Breastfeeding, overweight and obesity

First, the muddled news. A recent paper on breastfeeding from the Harvard Nurses' Health StudyReference Michels, Willett, Graubard, Vaidya, Cantwell and Sansbury¹ seemed to show that breastfeeding does not protect against overweight in adulthood. No correlation was found between breastfeeding duration or exclusiveness and overweight or obesity in the studied group. It was immediately trumpeted in the media in many parts of the world. The Swedish newspaper Dagens Nyheter stated ‘Breastfeeding does not protect against overweight’, with the subtitle ‘Breastfeeding does not protect against overweight. This is shown in a new study, which thereby contradicts previous research’². CNN.com stated ‘Breastfeeding link to adult weight challenged’³, while CBS News said ‘Study: Breastfed tots no thinner as adults – but research shows suggestion of protection during early childhood’⁴.

The results of the study are out of line with systematic literature reviews of studies of breastfeeding, overweight and obesity in childhood, which tracks into adult lifeReference Arenz, Ruckerl, Koletzko and von Kries^5, Reference Owen, Martin, Whincup, Davey-Smith, Gillman and Cook⁶.

So, what's wrong with the Harvard study? It probably does not matter that the study only takes girls into account (nurses). More worrying is that it relies on self-reporting of historical data for breastfeeding duration and exclusivity, as well as on self-reported body shape, height and weight for the mother and the child (nurse) in question. Also, while more than 35 000 women were followed, only a small fraction (n = 1916; 5.5%) of the mothers claimed that they breastfed for what is now the recommended duration – i.e. for 6 months exclusively⁷.

Assessment of exclusive breastfeeding as such or duration of exclusive breastfeeding in retrospective studies is obviously problematicReference Aarts, Kylberg, Hornell, Hofvander, Gebre-Medhin and Greiner^8, Reference Bland, Rollins, Solarsh, Van den Broeck and Coovadia⁹. So is reliance on self-reported height and – even more so – weightReference Niedhammer, Bugel, Bonenfant, Goldberg and Leclerc^10–Reference Roberts¹⁸. Overweight and obese people underreport their weight. So this study has relied on insecure data on breastfeeding exclusiveness and duration and compared those with insecure data on body size and mass. But even if these data were secure, they are from one female subpopulation in the USA, living in one of the most obesogenic environments in the world, within a general population with rather low breastfeeding rates.

The evidence base for breastfeeding and its relation to adult health should be further strengthened by preferably prospective cohort studies using measured data on height and weight rather than self-reports, and using current practices of breastfeeding rather than historical. ‘The jury is still out’, commented Laurence Grummer-Strawn from the CDC, in an interview by CBS⁴. And it is.

Salt and cardiovascular disease

Second the clear news. Scientists have agreed for many years on the importance of reducing salt intake for cardiovascular health¹⁹. The evidence base showing that reduced salt intake leads to lower blood pressure is solidReference Chobanian, Bakris, Black, Cushman, Green and Izzo²⁰. But so far evidence on hard clinical outcomes from sodium reduction has been lacking. Also, few of the previous studies include robust data on salt consumption, because food composition tables are unreliable, the added salt content of manufactured foods varies, and it is hard to measure the amount of salt used at the table and in home and restaurant cooking.

Help is at hand. The Trials of Hypertension Prevention (TOHP)^{21, 22} have used sodium excretion as a proxy for sodium intake. In a recent follow-upReference Cook, Cutler, Obarzanek, Buring, Rexrode and Kumanyika²³ of one of the two TOHP studies, the long-term effects of dietary sodium reduction were studied in relation to cardiovascular outcomes. The results show 30% reductions in cardiovascular events in the intervention group at follow-up.

Salt intake has been substantially reduced over the last hundred years, at least in higher-income countries, as it has become less important as a preservative. At the same time, rates of stroke and stomach cancer have also decreased. Maybe the invention and wide use throughout food systems of the refrigerator and freezer has done more for public health than medical intervention. The current study calls for increased adherence to the salt restriction recommendations¹⁹; previous analyses show the vast economic benefits when only taking blood pressure reduction into accountReference Selmer, Kristiansen, Haglerod, Graff-Iversen, Larsen and Meyer²⁴.

How can countries reduce salt consumption? The clear answer is by means of reducing the use of salt in production and manufacture. The spotlight falls on the food industry. Also what is needed is clear labelling of sodium content of foods, applied universally. The proposed labelling of foodsReference Scarborough, Rayner and Stockley²⁵ in the UK, including a ‘traffic light’ symbol, includes high salt content as one indicator of unhealthy foods. It is hoped that such a symbol would be visible enough for consumers to respond, and effective enough to encourage industry to lower the salt content of their products.