PD979 - ANOREXIA OF AGING IN NURSING HOME RESIDENTS: PREVALENCE, PREDICTORS, AND IMPACT ON MORTALITY

PD979

ANOREXIA OF AGING IN NURSING HOME RESIDENTS: PREVALENCE, PREDICTORS, AND IMPACT ON MORTALITY

E. ROTHENBERG^1,*, J. Bauer², L. Lindhagen³, J. Holm⁴, A. Modica⁵, T. Cederholm⁶

¹Health Science, Kristianstad University, Kristianstad, Sweden, ²2Centre for Geriatric Medicine, Heidelberg University, Heidelberg, Germany, ³Uppsala Clinical Research Center, Uppsala, ⁴Pfizer AB, Stockholm, Sweden, ⁵Medical Affairs, Pfizer AB, Stockholm, ⁶Department of Public Health and Caring Sciences, University of Uppsala, Uppsala, Sweden

Rationale: Anorexia of aging (AA) is defined by an age-related reduction in appetite. The aim of this longitudinal observational study was to examine the prevalence, predictors, and outcomes of AA among nursing home (NH) residents using data from the Swedish Senior Alert (SA) quality improvement registry.

Methods: NH residents aged ≥65 years were indexed from the SA at first Mini Nutritional Assessment-Short Form (MNA-SF) completion between June 2014 and December 2019. Residents with severe oral health issues were excluded. AA was estimated from MNA-SF responses. Residents with no decrease in food intake were not considered to have AA. Residents with decreased food intake were considered to have AA of increasing severity with greater recent weight loss. Follow-up continued until the end of the study, death, or censoring. Comorbidity-adjusted Cox models were used to evaluate the predictors of new AA and risk of mortality during follow-up in residents with and without dementia at index.

Results: Of the 132,127 NH residents in the study, 42,995 had dementia at index and 89,132 did not. Across groups, the annual prevalence of AA in those with a completed MNA-SF was 22% to 27%. Significant predictors of new AA during follow-up included older age (per 10 years older with dementia; hazard ratio [95% CI]: 1.17 [1.13, 1.20] and without dementia: 1.29 [1.26, 1.33]) and living in a small town/rural setting (1.28 [1.20, 1.36] and 1.37 [1.30, 1.44]). During follow-up, the risk of all-cause mortality was significantly higher in people who had AA at index, and the risk increased stepwise with the severity of AA.

Conclusion: AA was observed in around one in four NH residents and was associated with a higher risk of mortality. Identification of AA and providing effective intervention is key to improving the health of older people.

Disclosure of Interest: E. ROTHENBERG Grant / Research Support from: The study is financed by Pfizer AB, J. Bauer Grant / Research Support from: The study is financed by Pfizer, L. Lindhagen Grant / Research Support from: The study is financed by Pfizer, J. Holm Grant / Research Support from: The study is financed by Pfizer AB, A. Modica Grant / Research Support from: The study is financed by Pfizer AB, T. Cederholm Grant / Research Support from: The study is financed by Pfizer AB