Update: Dairy for children 'extends life'

So I got hold of the Bristol Uni/QIMH report from Heart. And there are some major problems here. Without boring you with a bunch of details on statistics, I’ll just point out that we use measures to estimate the variation in research and these often say more about the results than the results do. The authors of this study have ignored this, presenting information which has low statistical power as significant results in some cases and then drawing some pretty dodgy conclusions on the results. But they basically explain what is wrong with their conclusions in the text, which of course is not focused on by news media.

Just recapping quickly, the BBC article claimed that kids who ate the most dairy in childhood had the lowest risk of death from chronic heart disease (CHD) or stroke. But look what else the authors tell us about kids who ate the most dairy:

“Children in the highest group of dairy intake were more likely to be in a household with high food expenditure during childhood, more likely to have a father in a high occupational social class, and less likely to live in a deprived area or smoke as an adult than were those in the lowest intake group.”

“Children in the highest group of dairy intake were younger and their household’s intakes of vegetable and fat were higher compared to those in the lowest intake group. Dairy intake was highly correlated with total calcium intake. Intake of fruit, vegetable, egg and egg dishes, protein, fat, and energy intake was also highest in those with the highest baseline calcium intake.”

So basically they had the best all round nutrition. And since they were wealthier we can also assume they’ve had better welfare/health care and perhaps also better education which means they could be actively reducing their other risk factors such as smoking.

The authors then go on to explain that high dairy or calcium intake was associated with reduced risks of mortality from various different causes, even suicide. I can only assume that the only analyses they performed was related to dairy and dietary calcium, as they don’t report on whether the other correlations with high dairy and calcium intake (ie fruit, vegetable intake etc) also showed the same relationships or gave different results. If the agency that funded the research (World Cancer Research Fund, who was also involved in the American Institute for Cancer Research’s review of dietary cancer research) only were interested in the associations with dairy intake then this is a reasonable conclusion.

But what really stands out in this material is the fact that the data is obtained from what families ate for a single week in the 1930’s and was extrapolated out for their entire lives. More importantly, the data didn’t provide specific information on who ate what, so they made assumptions that of the foods consumed in a single household, the number of children in that household ate a given proportion and they further assumed that it was children consuming the most dairy. The authors note this, but then go on to make more assumptions on their already weak data:

“Data from the 1997 follow-up study in a sub-group of the study participants showed that the correlation between milk intake in childhood and that in adulthood was very low and thus these results are unlikely explained by similar dairy and calcium consumption patterns in adulthood and probably point to a childhood-specific effect.”

What? Probably a childhood specific effect? Or could it be that there was some other factor in the first place like overall nutrition that drives this relationship?

In closing, they point out another weak point in their material.

“The macro-environment of participants has varied considerably between the time of participation in the study at baseline, as children in 1930-ies pre-war Britain, and the time of their death. Conditions during the second World War, and participants’ environment as adults in the second half of the twentieth century will have been considerably different from those in the 1930ies. Food availability, general income, and health and housing standards have varied considerably during the study period, but such changes could not be considered in the analyses.”

To say nothing of the incredible differences in dairy and other food production systems occurring between now and then. That is what really bugs me about this study and the articles which sprang up all over the web suggesting that feeding kids dairy will help them live longer. The dairy products these people ate are not like dairy products today, thanks to the technology associated with intensive milk production. But since there are obvious holes in this report which require further explanation, I would go as far as saying this report is misleading.

So there you have it. This research teaches us nothing about nutrition, but it does reiterate the need for good data and thorough analysis. As George Fuechsel said, “Garbage in, garbage out”. Pity the BBC didn’t approach this material from the same perspective when they suggest that it backs the practice of giving dairy to children beyond infancy.

References:
Jolieke van der Pols, David Gunnell, Gail Williams, Jeff Holly, Chris Bain and Richard Martin (2009). Childhood dairy and calcium intake and cardiovascular mortality in adulthood: 65-y follow-up of the Boyd Orr cohort. Heart published online 29 Jul 2009 doi:10.1136/hrt.2009.168716.