The study, which is published in the scientific journal BMJ, also suggests that some of the adverse consequences of being underweight may be overstated.
Numerous studies have already investigated the link between body mass index (BMI) and mortality. They show that high BMI is associated with higher rates of death from cardiovascular causes, diabetes, and some cancers, while low BMI is associated with increased mortality from other causes, such as respiratory disease and lung cancer.
However, there are inconsistencies in the evidence that low body mass index actually increases the risk of causes of death such as respiratory disease and lung cancer. Some argue that this association may be biased by a process called reverse causality, where a severe illness, such as lung cancer, leads to both weight loss and higher mortality. Other factors such as smoking and poor socioeconomic circumstances may also lead to biasing estimates. This is known as confounding
Now a team of researchers from Karolinska Institutet and the University of Bristol, Britain, has found a way to obtain a valid estimate of the association between BMI and mortality by comparing the BMI of offspring when young adults with mortality among their parents for over one million Swedish parent-son pairs. Using offspring BMI as an indicator of parental BMI avoids problems of reverse causality and is less influenced by confounding.
The analysis shows strong associations between high offspring BMI (used as a so-called instrumental variable) and parental mortality from cardiovascular disease, diabetes, and some cancers, as reported in other studies of own BMI with mortality.
“This confirms and emphasizes the results in earlier studies that show a connection between high BMI and increased risk of early death in certain deceases”, says Professor Finns Rasmussen, Karolinska Institutet. “However, unlike previous studies, we did not find evidence of an association between low BMI and an increased risk of respiratory disease and lung cancer mortality, so this connection is probably highly overestimated.”
The association between body mass index and mortality using offspring BMI as an indicator of own BMI: large intergenerational mortality study
BMJ online 23 december 2009.