Rate and Timing of Long-Term Condition Accumulation Among Adults with Multiple Long-Term Conditions in England

• Using primary care data on 7.2M adults in England, we examined MLTC accumulation. • Time between LTC diagnoses fell with condition count, from 32.6 months to 5.1. • Most additional conditions occurred within three years of the second LTC diagnosis. • ML identified age, female sex, deprivation, Asian ethnicity as MLTC predictors. Multiple long-term conditions (MLTCs), defined as two or more long-term conditions (LTCs), represent a growing public health challenge. However, the rate and timing at which additional conditions are acquired after MLTC onset remain poorly understood. It is unclear when individuals are at greatest risk of further accumulation or which sociodemographic groups experience faster progression. We aimed to quantify the rate and timing of MLTC accumulation in England and assess sociodemographic influences following MLTC onset. We conducted a retrospective cohort study using Clinical Practice Research Datalink (CPRD) Gold and Aurum data from 1987 to 2020. Adults aged ≥18 years with at least two LTCs were followed longitudinally through successive diagnoses up to their tenth condition. Median transition times between diagnoses were estimated to characterise accumulation patterns. To assess sociodemographic influences on MLTC accumulation, we applied random forest regression with one-hot encoded sociodemographic and clinical features to predict the number of LTCs acquired within 10 years after the second diagnosis. Feature importance metrics and R² were used to evaluate predictors. Among 7,286,481 adults, the median time between diagnoses declined from 32.6 months between the second and third condition to 5.1 months between the ninth and tenth. Most additional conditions were diagnosed within three years following the second LTC. In the random forest model, older age was the strongest predictor of faster accumulation, followed by cardiovascular and mental health conditions at baseline, female sex, greater deprivation, and Asian ethnicity. The model explained 32.4% of the variance in 10-year LTC accumulation (R² = 0.324). MLTC accumulation accelerated after the second diagnosis, identifying an early period of increased risk. Random forest analyses showed that sociodemographic and clinical factors were associated with faster accumulation, although predictive performance was moderate. These findings highlight a potential window for targeted intervention following MLTC onset.