Obesity measures in the Kiribati population: a need to reclassify body mass index cut-points

BMC Public Health. 2020 Jul 11;20(1):1092. doi: 10.1186/s12889-020-09217-z.

Abstract

Background: Obesity is a public health problem in Micronesia. The objective of the study was to assess obesity, the relationship between body mass index (BMI) and body fat percentage (BF%) among adults, and determine the appropriate BMI cut-points in Kiribati.

Methods: A cross-sectional study was undertaken among 483 adults randomly selected from South Tarawa (ST) and Butaritari (BT). Weight, height, BF% and physical activity level (PAL) was measured using standard methods. Linear and quadratic regression analyses were conducted to assess the association between BF% and BMI whilst controlling for age and gender. Receiver operating characteristics (ROC) curve analyses were used to assess whether for the Kiribati population alternative BMI cut-off points for obesity are needed.

Results: Approximately 75% of participants were obese using standard BMI and BF% cut-offs, with the highest prevalence observed in South Tarawa. BF% was significantly (p < 0.001) and positively associated with age (males, r = 0.78; females, r = 0.67; p < 0.001) and BMI. Based on ROC-curve analyses the BMI cut-offs for predicting high BF% among I-Kiribati people were 24.5 kg/m2 for males and 32.9 kg/m2 for females.

Conclusions: In conclusion, the majority of adults in Kiribati were either obese or overweight and had high BF%. We suggest that ethnic-specific BMI cut-points to define obesity for the population of Kiribati may be more appropriate than the currently used international cut-points.

Keywords: Adults; Body fat percent; Body mass index; Kiribati; Predictor; Relationship.

MeSH terms

  • Adipose Tissue*
  • Adult
  • Body Composition
  • Body Mass Index*
  • Body Weight
  • Cross-Sectional Studies
  • Exercise*
  • Female
  • Humans
  • Male
  • Micronesia
  • Middle Aged
  • Obesity / epidemiology*
  • Prevalence
  • Public Health
  • ROC Curve
  • Reference Values
  • Regression Analysis
  • Waist-Height Ratio