Body-mass index and risk of obesity-related complex multimorbidity: an observational multicohort study

Abstract

Background: The accumulation of disparate diseases in complex multimorbidity makes prevention difficult if each disease is targeted separately. We aimed to examine obesity as a shared risk factor for common diseases, determine associations between obesity-related diseases, and examine the role of obesity in the development of complex multimorbidity (four or more comorbid diseases).

Methods: We did an observational study and used pooled prospective data from two Finnish cohort studies (the Health and Social Support Study and the Finnish Public Sector Study) comprising 114 657 adults aged 16-78 years at study entry (1998-2013). A cohort of 499 357 adults (aged 38-73 years at study entry; 2006-10) from the UK Biobank provided replication in an independent population. BMI and clinical characteristics were assessed at baseline. BMIs were categorised as obesity (≥30·0 kg/m²), overweight (25·0-29·9 kg/m²), healthy weight (18·5-24·9 kg/m²), and underweight (<18·5 kg/m²). Via linkage to national health records, participants were followed-up for death and diseases diagnosed according to the International Classification of Diseases 10th Revision (ICD-10). Hazard ratios (HRs) with 95% CIs and population attributable fractions (PAFs) for associations between BMI and multimorbidity were calculated.

Findings: Mean follow-up duration was 12·1 years (SD 3·8) in the Finnish cohorts and 11·8 years (1·7) in the UK Biobank cohort. Obesity was associated with 21 non-overlapping cardiometabolic, digestive, respiratory, neurological, musculoskeletal, and infectious diseases after Bonferroni multiple testing adjustment and ignoring HRs of less than 1·50. Compared with healthy weight, the confounder-adjusted HR for obesity was 2·83 (95% CI 2·74-2·93; PAF 19·9% [95% CI 19·3-20·5]) for developing at least one obesity-related disease, 5·17 (4·84-5·53; 34·4% [33·2-35·5]) for two diseases, and 12·39 (9·26-16·58; 55·2% [50·9-57·5]) for complex multimorbidity. The proportion of participants of healthy weight with complex multimorbidity by age 75 years was observed by age 55 years in participants with obesity, and degree of obesity was associated with complex multimorbidity in a dose-response relationship. Compared with obesity, the association between overweight and complex multimorbidity was more modest (HR 2·67, 95% CI 1·94-3·68; PAF 13·3% [95% CI 9·6-16·3]). The same pattern of results was observed in the UK Biobank cohort.

Interpretation: Obesity is associated with diverse, increasing disease burdens, and might represent an important target for multimorbidity prevention that avoids the complexities of multitarget preventive regimens.

Funding: Wellcome Trust, Medical Research Council, National Institute on Aging.

Figure 1

Study profile

Figure 2

Associations between obesity and incidence of 78 diseases and death ICD-10=International Classification of Diseases 10th Revision. HR=hazard ratio. *HR for obesity (BMI ≥30·0 kg/m²) versus healthy weight (18·5–24·9 kg/m²), adjusted for age, sex, cohort (Finnish cohorts), education, and neighbourhood deprivation (basic model); HRs greater than or equal to 1·50 and significant at p<0·0006 (Bonferroni correction for multiple testing) are shown in bold. †Non-overlapping disease diagnoses included in analysis of obesity-related multimorbidity.

Figure 3

Cumulative incidence of obesity-related simple multimorbidity and complex multimorbidity by age with obesity versus healthy weight (A) Finnish cohorts. (B) UK Biobank cohort. HR=hazard ratio. *Hazard ratio (95% CI) adjusted for age, sex, ethnicity (UK Biobank), and cohort (Finnish cohorts). †Maximum number of participants included in follow-up.