Another sign that breast is best

“Babies who are breastfed have lower cholesterol levels in later life”, the Daily Express reports. It says that researchers who reviewed the histories of 4,608 children given formula milk and 12,890 breastfed children found that in adulthood, those who were breastfed had a lower overall level of cholesterol.

This story was based on a review that combined the results of several studies on infant feeding practices and cholesterol levels in adult life. Although the results did show a very small reduction in cholesterol in adults who had ever been breastfed compared to those formula-fed, it cannot prove that this was caused by breastfeeding. This is not the due to the quality of the review, which was well conducted, but due to limitations that are inherent when combining data from several studies that have different designs and methods. However, despite offering no conclusions on whether breastfeeding reduces cholesterol in later life, this review provides another indication that breast may be best.

Where did the story come from?

Dr Christopher Owen and colleagues from the University of London, University of Bristol, Umea University, Sweden, and several other institutions across the UK and worldwide, carried out the research. The study was funded by the British Heart Foundation and was published in the peer-reviewed medical journal: American Society for Nutrition.

What kind of scientific study was this?

In this systematic review, the researchers looked at the current published evidence on whether breastfeeding is associated with lower blood cholesterol in adult life.

A literature search was carried out in a number of electronic databases to look for published observational studies, review articles or letters that had recorded infant feeding practices and measured blood cholesterol levels at some point in adult life. Following the exclusion of studies that had not collected the necessary data, the researchers were left with 18 publications of 17 studies (mostly North European) that involved 17,498 participants.

Where possible, the researchers contacted the authors of the primary research (13 studies) to obtain the original study data. Where this was not possible, they used the data from the published reports (four studies). They were looking for data on groups of exclusively breastfeeding and exclusively bottle feeding mothers. They were also interested in the duration that the child was breast fed or bottle fed, or the duration that they were given both.

All but four of the studies had measured cholesterol after the subjects had fasted. The data was analysed to look for differences in total cholesterol levels in adults who had been characterised as being breastfed compared to those characterised as formula-fed. The analyses took into account various potential confounders such as age, sex, BMI, smoking and socioeconomic variables. Attempts were also made to consider the type of formula the child was given and its year of birth.

What were the results of the study?

It was possible to compare exclusively breastfed babies with exclusively bottle-fed babies in seven of the studies. In the other 10 studies, it was not known whether the type of feeding was exclusive. When the researchers combined all 17 studies, they found that there was marginally significant reduction in total cholesterol among adults who had been breastfed compared to those formula-fed (reduction of 0.04mmol/L; 95% confidence interval -0.08 to 0.00mmol/L). The studies had different designs and methods, and there were significant differences in the results.

When the researchers combined the data from just the seven studies that could determine exclusive feeding patterns, they found the reduction in cholesterol in those people exclusively breastfed compared to those exclusively bottle-fed was greater (reduction of -0.15mmol/L; 95% confidence interval -0.23 to -0.06mmol/L). In addition, the results of these seven studies did not differ significantly.

A separate analysis of the 10 studies that could not say for certain whether feeding patterns were exclusive, did not show any significant difference in cholesterol levels between the two groups. There was little effect upon the results when confounding factors were taken into account.

What interpretations did the researchers draw from these results?

The researchers concluded that breastfeeding, ‘particularly when exclusive’, may be linked to lower blood cholesterol concentrations in later life.

What does the NHS Knowledge Service make of this study?

This was a well-designed and carefully conducted review; however, there are unavoidable complications which limit interpretation of its results.

• It is likely that the studies that were included had various methods and designs. These studies were from different countries, assessed feeding patterns in different ways (for example some at the time of infancy, others from later questionnaires or parent recall), covered births from different time periods (with a range of birth years from 1919 to 1982), and assessed adult cholesterol at various ages (from age 17 to age 71). When studies that vary like this are combined, the reliability of the results is affected.
• The researchers made stringent attempts to obtain data from the original studies to compare groups of mothers who had either exclusively breastfed or bottle-fed. This was not possible for most of the studies, and those studies that did report children as either being exclusively bottle-fed or breastfed are subject to a number of unavoidable errors. These possible errors stem from the facts that exclusivity of infant feeding practice was reported by individual mothers, what is considered to be ‘exclusive’ feeding can be viewed as subjective, and variations in the number of months that this type of feeding lasted. In the studies where it was not possible to get information on exclusive feeding, the categories of ‘ever breastfed’ or ‘ever bottle-fed’ are going to cover a very wide range of feeding patterns.
• The researchers say that ‘bottle feeders were assumed to have been fed formula milk, not human milk’; but this may not be so in all cases.
• The difference in total cholesterol levels between the groups was minimal and it is not possible to say whether this small difference would have any health implications. No disease outcomes in the adults were examined (e.g. heart disease, strokes, blood pressure).
• The researchers reported that they chose to analyse total cholesterol rather than other measures in the blood as it had been more widely reported in studies. However, more specific markers of low-density lipoproteins (‘bad’ cholesterol) or high-density lipoproteins (‘good’ cholesterol) may have given a clearer picture of disease risk modification.
• It is difficult to account for other confounding factors that may affect both the likelihood of being breastfed and the likelihood of developing higher cholesterol in adult life. For example, parents who have healthy lifestyle behaviours, such as a good diet and plenty of exercise, may also be more likely to breastfeed their children; in turn these children may be more likely to adopt healthier lifestyle behaviours which could influence their cholesterol levels. As such, it is not possible to say for sure why the differences in later life cholesterol may be observed, and it should not be assumed at this stage that it is attributable to the constituents within formula milk (cholesterol content of breast milk is in fact higher).
• The most reliable method of answering the question of whether breastfeeding reduces cholesterol in later life would be a randomised controlled trial that balanced all other confounding factors that could affect the likelihood of breastfeeding and cholesterol risk. This, however, would involve randomly allocating children to either being breastfed or bottle-fed, which is clearly unethical and therefore not possible.

Although the question of whether breastfeeding reduces cholesterol in later life cannot be conclusively answered by this review, the study does present another indicator that breast may be best.