Cardiometabolic disease trajectories and falls risk in older adults: a cohort study using electronic health records

Introduction

Cardiometabolic diseases accumulate with age. Falls are a leading cause of injury, hospitalisation and loss of independence in older adults, and individual cardiometabolic conditions are known to increase falls risk. However, it is unclear how the sequence in which cardiometabolic conditions develop influences subsequent falls risk. We aimed to identify common cardiometabolic disease trajectories and estimate their association with incident falls.

Method

We used UK primary care data (Clinical Practice Research Datalink), linked to hospital admissions data (HES) in a cohort study of older adults aged ≥60 years. Incident falls were defined using diagnosis codes in the primary care and hospital data.  Sequential pattern mining identified common cardiometabolic disease sequences. Time-updated Cox proportional hazards models estimated associations of these sequences with incident falls, adjusting for age and sex.  Associations were evaluated by age, frailty, non-cardiometabolic multimorbidity, and polypharmacy.

Results

Among 5,709,756 older adults, sequential pattern mining identified 63 cardiometabolic trajectories occurring in >1% of the cohort. Hypertension was the most common first condition and the most frequent transitions were to chronic kidney disease, type 2 diabetes, atrial fibrillation and coronary heart disease. Falls risk varied substantially across cardiometabolic disease sequences. Among the highest risk sequences were type 2 diabetes→stroke (hazard ratio [HR] 2.87 [95% CI 2.61–3.15]), type 2 diabetes→heart failure (HR 2.62 [2.23–3.08]) and coronary heart disease→atrial fibrillation→heart failure (HR 2.44 [2.17–2.75]), compared with no cardiometabolic conditions. Associations also differed by age, frailty, non-cardiometabolic multimorbidity and polypharmacy.

Conclusion 

Cardiometabolic disease accumulation is associated with increased falls risk in older adults, with risk varying substantially between disease trajectories. Identifying high-risk trajectories may help inform falls risk stratification and support more targeted and timely fall prevention strategies. Future work will examine how falls risk changes by time since disease transitions to identify potential critical intervention windows.