstract] ABSTRACT: To examine the effect of measurement error in dietary data on the relationship between diet and body mass index (BMI). To correct for the effect of measurement error on diet-BMI association by using replicate measurements of diet. The effect of measurement error on diet--BMI relationship was simulated, and its implications are discussed.
Prospective study design.
The first and second China Health and Nutrition Survey conducted in 1989 and 1991, respectively.
Three thousand, four hundred and seventy-nine adults age 20-45 y at the 1989 survey.
Statistical methods were used to demonstrate the effect of measurement error in dietary data on the diet-BMI association.
By using the average of three replicate 24 h dietary recalls, the attenuation of diet-BMI association was reduced substantially. The regression coefficients of fat and energy intakes differed markedly from those computed by using only single measurement of diet.
Measurement error in dietary data may significantly attenuate the diet-disease association. Where appropriate, specific emphasis may be needed to address the problem of measurement error in the study of diet-disease relationship.
European Journal of Clinical Nutrition 11/1998; 52(10):722-7. · 2.76 Impact Factor
[show abstract][hide abstract] ABSTRACT: We examined the cross-sectional association between intake of carotenoids with provitamin A activity and carotid artery plaques in 12,773 participants of the Atherosclerosis Risk in Communities Study aged 45-64 y. Usual diet was assessed with a 66-item food-frequency questionnaire. Plaques were examined by B-mode ultrasound of multiple carotid artery segments. In both women and men, those in the highest quintile of carotenoid consumption had a lower prevalence of plaques (women, 25.4%; men, 36.0%) than those in the lowest quintile of carotenoid consumption (women, 29.3%; men, 39.8%). The prevalence odds ratios contrasting extreme intake quintiles were 0.82 (95% CI: 0.70, 0.97) in women and 0.85 (95% CI: 0.72, 1.01) in men. The associations diminished slightly after potential confounders were adjusted for. In women, the inverse association was particularly strong for current smokers (adjusted odds ratio contrasting extreme quintiles: 0.67; 95% CI: 0.45, 0.98). In men, no such effect modification by smoking was seen. The inverse association was somewhat stronger in men aged 55-64 y than in those aged 45-54 y, whereas age made little difference in women. These findings, together with previous findings that carotenoid intake was unrelated to average carotid artery wall thickness, suggest that carotenoids may exert their influence later rather than earlier in the atherosclerotic process, and support the hypothesis that carotenoids or other plant-derived compounds may play a role in preventing arterial plaque formation.
American Journal of Clinical Nutrition 10/1998; 68(3):726-33. · 6.50 Impact Factor
[show abstract][hide abstract] ABSTRACT: To determine the role of fruit and vegetable consumption and dietary intake of folic acid and related nutrients such as methionine, cysteine and alcohol in the aetiology of breast cancer.
Population based case-control study.
Part of the European Community Multicentre Study on Antioxidants, Myocardial Infarction, and Cancer of the Breast (EURAMIC) in Berlin, Germany.
As part of the EURAMIC study, dietary intake data were collected in 43 postmenopausal women diagnosed with breast cancer between 1991 and 1992 in Berlin, Germany, and compared to 106 population-based controls.
Odds ratios (ORs) adjusted for major risk factors of breast cancer but not for total energy intake showed a non-significant inverse association between a high intake of vegetables (OR=0.76, 95% CI=0.48-1.20) and fruits (OR=0.74, 95% CI=0.48-1.15) and breast cancer. Once results were adjusted for total energy intake the associations became much weaker (vegetables: R=0.86, 95% CI=0.51-1.46; fruits: OR=0.82, 95% CI=0.51-1.32). For all nutrients, the effect of energy adjustment was more profound and the inverse associations disappeared when results were adjusted for energy intake (total folate-not energy adjusted: OR = 0.79, 95% CI=0.51-1.21; energy adjusted: OR=1.14, 95% CI=0.73-1.79; folate equivalents-not energy adjusted: OR=0.81, 95% CI=0.53-1.23; energy adjusted: OR=1.16, 95% CI=0.78-1.74; methionine-not energy adjusted: OR=0.60, 95% CI=0.35-1.03; energy adjusted: OR=1.29, 95% CI=0.76-2.19; cysteine-not energy adjusted: OR=0.52, 95% CI=0.29-0.94; energy adjusted: OR=1.22, 95% CI=0.75-1.97). Alcohol intake was inversely associated with breast cancer in a non-significant way, possibly due to the relatively low alcohol intake of the study population.
The results of this study do not provide firm evidence that a high intake of fruits and vegetables, folic acid, methionine or cysteine reduces the risk of getting breast cancer.
Public Health Nutrition 10/1998; 1(3):147-56. · 2.25 Impact Factor
[show abstract][hide abstract] ABSTRACT: The strongest evidence that monunsaturated fat may influence breast cancer risk comes from studies of southern European populations, in whom intake of oleic acid sources, particularly olive oil, appears protective. No previous study has examined the relation of adipose tissue fatty acid content to breast cancer in such a population. We used adipose biopsies with diverse fat intake patterns gathered in 5 European centers, including southern Europe (Malaga, Spain), to test the hypothesis that stores of oleic acid or other monounsaturates are inversely associated with breast cancer. Gluteal fat aspirates were obtained from 291 postmenopausal incident breast cancer patients and 351 control subjects, frequency-matched for age and catchment area. Logistic regression was used to model breast cancer by monounsaturates, with established risk factors controlled for. Oleic acid showed a strong inverse association with breast cancer in the Spanish center. The odds ratio for the difference between 75th and 25th percentiles was 0.40 (95% CI: 0.28, 0.58) in Malaga and 1.27 (0.88, 1.85) in all other centers pooled, with a peak at 2.36 (1.01, 5.50) for Zeist. Palmitoleic and myristoleic acids showed evidence of an inverse association outside Spain, and cis-vaccenic acid showed a positive association in 3 centers. These data do not support the hypothesis that increasing tissue stores of oleic acid are protective against breast cancer in non-Spanish populations. This finding implies that the strong protective associations reported for olive oil intake in dietary studies may be due to some other protective components of the oil and not to the direct effect of oleic acid uptake. Alternatively, high olive oil intake may indicate some other protective aspect of the lifestyle of these women.
American Journal of Clinical Nutrition 08/1998; 68(1):134-41. · 6.50 Impact Factor
[show abstract][hide abstract] ABSTRACT: It has been stated that energy adjustment can control for recall bias in case-control studies. Simulation of recall bias and cases and controls in a nutritional survey of German adults was conducted to examine its impact on five dietary effects, (adding a macronutrient, substituting one macronutrient for another, adding a macronutrient while keeping the other energy sources constant, and changing the macronutrient to energy ratio through addition or substitution) using various energy adjustment models. If energy adjustment were an effective means of correcting measurement error, the energy adjusted dietary effects, after a subtraction of energy and fat intake, should equal those in the original data set. Simulation of differential under-reporting of fat and energy intake by cases but not controls showed this to dramatically impact all five considered dietary effects, even after energy adjustment. The influence of the assumed recall bias on the different effects depends on the error type structure, inflating an odds ration of 1.8 to as much as 12.3 or reducing it to 0.45 when 100 kcal of fat was substituted for 100 kcal of other macronutrients. Although energy adjustment may serve many functions, it cannot correct for differential error. Depending upon the nature of the hypothesized effect and the error type, energy adjustment may also distort risk ratios in the presence of non-differential bias. The concern that cases and controls report their energy intakes with different degrees of error remains a critical consideration that must be addressed through improved measurements, and not energy adjustment under any of the currently used models.
Journal of Clinical Epidemiology 06/1998; 51(5):393-8. · 5.33 Impact Factor
[show abstract][hide abstract] ABSTRACT: This study evaluated the iron sufficiency of the Russian diet.
Data were obtained from 24-hour dietary recalls conducted in 4 rounds (1992 through 1994) of a nationally representative longitudinal survey of 10,548 women and children. Iron bioavailability was estimated via algorithms adjusting for enhancers (heme, vitamin C) and inhibitors (tannins in tea, phytates in grains) consumed at the same meal.
Dietary iron intakes were deficient in the most vulnerable groups: young children and women of reproductive age. Poverty status was strongly associated with deficiency. After adjustment for enhancers and inhibitors, estimated bioavailable iron intakes at 3% to 4% of total iron were inadequate in all women and children.
These dietary data suggest that Russian women and children are at high risk of iron deficiency. Grain products rich in phytates, which inhibit absorption, were the major food source of iron in Russia. High intakes of tea and low consumption of vitamin C also inhibited iron bioavailability. Since changes in eating behavior could potentially double iron bioavailability, educational programs should be explored as a strategy for improving iron nutriture.
American Journal of Public Health 04/1998; 88(4):576-80. · 3.93 Impact Factor
[show abstract][hide abstract] ABSTRACT: The fatty acid content of adipose tissue in postmenopausal breast cancer cases and controls from five European countries in the European Community Multicenter Study on Antioxidants, Myocardial Infarction, and Cancer (EURAMIC) breast cancer study (1991-1992) was used to explore the hypothesis that fatty acids of the omega-3 family inhibit breast cancer and that the degree of inhibition depends on background levels of omega-6 polyunsaturates. Considered in isolation, the level of omega-3 or omega-6 fat in adipose tissue displayed little consistent association with breast cancer across study centers. The ratio of long-chain omega-3 fatty acids to total omega-6 fat showed an inverse association with breast cancer in four of five centers. In Malaga, Spain, the odds ratio for the highest tertile relative to the lowest reached 0.32 (95% confidence interval 0.13-0.82). In this center, total omega-6 fatty acid was strongly associated with breast cancer. With all centers pooled, the odds ratio for long-chain omega-3 to total omega-6 reached 0.80 for the second tertile and 0.65 for the third tertile, a downward trend bordering on statistical significance (p for trend = 0.055). While not definitive, these results provide evidence for the hypothesis that the balance between omega-3 and omega-6 fat may play a role in breast cancer.
American Journal of Epidemiology 02/1998; 147(4):342-52. · 4.78 Impact Factor
[show abstract][hide abstract] ABSTRACT: Associations between reported consumption of animal products and chlorinated hydrocarbon concentrations were examined in 297 elderly people who lived in Germany. Consumption of beef and lamb was correlated positively with hexachlorobenzene (HCB), beta-hexachlorocyclohexane (beta-HCH), total polychlorinated biphenyls (PCBs), and total dichloro-diphenyl-trichloroethane (DDT) (r = .13-.19, p < .05). Consumption of saltwater fish was correlated positively with alpha-HCH, dieldrin, and PCBs (r = .12-.26, p < .05). Other univariate predictors were body mass index, plasma cholesterol, pork consumption, poultry consumption, and age. Multivariate linear models of predictors of each chlorinated species were constructed, and some form of meat was used as the main predictor; the sum of all meats (exclusive of fish) was the best predictor of dieldrin and In(alpha-HCH) concentrations. Beef and lamb consumption was a positive predictor of HCB, heptachlor epoxide, total DDT, and beta-HCH. Saltwater fish was the major dietary predictor of PCBs.
Archives of Environmental Health An International Journal 01/1998; 53(2):147-55.
[show abstract][hide abstract] ABSTRACT: Study of the effects of most individual biologically active dietary fatty acids on human disease requires the use of biomarkers of long-term intake in well-designed epidemiologic studies. Several small studies of tissue taken from women undergoing surgery for breast abnormalities have compared fatty acid profiles of women with ascertained metastatic breast cancer with those with other abnormalities. These studies, although often flawed in design and generally of inadequate statistical power to determine significant differences, provide some evidence. Human studies are generally consistent with animal models suggesting a protective effect of n-3 fatty acids, a detrimental effect of high n-6 fatty acids, and the possible importance of the ratio of these two classes of dietary fatty acids on both breast cancer incidence and recurrence. High intakes of monounsaturated fatty acid were also often negatively associated with breast cancer. The effects of trans fatty acids have rarely been studied, but there are some indications that they may enhance risk. In general, the study of individual fatty acids is in its infancy. Larger well-designed studies with diverse population and modern analyses of individual fatty acids are needed.
American Journal of Clinical Nutrition 01/1998; 66(6 Suppl):1548S-1556S. · 6.50 Impact Factor
[show abstract][hide abstract] ABSTRACT: Epidemiological and experimental evidence indicates that consumption of fried meats in conjunction with certain genotypes of phase I and II metabolism genes poses an elevated risk for colorectal cancer. Parallel to this, the consumption of cruciferous vegetables is associated with a reduced risk of colon cancer. Therefore, we designed a 6-week pilot feeding study to evaluate the effect of these variables on urinary mutagenicity, which is a biomarker associated with fried-meat consumption. Eight subjects were fed fried meats daily for six weeks; four ate cruciferous vegetables, and four ate non-cruciferous vegetables. Urinary mutagenicity was evaluated in the presence of S9 in strain YG1024 of Salmonella, which is a frameshift strain that overproduces acetyltransferase. C18/methanol extracts of 24-h urines collected once each week were tested unhydrolyzed (free mutagenicity) and hydrolyzed (total mutagenicity); the difference between the two was the conjugated mutagenicity. Although not significant, the levels of conjugated urinary mutagenicity doubled among crucifera consumers and decreased to 30% of the initial levels among non-crucifera consumers, suggesting the possibility that crucifera may enhance the level of conjugated urinary mutagenicity resulting from consumption of fried meats. Such an effect would be consistent with the documented ability of cruciferous vegetables to induce phase II enzymes. The NAT2 rapid phenotype was significantly associated with approximately 2-fold increases in conjugated (p = 0.05) and total (p = 0.004) urinary mutagenicity relative to NAT2 slow subjects, consistent with the elevated risk confirmed by the NAT2 rapid phenotype for colorectal cancer among meat consumers. An approximately 2-fold increase in urinary mutagenicity among the GSTM1- subjects relative to the GSTM1+ subjects approached significance for free (p = 0.18) and total (p = 0.13) urinary mutagenicity. This is the first report on (a) the mutagenicity of hydrolyzed urine, which was consistently more mutagenic than unhydrolyzed urine; (b) the potential enhancement of conjugated urinary mutagenicity by crucifera; and (c) the association of the rapid NAT2 and possibly the GSTM1- phenotype with elevated levels of fried meat-associated urinary mutagenicity.
Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis 12/1997; 381(1):83-96. · 3.90 Impact Factor
[show abstract][hide abstract] ABSTRACT: A multicenter case-control study was conducted to evaluate the relations between antioxidant status assessed by biomarkers and acute myocardial infarction. Incidence cases and frequency matched controls were recruited from 10 European countries to maximize the variance in exposure within the study. Adipose tissue needle aspiration biopsies were taken shortly after the infarction and analyzed for levels of carotenoids and tocopherols. An examination of colinearity including all covariates and the three carotenoids, α-carotene, β-carotene, and lycopene, showed that the variables were sufficiently independent to model simultaneously. When examined singularly, each of the carotenoids appeared to be protective. Upon simultaneous analyses of the carotenoids, however, using conditional logistic regression models that controlled for age, body mass index, socioeconomic status, smoking, hypertension, and maternal and paternal history of disease, lycopene remained independently protective, with an odds ratio of 0.52 for the contrast of the 10th and 90th percentiles (95% confidence interval 0.33–0.82, p = 0.005). The associations for α- and β-carotene were largely eliminated. We conclude that lycopene, or some substance highly correlated which is in a common food source, may contribute to the protective effect of vegetable consumption on myocardial infarction risk.
American Journal of Epidemiology 11/1997; · 4.78 Impact Factor
[show abstract][hide abstract] ABSTRACT: To investigate the relationship between trans fatty acids and postmenopausal breast cancer in European populations differing greatly in their dietary fat intakes, a case control study using adipose tissue stores of trans fatty acids as a biomarker of exposure was conducted. Subjects included 698 postmenopausal incident cases of primary breast cancer and controls randomly drawn from local population and patient registries, ages 50-74 Concentrations of individual trans fatty acids in gluteal fat biopsies were measured in these women. The adipose concentration of trans fatty acids showed a positive association with breast cancer. The covariate-adjusted association with breast cancer. The covariate-adjusted OR was 1.40 (95% confidence interval: 1.02, 1.93) for the difference between the 75th and 25th percentiles of total adipose trans. The adjusted OR for trans in the lowest tertile of polyunsaturated fatty acid reached 3.6 (2.2, 6.1). These associations were not attributable to differences in age, body mass index, exogenous hormone use, or socioeconomic status. These findings suggest an association of adipose stores of trans fatty acids with postmenopausal breast cancer in European women. They require confirmation in other populations, with concomitant consideration of the potential roles of dietary saturated and monounsaturated fats.
[show abstract][hide abstract] ABSTRACT: Although determining iron intakes is essential in assessing adequacy of iron in the diet, estimating iron availability may be more useful for evaluating whether iron requirements are met. Our objectives were to describe the dietary information, analytical steps, and computer algorithms needed for iron bioavailability adjustments and to demonstrate the effects of various dietary factors on calculated iron absorption. Our study was based on 9890 women and children participating in the Russian Longitudinal Monitoring Survey. Between August 1992 and February 1993, two 24-h recalls were collected from each participant, and total, heme and nonheme iron intakes were calculated. Nonheme iron availability was adjusted for meat, fish and poultry and vitamin C consumed in the same meal and then further adjusted for tea and phytates. We found mean total iron intakes to be comparable to those of women of reproductive age in the United States and lower than those of United States children. When these intakes were adjusted for enhancers and inhibitors of absorption, the iron bioavailability in these vulnerable Russian groups was extremely low. Mean bioavailable iron as well as the 25th-75th percentile ranges of intake were below the bottom of the range of requirements, indicating that iron adequacy in this population may be considerably less than expected based on total iron intakes alone. Furthermore, rural and urban food availability had a significant effect on iron bioavailability. Future research on dietary iron adequacy should be based on estimates of available iron by collecting meal-level dietary data and using detailed information on mixed dishes and phytates.
Journal of Nutrition 09/1997; 127(8):1456-68. · 4.20 Impact Factor