October 28, 2014

expert reaction to milk consumption, fractures and mortality

A paper published in The BMJ reported an association between higher intake of milk and higher mortality in cohorts of men and women. In the cohort of women, they also observed a higher incidence of fractures with higher milk intake. The researchers recommend cautious interpretation of the data due to it being an observational study, and possible confounding effects.

Catherine Collins, Principal Dietitian at St George’s Hospital  NHS Trust, said:

“The study as reported creates more questions than provides answers, and whilst an interesting read adds nothing to influence current UK dietary recommendations related to dairy products and health. It fails to confirm the hypothesis that a component of milk sugar – galactose – is clearly implicated in bone fracture and earlier mortality. The results were dietetically ambiguous, and there is no need to recommend change in peoples’ usual consumption of milk and dairy products based on this work – even for those consuming twice the average intake.

“The authors hypothesised that galactose, a simple sugar found mainly in milk, contributes to oxidative stress and inflammation within the body, increasing the risk of fracture and earlier mortality. The naturally lower lactose/galactose content of yoghurt and other fermented dairy products, and the negligible content in cheese, would hypothetically generate different risks if galactose was directly involved in fracture risk and mortality. This hypothesis remains unproven.

“There are several limitations relating to this study. Firstly, the authors did not define the type of physical activity taken by participants, yet physical activity of a weight bearing nature such as running has a greater impact on maintaining bone ‘strength’ than non-weight bearing exercise such as swimming. Only duration and intensity of activity, translated into a standard unit of ‘METS’ was reported, and this value was insufficient to define the impact of subject activity on bone health.

“Secondly,  whilst excluding ‘pathological’ bone fractures (caused by a health condition, such as breast cancer) from the data, the inclusion of high impact trauma fractures (eg fractures resulting from a car crash) with low impact fractures (eg person trips on rug at home, falls and sustains a hip fracture) is unusual. Without clarification the additional yearly fracture risk suffered by 1 in every 1000 women who reported being avid milk drinkers could merely reflect poor driving skills and loss of collision damage waiver.

“There is also the concern that dietary habits and exercise levels change over the years in many people, so a ‘snapshot’ of diet and exercise taken over a decade ago often fails to reflect current habitual diet. Furthermore, women consuming over a pint of semi skimmed milk per day would on average be consuming 550kcals (25-30% calorie needs in this age group) than women consuming the UK average of 1/3 pint a day, yet BMI remained remarkably consistent across all groups irrespective of energy intake.  This discrepancy highlights the inaccuracies of longitudinal studies on diet and lifestyle in representing true intakes, and serves only to demonstrate how trends, rather than absolute amounts of a particular food, may influence the health of a population.”

Prof Sue Lanham-New, Head of the Department of Nutritional Sciences, University of Surrey, said:

“This new study in BMJ has a number of limitations.  Indeed the authors refer to this in the final sentence of their abstract.  The study is only observational and the method used to assess diet is weak with no estimation of total energy intake or the extent of under reporting which means we are not sure how accurate the cut off values used for frequency of milk consumption is.  Neither do we have a feel for the influence of physical activity or other life style habits important to bone or overall mortality. The effect of increasing BMI has not been fully investigated in this study.

“Also key is that all milk in Sweden is fortified with vitamin A.  This may well be an important confounding factor that the authors do not address.

“Milk and dairy products in the UK provide 50-60 percent of the calcium in our diet.  We know that low calcium intakes (less than 400mg per day) is a risk factor for osteoporosis.  Individuals should still be encouraged to consume a balanced diet from the five key food groups of which milk and dairy are key.”

Gaynor Bussell, Dietitian and Public Health Nutritionist, said:

“An interesting study and well thought out, but as the authors say, some caution required here.  Firstly food frequency questionnaires are not the gold standard for estimating food intake, as the authors point out. Secondly, we cannot assume cause and effect as there may be another factor that is hard to identify causing the increased mortality and fracture rate in women, but seems to track with milk consumption.

“We do know that increase in inflammation is associated with reduced bone density and so the effect of galactose certainly needs to be looked at in further studies. Calcium is required in the diet for bone health and we require approximately the daily amount of calcium that’s in about a pint of milk.  So milk is a very convenient source of calcium as well as many other vitamins and minerals.  So as the author said, some caution is required here in interpreting the results and so I would urge some more research in this area that can back or refute these findings.  One such study is insufficient to base public health decisions on.”

Prof Brian Ratcliffe, Professor of Nutrition, Robert Gordon University, Aberdeen, said:

“This is a thought-provoking study.  The authors seem to have taken care with the design and analysis of the study and they have clearly pointed out that their conclusions should not be over interpreted because of the possibilities of confounding of the data for which they have not accounted. The effects that they report for the highest consumption of milk compared with the lowest seem to be most marked in women compared with men.  Although the authors have taken various factors into account such as smoking (the prevalence was higher in the high milk drinkers), BMI, and use of vitamin D supplements (lowest in the high milk consumers), they seem to have understated the possible role of alcohol intake.

“From their data, it looks as though most of the extra mortality in the highest milk drinkers is related to increased death rates from cardiovascular disease (there is little difference across the categories of intake for cancer) but the lowest level of milk consumption was associated with the highest consumption of alcohol (63% higher for women and 42% higher for men) with a graded inverse relationship with milk intake.  There is considerable evidence for a cardio-protective effect of alcohol at modest intakes and this may be the source of more confounding than the authors’ acknowledge.

“Furthermore, the lowest category of milk intake also had the lowest absolute intake of saturated fatty acids (and, in women, this was the lowest intake expressed as a percentage of energy intake).  The results showing higher levels of biomarkers of oxidative stress and inflammation in relation to higher milk consumption is worthy of further investigation.  Certainly the effects reported are only for milk, since the authors did not see detrimental effects for other dairy products, and the consumption of liquid milk in the UK has fallen considerably over the last three decades.”

‘Milk intake and risk of mortality and fractures in women and men: cohort studies’ by Michaëlsson et al. published in The BMJ on Tuesday 28^th October.

Declared interests

Prof Sue Lanham-New: Received funding from National Dairy Council and Yoghurt Council.  Views expressed are those of the author alone.

No other conflicts declared.