Comparison of body mass index range criteria and their association with cognition, functioning and depression: a cross-sectional study in Mexican older adults

Background: World population is living longer, demanding adjustments in public health policies. Body mass index (BMI) is widely known and used as a parameter and predictor of health status although an adapted criterion for older adults is usually overlooked. BMI has been extensively analysed in relation to mortality but fewer studies address its association with cognition, functioning and depression in older adults. The present study aimed at 1) comparing BMI distribution according to the ranges proposed by the World Health Organization (WHO) and the United States National Research Council Committee on Diet and Health (CDH), 2) analysing their association with cognitive functioning, physical functioning and depression and 3) analysing a possible, interaction of BMI criteria with sex on the outcome measures. Methods: This cross-sectional study included 395 participants recruited by convenience sampling; 283 (71.6%) women and 112 (24.58%) men. Mean age was 74.68 (SD = 8.50, range: 60-98). Outcome measures included the Short Portable Mental State Questionnaire for cognitive status, the Barthel's Index of Activities of Daily Living for physical functioning, and the Geriatric Depression Scale. Results: WHO criterion classified most cases (65.3%) as overweight, followed by normal weight (32.2%) and underweight (2.5%) whereas CDH criterion considered most (48.1%) as normal weight, and followed by overweight (31.4%) and underweight (20.5%). Analysing cognitive status, independent physical functioning and depression mean scores, significant differences (p = .001) were found when comparing the three weight groups (underweight, normal weight and overweight) using either the WHO- or the CDH criterion. Post-hoc tests revealed that in all comparisons the underweight group scored the lowest in all three outcome measures. According to the CDH criterion, overweight was favourable for females but unfavourable for males regarding cognitive status (interaction F(2,389) = 4.52, p = .01) and independent functioning (interaction F(2,389) = 3.86, p = .05). Conclusions: BMI and its associations to relevant outcome measures in the older adults must rely on criteria that take into account the particular features of this population, such as the CDH criterion. Underweight was associated with decremented cognition, less independent physical functioning and more depression. Overweight seemed favourable for women but unfavourable for men.