Specific multimorbidity patterns modify the impact of an exercise intervention in older hospitalized adults.

Background: Different multimorbidity patterns present with different prognoses, but it is unknown to what extent they may influence the effectiveness of an individualized multicomponent exercise program offered to hospitalized older adults.

Methods: This study is a secondary analysis of a randomized controlled trial conducted in the Department of Geriatric Medicine of a tertiary hospital. In addition to the standard care, an exercise-training multicomponent program was delivered to the intervention group during the acute hospitalization period. Multimorbidity patterns were determined through fuzzy c-means cluster analysis, over 38 chronic diseases. Functional, cognitive and affective outcomes were considered.

Results: Three hundred and six patients were included in the analyses (154 control; 152 intervention), with a mean age of 87.2 years, and 58.5% being female. Four patterns of multimorbidity were identified: (26.8%); (20.6%); (16%); and an pattern (36.6%). The Short Physical Performance Battery (SPPB) test improved across all patterns, but the intervention was most effective for patients in the pattern (2.48-point difference between intervention/control groups, 95% CI 1.60-3.35). Regarding the Barthel Index and the Mini Mental State Examination (MMSE), the differences were significant for all multimorbidity patterns, except for the pattern. Differences concerning quality of life were especially high for the p pattern (16.9-point difference between intervention/control groups, 95% CI 4.04, 29.7).

Conclusions: Patients in all the analyzed multimorbidity patterns improved with this tailored program, but the improvement was highest for those in the metabolic pattern. Understanding how different chronic disease combinations are associated with specific functional and cognitive responses to a multicomponent exercise intervention may allow further tailoring such interventions to older patients' clinical profile.