Is the Utility of the GLIM Criteria Used to Diagnose Malnutrition Suitable for Bicultural Populations? Findings from Life and Living in Advanced Age Cohort Study in New Zealand (LiLACS NZ).

Objectives: To investigate associations between nutrition risk (determined by SCREEN-II) and malnutrition (diagnosed by the GLIM criteria) with five-year mortality in Māori and non-Māori of advanced age.

Design: A longitudinal cohort study.

Setting: Bay of Plenty and Lakes regions of New Zealand.

Participants: 255 Māori; 400 non-Māori octogenarians.

Measurements: All participants were screened for nutrition risk using the Seniors in the Community: Risk Evaluation for Eating and Nutrition (SCREEN-II). Those at high nutrition risk (SCREEN-II score <49) had the Global Leadership Initiative in Malnutrition (GLIM) criteria applied to diagnose malnutrition or not. Demographic, physical and health characteristics were obtained by trained research nurses using a standardised questionnaire. Five-year mortality was calculated from Government data. The association of nutrition risk (SCREEN-II) and a malnutrition diagnosis (GLIM) with five-year mortality was examined using logistic regression and cox proportional hazard models of increasing complexity.

Results: 56% of Māori and 46% of non-Māori participants had low SCREEN-II scores indicative of nutrition risk. The prevalence of GLIM diagnosed malnutrition was lower for both Māori and non-Māori (15% and 19% of all participants). Approximately one-third of participants (37% Māori and 32% non-Māori) died within the five-year follow-up period. The odds of death for both Māori and non-Māori was significantly lower with greater SCREEN II scores (better nutrition status), (OR (95% CI); 0.58 (0.38, 0.88), P < 0.05 and 0.53 (0.38, 0.75), P < 0.001, respectively). GLIM diagnosed malnutrition was not significantly associated with five-year mortality for Māori (OR (95% CI); 0.88 (0.41, 1.91), P >0.05) but was for non-Māori. This association remained significant after adjustment for other predictors of death (OR (95% CI); 0.50 (0.29, 0.86), P< 0.05). Reduced food intake was the only GLIM criterion predictive of five-year mortality for Māori (HR (95% CI); 10.77 (4.76, 24.38), P <0.001). For non-Māori, both aetiologic and phenotypic GLIM criteria were associated with five-year mortality.

Conclusion: Nutrition risk, but not malnutrition diagnosed by the GLIM criteria was significantly associated with mortality for Māori. Conversely, both nutrition risk and malnutrition were significantly associated with mortality for non-Māori. Appropriate phenotypic criteria for diverse populations are needed within the GLIM framework.