Research from the past few years has demonstrated that far better correlations between weight and mortality can be obtained by considering the history of individual weight rather than just taking snapshots of populations at a moment in time. In particular, some studies produced results suggesting that being overweight has a lower mortality rate in older age, but this happened as a result of failing to consider weight changes in the studied populations; following work has fairly comprehensively torn down those results. Consider that, for example, a fair number of thinner old people were overweight when younger but suffer from chronic medical conditions that produce both weight loss and much higher mortality. They distort the data, being completely different from people who were always thinner and are as a result healthier in old age. This is why you can't put all people who are thin at a given moment in time into one statistical bucket and expect sensible data on the other side of the calculations: garbage in, garbage out. The publicity materials linked here cover studies that provide more data along these same lines:
People who are lean for life have the lowest mortality, while those with a heavy body shape from childhood up to middle age have the highest mortality, reveal findings of a large study. Researchers tracked the evolution of body shape and associated mortality among two large cohort studies. In total, 80,266 women and 36,622 men enrolled in the Nurses' Health Study and Health Professionals Follow-up Study, recalled their body shape at ages 5, 10, 20, 30, and 40 years. They also provided body mass index at age 50, and were followed from age 60 over a median of 15-16 years for death. They answered detailed questionnaires on lifestyle and medical information every two years, and on diet every four years.
Among the cohort, five distinct body shapes were identified from age 5 to 50: lean-stable, lean-moderate increase, lean-marked increase, medium-stable/increase, and heavy-stable/increase. Results showed that people who remained stably lean throughout life had the lowest mortality, with a 15-year risk of death being 11.8% in women, and 20.3% in men. Those who reported being heavy as children and who remained heavy or gained further weight, especially during middle age, had the highest mortality, with a 15-year risk of death being 19.7% in women and 24.1% in men. The authors conclude: "our findings provide further scientific rationale for recommendations of weight management, especially avoidance of weight gain in middle life, for long-term health benefit."
In a second study, an international team of researchers confirm that increasing levels of body mass index (BMI) are associated with higher risks of premature death. The BMI is an established way of measuring body fat from the weight and height of a person, but the optimal BMI associated with the lowest mortality risk is not known. It's expected that a higher BMI is associated with a reduced life expectancy, but the largest previous study showed that when compared with normal weight, overweight was associated with reduced mortality, and only high levels obesity were associated with increased mortality.
So researchers in the current study sought to clarify this association by carrying out a large meta-analysis of 230 prospective studies with more than 3.74 million deaths among more than 30.3 million participants. They analysed people who never smoked to rule out the effects of smoking, and the lowest mortality was observed in the BMI range 23-24 among this group. Lowest mortality was found in the BMI range 22-23 among healthy never smokers, excluding people with prevalent diseases. And among people who never smoked, and studied over a longer duration of follow up of more than 20 and 25 years, where the influence of prediagnostic weight loss would be less, the lowest mortality was observed in the BMI range 20-22.