Research Update

Full Breastfeeding Duration and Risk for Iron Deficiency in U.S. Infants

A study published in Breastfeeding Medicine as the lead article by Caroline Chantry et al. titled “Full Breastfeeding Duration and Risk for Iron Deficiency in U.S. Infants” suggests that infants fully breastfed for 6 months or more in the United States may be at increased risk of iron deficiency and anemia. 

This study analyzed data from two nationally conducted cross-sectional surveys, the National Health and Nutrition Examination Survey (NHANES) III, which took place from 1988-1994 and NHANES 1999-2002.  The NHANES III included 2268 infants and children, aged 6 to 24 months. Anemia and iron status were compared for five groups: formula fed only (n = 1142), full breastfeeding <1 month (n = 425), full breastfeeding 1-<4 months (n = 343), full breastfeeding 4-<6 months (n = 222), or full breastfeeding > 6 months (n = 136).  A similar analysis was conducted for 526 children, age 12 to 24 month olds from NHANES 1999-2002 with the following groups:  formula fed only (n = 181), full breastfeeding <1 month (n = 113), full breastfeeding 1-<4 months (n = 116), full breastfeeding 4-<6 months (n = 28), or full breastfeeding > 6 months (n = 88).  “Full breastfeeding” was defined as the use of breastmilk as the overwhelming majority of the diet.

In this study, there was one measure of iron deficiency: low serum ferritin (< 7 m  /L at age 12 - <24 months), and two measures of anemia: low hemoglobin (<10.5 g % 12- <24 months) and parentally reported history of anemia.  Note that lab data was only available for 745 children aged 12- <24 months old from the NHANES III data set and for 294 children from the NHANES 1999-2002 dataset.  Analysis was performed on each group for these three outcomes relative to the breastfeeding >6 months group.

With respect to iron deficiency, the study found that there was no significant association of low serum ferritin and any feeding status for the NHANES III survey.  However, in the NHANES 1999-2002 survey, there was a five fold increased risk of low serum ferritin for full breastfeeding >6 months.  The reason for the discrepancy between the two datasets, the author states, is unclear.

As for reported anemia, the study found that NHANES III infants who were fully breast-fed for >6 months had 3-5 times the risk of having a reported history of anemia compared to those fully breastfed for 1 to <6 months and twice the risk of reported history of anemia compared to formula fed infants.  The association held up even when the analysis controlled for factors typically associated with iron deficiency, such as birth weight, demography, race/ethnicity, gender, socioeconomic, status, and education.  However, in the NHANES 1999-2002 dataset, no significant association was seen between any groups and a history of anemia.  One explanation for these findings is the age discrepancy in reported anemia between the 2 datasets; information was available for children 6 to 24 months in NHANES III while NHANES 1999-2002 was limited to 12 to 24 months.  Lab data was not available for 6 to 12 month olds in the NHANES III study.  These age differences and differences in sample size could potentially account for the differences in outcomes.

When the authors examined hemoglobin status, they found that NHANES III infants who were breastfed for >6 months were at a significant less risk for low hemoglobin than infants fully breastfed <4 months (including formula fed).  The authors explain the inconsistent findings of hemoglobin and history of anemia in relation to breastfeeding by saying children diagnosed with anemia by standard screening at 9-12 months received treatment and thus were more likely to be partially iron repleted.  Therefore, they were less likely to be anemic, as measured by hemoglobin status.  No significant association was observed for the NHANES 1999-2002 children for breastfeeding length and hemoglobin status.

According to the authors, the study has several limitations, besides the inconsistent results upon analysis of the two NHANES datasets.  For history of anemia, there is a potential for recall bias, since the parentally reported information was not verified by doctor’s records.  However, the authors disregard this bias as a confounder, as it is “unlikely that recall bias would differ by breast-feeding status.”  Additionally, there was also incomplete information regarding iron supplementation (data only available for those who took supplements within the past 30 days).  In this study, there was limited laboratory data for both ferritin and hemoglobin with no laboratory data available for ages 6 to 12 months of age in the NHANES III dataset.  Despite these limitations, the authors conclude that their finding “needs confirmation,” but recommend that physicians stress the importance of foods rich in iron to children 6 months of age or older.  For commentary on this article from the Editor of Breastfeeding Medicine, Ruth Lawrence, who acknowledges “the work as valid [and] the report as accurate,” click here.

Chantry et al.  Full Breastfeeding Duration and Risk for Iron Deficiency in U.S. Infants.  Breastfeed Med; 2007; Jun; 2 (2): 63-73.