Objective: To estimate the excess deaths associated with weight loss in combination with leisure time physical activity among overweight or obese people.
Design: Prospective cohort study.
Subjects: In two consecutive examinations in 1976-1978 and 1981-1983, 11 135 people participated in the Copenhagen City Heart Study. Of these, 3078 overweight or obese participants lost weight or remained weight stable from 1976-1978 to 1981-1983, and were without pre-existing diagnosis of diabetes, stroke, ischaemic heart disease or cancer in 1981-1983. They were followed up until 2007 in the Danish Civil Registration System, with a <0.2% loss to follow-up only.
Measurements: The following measurements were taken: body mass index (BMI) and physical activity in 1976-1978 and 1981-1983 and hazard ratio (HR) of mortality during 53 976 person-years of follow-up.
Results: Of the initially overweight or obese subjects who either lost weight or remained weight stable, 2060 died. Overall, weight loss was associated with excess mortality when compared with weight stability. Weight loss was associated with a higher mortality among those who became physically inactive, compared with those who remained active while losing weight (men: HR 2.25, 95% confidence interval 1.31-3.84; women: 1.43, 1.07-1.91). However, losing weight while remaining physically active was still associated with excess mortality when compared with those who were weight stable and initially active (men: 1.72, 1.27-2.34; women: 1.57, 1.06-2.31). Among those who remained physically inactive, weight loss seemed associated with excess mortality when compared with weight loss among those who became active, although not statistically significant (men: 2.00, 0.94-4.29; women: 1.40, 0.82-2.39). Finally, weight loss among those who became physically active was not associated with excess mortality when compared with those who were weight stable and initially inactive (men: 1.12, 0.61-2.07; women: 1.19, 0.58-2.43).
Conclusion: Weight loss among the overweight or obese seemed hazardous to survival. However, weight loss seemed less hazardous to survival among those who remained physically active or those who became active.