Longitudinal changes in comorbidity patterns over time in relation to mortality in older Korean adults.

This study investigates the different pattern of change in comorbidity over time and its effect on mortality in older adults in South Korea. Data are from the National Health Insurance Service-Korean Senior (NHIS-Korean Senior) claim database, consisting of a random sample of 10% of adults aged 60 years and older. At baseline in 2002 121,733 subjects, who were admitted to the hospital at least once a year between 2002 and 2008, were analyzed. We used the Cox proportional hazards models after growth mixture modeling (GMM) to estimate trajectories in the combined comorbidity Index scores which is the Gagne's method between 2002 and 2008, and their 1-, 3-, and 5-year mortality rates. Five comorbidity trajectory groups were generated: consistently low, increased, decreased/low, decreased/high, and consistently high. After adjusting for all confounders, compared with decreased/low, the consistently high group presented the highest mortality risk (p < 0.001; Ptrend  < 0.001), with hazard ratios (HR) of 3.48 at 1-year, 2.53 at 3-year, and 1.92 at 5-year follow-up, followed by the increased and decreased/high groups. Five distinct comorbidity trajectories were identified that predicted increased risk of mortality. The group with the consistently high comorbidity scores over time exhibited the highest mortality risk.