To assess body composition, nutritional status and its differences between genders in a sample of Portuguese centenarians.

Observational cross-sectional study.

Centenarians recruited in Portugal, able to give informed consent.

A total of 252 subjects, with a median age of 100 years, mostly women (77.8%) who accepted to participate in the study, during the period of 2012 to 2014.

Anthropometric data collected (weight, height, BMI, waist circumference, hip and waist/hip ratio) were evaluated according to WHO criteria. A portable tetrapolar bioimpedance analyzer was used to calculate body composition and to assess resting metabolism. Nutritional status was evaluated according to three different criteria: BMI, waist circumference and body fat percentage using anthropometric equations and bioimpedance.

We observed an overall mean weight of 51.02±11.03Kg, height of 1.55±0.07m and a BMI of 21.07±3.69Kg/m². For most of the evaluated parameters, we found substantial differences between genders. The prevalence of underweight and overweight were 25.3% and 13.3%, respectively. Only 5 subjects were obese. Overweight subjects were mostly men (W=10.6% vs. M=22.6%), whereas women were more underweight (W=28.7% vs. M=13.2%). When considering the waist circumference, 26.5% were above the cut-off value. Most of centenarians (72.9%) had a healthy level of visceral fat. This measurement was highly correlated with waist circumference (r= 0.687, p<0.001). The mean of body fat mass was 10.69±6.50Kg, fat-free mass 40.87±7.60Kg and total body water 27.54±6.25Kg. According to body fat mass criteria assessed by bioimpedance, the prevalence of obesity in study population was 6.0% with no gender differences (p = 0.225). Obesity prevalence using anthropometric equations was higher (Deurenberg: 77.7% and Gallagher: 42.8%) than the obtained value by bioimpedance analysis, although according to Bland-Altman analysis both equations showed a good agreement (Deurenberg: 95.8% and Gallagher: 97%) with bioimpedance method. The prevalence of hypohydration (12.9%) was tendentiously higher in women compared to men (W= 15.4% vs. M= 5.0%, p=0.087). Despite the frequency of osteoporosis was higher in women (W = 71.85% vs. M = 28.15%), 95% of men revealed criteria for osteoporosis. Resting metabolic rate (RMR) was significantly different between genders using bioimpedance analysis (W= 1123.33± 173.91; M= 1350.10± 188.88; p<0.001) or Harris Benedict equation (W= 934.92± 102.60; M= 1018.85± 171.68; p=0.001). Bland- Altman analysis between methods indicate that there was an agreement of 97.6%. The overall mean metabolic age obtained was 83.52±1.11 years, well below the chronologic age (p<0.001).

In Portuguese centenarians, clinical and nutritional approach should be improved on the gender basis. In relation to nutritional status, centenarians were more frequently underweight than overweight. The thinness could be a natural process, contributing for the longevity being rather overweight a reducing factor in life expectancy. BMI and waist circumference showed a good correlation with body fat percentage. Despite the results of Bland- Altman analysis, Deurenberg and Gallagher equations are not suitable to evaluate obesity prevalence in centenarians. Harris Benedict equation seems to be a good option to measure RMR in centenaries, when BIA is not available. Body composition and nutritional characterization of Portuguese centenarians are relevant contribution in scientific evidence production for the action plan of healthy ageing in Europe (2012–2020) and also for clinical practice.