Prevalence and Predictors of Insulin Resistance in Non-Obese Healthy Young Females in Qatar

Abstract

The state of Qatar suffers from diabetes epidemic due to obesity-associated metabolic syndrome. However, the prevalence of insulin resistance prior to obesity, which could play an important role in the high prevalence of diabetes, has not yet been described. This study aims to compare the prevalence of insulin resistance in apparently healthy non-obese and obese participants from Qatar and identify the predictors of insulin resistance in different body mass index (BMI)-groups. In this cross-sectional study, 150 young healthy females from Qatar were dichotomized into four groups (underweight, normal weight, overweight and obese) based on their BMI. Anthropometric measures as well as fasting plasma levels of lipids, adipokines, blood glucose and insulin were recorded. The prevalence of insulin resistance as per homeostatic model assessment of insulin resistance (HOMA-IR) was estimated and differences between insulin sensitive and insulin resistant were compared. Linear models were used to identify predictors of insulin resistance in every BMI group. Prevalence of insulin resistance in non-obese healthy females from Qatar ranges between 7% and 37% and increases with BMI. Overall, predictors of insulin resistance in the Qatari population are triglycerides/high-density lipoprotein (HDL) ratio and free fat mass but vary according to the BMI group. The main predictors were triglycerides in normal weight, triglycerides/HDL in overweight and triglycerides/HDL and interleukin-6 (IL-6) in obese individuals. The high prevalence of insulin resistance in non-obese Qataris may partially explain diabetes epidemic. Larger studies are warranted to confirm these findings and identify underlying causes for insulin resistance in non-obese individuals in Qatar, aiming at targeted intervention before diabetes onset.

Keywords: BMI; Qatar; insulin resistance; non-obese; prevalence.

Figure 1

Orthogonal partial least square discriminate analysis (OPLS-DA) model comparing mediators of metabolic disease in IS and IR participants. (A). A score plot showing the class-discriminatory component 1 (x-axis) versus orthogonal component (y-axis). (B). An updated score plot that reveals that the orthogonal component (y-axis) mostly represent BMI groups (underweight, normal weight, overweight and obese). (C). The corresponding loading plot showing mediators of metabolic syndrome at either ends of the discriminatory components along the x-axis and y-axis.

Figure 2

ROC curve for triglycerides, triglycerides/HDL and body fat mass as a predictor for the HOMA-IR index in normal, overweight/obese and all participants.