OBJECTIVE
To study the use of hypertriglyceridemic waist (HTGW) to predict the occurrence of diabetes. Also to independently study whether there was an interaction between HTGW and impaired fasting glucose impaired fasting glucose (IFG) on the cause of diabetes.
METHODS
We undertook a cohort study based on data from the "Prevention of Multiple Metabolic Disorders and Metabolic Syndrome (MS) Study in Jiangsu Province, China". We used the logistic regression model to analyze the relations between both HTGW, IFG and diabetes mellitus and to evaluate the multiplied interaction between HTGW and IFG to include product terms method. Counting additive interaction was carried out under the Excel Calculation Sheet, compiled by Anderson and his colleagues.
RESULTS
After adjusted for general risk factors and baseline fasting plasma glucose (FPG), results showed that subjects with HTGW had a 2.10 (95% CI: 1.36 - 3.25) adjusted relative risk (HR) of developing a diabetes when compared with those individuals without HTGW at the baseline study. When IFG was stratified, participants with HTGW were significantly associated with diabetes, regardless of IFG. The multi-adjusted HRs of diabetes were 3.09 (1.70 - 5.61) and 2.09 (1.08 - 4.02), respectively. Compared to the participants with non-HTGW and their FPG level below the threshold, those having HTGW and their FPG level was above the threshold, had the highest adjusted HR values [12.05 (95% CI: 6.89 - 21.07)]. Data from the additive interaction analysis showed that RERI as 7.00 (95% CI: 0.49 - 13.51), AP as 0.57 (95% CI: 0.32 - 0.82) and SI as 2.66 (95% CI: 1.36 - 5.21).
CONCLUSION
HTGW could predict the occurrence of diabetes, independent from IFG while the presence of HTGW with IFG could have an additive interaction on the cause of diabetes.