Association of two-timepoint clinical obesity transitions with incident anxiety disorder: a UK Biobank cohort study

Clinical obesity incorporates excess adiposity and evidence of reduced organ or tissue function. Whether two-timepoint changes in clinical-obesity state are associated with incident anxiety disorder remains unclear. We analyzed 51,700 UK Biobank participants with a definable clinical-obesity state at baseline and a repeat assessment. Participants were classified as having no obesity, preclinical obesity, or clinical obesity at each assessment, generating eight observed two-timepoint transitions. Follow-up began at the repeat assessment and continued until incident hospital-recorded anxiety disorder, death, or administrative censoring. Cox proportional hazards regression was used to analyze associations separately within each baseline clinical-obesity state. During a median follow-up of 4.83 years, 1,094 incident anxiety disorders occurred. In analyses comparing participants within the same baseline clinical-obesity state, transition from no obesity to clinical obesity was associated with higher anxiety risk than persistent no obesity (adjusted hazard ratio [aHR], 1.87; 95% confidence interval [CI], 1.35–2.60); no other transition was significantly associated. When all eight transition groups were compared with persistent no obesity, transition from preclinical obesity to clinical obesity (aHR, 1.29; 95% CI, 1.03–1.63) and persistent clinical obesity (aHR, 1.46; 95% CI, 1.16–1.84) were also associated with higher risk. Transition from no obesity to clinical obesity was associated with higher incident anxiety in the primary analysis. When all transition groups were compared with persistent no obesity, transition from preclinical obesity to clinical obesity and persistent clinical obesity were also associated with higher risk. These findings suggest that transitions resulting in clinical obesity, or persistence of clinical obesity, may identify individuals at higher risk of incident anxiety disorder.