Send to

Choose Destination
PLoS One. 2019 Feb 22;14(2):e0212862. doi: 10.1371/journal.pone.0212862. eCollection 2019.

Body mass index trajectories preceding first report of poor self-rated health: A longitudinal case-control analysis of the English Longitudinal Study of Ageing.

Author information

Department of Public Health, Aarhus University, Aarhus, Denmark.
Danish Diabetes Academy, Odense, Denmark.
Steno Diabetes Center Aarhus, Aarhus University Hospital, Aarhus, Denmark.



Studies have consistently found that obesity is associated with poor self-rated health, but how body mass index (BMI) developed in the lead up to poor self-rated health is unknown.


We nested a longitudinal case-control study in the English Longitudinal Study of Ageing (1998-2015) to investigate BMI trajectories in the years preceding a first self-report of poor health. Participants rated their health at each data collection; every other collection included a BMI assessment by a nurse. Case status was defined as a first report of poor health during follow-up. Three age- and sex-matched controls were identified per case using density sampling. BMI trajectories were fitted to time backwards prior to first report of poor health using mixed-effects models. Age and sex were potential modifiers. We conducted subgroup analyses of those not reporting certain chronic diseases or smoking.


We identified 732 cases and 2195 controls. Age, but not sex, modified the association between BMI and self-rated health. Participants reporting poor health at age 60 had a 1.5 kg/m2 (95%CI: 0.8 to 2.1) higher BMI at the time of reporting than controls, and their BMI had previously increased markedly (1.3 kg/m2 95%CI: 0.9 to 1.8 over ten years). After age 75, cases no longer had higher BMI than controls, and their BMI had decreased sharply prior to reporting poor health (e.g. -2.0 kg/m2 95%CI: -2.6 to -1.5 per decade on average for those reporting poor health at age 90). Age was also an effect modifier among those without diabetes, however BMI trajectories were more similar among the middle-aged. The subgroup analysis of those without cardiovascular disease, cancer and chronic lung disease showed similar results to the main findings.


Development of BMI was associated with poor self-rated health; however, the nature of the association depended markedly on age.

Conflict of interest statement

The authors have declared that no competing interests exist.

Supplemental Content

Full text links

Icon for Public Library of Science Icon for PubMed Central
Loading ...
Support Center