January 22, 2019

expert reaction to study looking at body size and lifespan

Research published in the Journal of Epidemiology & Community Health indicates that body size, physical activity and gender are related to the likelihood of reaching 90 years of age.

Dr David Carslake, Senior Research Associate, MRC Integrative Epidemiology Unit, University of Bristol, said:

“It is now very clear that overweight, obesity and sedentary lifestyles are harmful to health.  Furthermore, they are becoming so common that it is important that we study their effects in more detail than has been done so far.  Studies like this one, which examine the shape of the associations, and ask whether they are the same in different groups (here, men and women), will be increasingly important.  Distinguishing causal effects of body weight from mere association remains a major challenge in the field.

“These results provide interesting hints that men’s and women’s health might respond differently to BMI, height and exercise but they have two major limitations and should not be over-interpreted.

“First, as the authors explain, there are several medical conditions which could cause weight loss – particularly given that the participants in the study were already 70 years old.  If those conditions can also lead to a premature death, we risk underestimating the harmful effects of overweight in this type of analysis.  This ‘reverse causation’ applies particularly to diseases of the lung, which are more common in men due to their higher rates of smoking and industrial exposure.  In other words, underestimation of the harmful effects of overweight could have been greater in men than in women.

“A second limitation concerns the interpretation of the statistical results.  Although the Netherlands Cohort study is large, restriction to those born 1916-1917 reduced the numbers considerably.  Furthermore, while 34% of women survived to 90, only 17% of men did.  This meant that estimates of the effects of BMI, height and exercise were much less precise among men than among women.  The wide 95% confidence intervals around the estimates for men mean that we cannot rule out the possibility that the values would have been the same in both sexes, if a larger sample had been analysed.”

*Dr Carslake’s comment was amended at his own request to correct ‘Distinguishing casual effects’ to ‘Distinguishing causal effects ‘ (10:28, 23/01/2019)

‘Body size, non-occupational physical activity and the chance of reaching longevity in men and women: findings from the Netherlands Cohort Study’ by Lloyd Brandts and Piet A van den Brandt was published in the Journal of Epidemiology & Community Health at 23:30 UK time on Monday 21 January 2019.

Declared interests

Dr David Carslake: “I don’t think I have any conflicts of interest here – I’ve not knowingly had any interaction with the authors and I have no related financial interests.  I have published previously in the field.”