Abstract
BACKGROUND: Little is known about the heterogeneous clinical profile of physical frailty and its association with cognitive impairment in older U.S. nursing home (NH) residents. METHODS: Minimum Data Set 3.0 at admission was used to identify older adults newly-admitted to nursing homes with life expectancy ≥6 months and length of stay ≥100 days (n = 871,801). Latent class analysis was used to identify physical frailty subgroups, using FRAIL-NH items as indicators. The association between the identified physical frailty subgroups and cognitive impairment (measured by Brief Interview for Mental Status/Cognitive Performance Scale: none/mild; moderate; severe), adjusting for demographic and clinical characteristics, was estimated by multinomial logistic regression and presented in adjusted odds ratios (aOR) and 95% confidence intervals (CIs). RESULTS: In older nursing home residents at admission, three physical frailty subgroups were identified: "mild physical frailty" (prevalence: 7.6%), "moderate physical frailty" (44.5%) and "severe physical frailty" (47.9%). Those in "moderate physical frailty" or "severe physical frailty" had high probabilities of needing assistance in transferring between locations and inability to walk in a room. Residents in "severe physical frailty" also had greater probability of bowel incontinence. Compared to those with none/mild cognitive impairment, older residents with moderate or severe impairment had slightly higher odds of belonging to "moderate physical frailty" [aOR (95%CI)(moderate cognitive impairment): 1.01 (0.99-1.03); aOR (95%CI)(severe cognitive impairment): 1.03 (1.01-1.05)] and much higher odds to the "severe physical frailty" subgroup [aOR (95%CI)(moderate cognitive impairment): 2.41 (2.35-2.47); aOR (95%CI)(severe cognitive impairment): 5.74 (5.58-5.90)]. CONCLUSIONS: Findings indicate the heterogeneous presentations of physical frailty in older nursing home residents and additional evidence on the interrelationship between physical frailty and cognitive impairment.