Abstract
Background and rationale: Diabetes mellitus (DM) is the major public health burden in Thailand. DM is the significant risk factors of cardiovascular disease, end stage renal disease and cancer. Therefore, to decrease morbidity and mortality from diabetic complications, prevention of DM is an important issue. Insulin resistance and pancreatic β-cell dysfuction are the major mechanisms of developing DM. These 2 mechanisms lead to abnormal glucose regulation, called prediabetes that comprises of impaired fasting glucose or impaired glucose tolerance. Previous evidences found that prediabetes is a significant risk factor of DM. Around 60% of diabetic patients had prediabetes 5 years before developing DM. Until now, there has been no study that investigate about the progression rate to DM in Thai prediabetes patients. In addition, currently, there is new proposed risk factors of DM such as serum uric acid and sleep factors. Therefore, this study aims to estimate the progression rate to DM and assess the association between sleep factors, serum uric acid, and HbA1C level. Methods: This study was retrospective cohort study. Prediabetes patients who visited the outpatient clinic at Department of Family Medicine, Ramathibodi hospital during September 2015 to March 2018 were recruited to this study. The independent variables was divided into 7 categories, including 1) demographic data (i.e. age, sex, education level, family history of diabetes) 2) underlying diseases (i.e. hypertension, dyslipidemia) 3) health risk behavior (i.e. smoking, alcohol drinking) 4) sleep factors (i.e. sleep duration, sleep quality, risk of obstructive sleep apnea (OSA), circadian factors) 5) depressive symptoms 6) physical examination (i.e. body weight, height, waist circumference) 7) laboratory (i.e. fasting plasma glucose (FPG), HbA1C, serum uric acid). Outcome of interest was incidence of DM defined according to criteria of American Diabetes Association 2014. Progression rates of developing DM at 24-, 48-, 72-, 96-, 120- months after diagnosis of prediabetes were calculated using Kaplan-Meier method. Multivariate Cox regression analysis was applied to explore the independent relationship between factors and risk of DM. Association between sleep factors and HbA1C was estimated using multivariate linear regression model. Direct and indirect effects of serum uric acid on HbA1c level were investigated using mediation analysis. Results: Progression rates to DM 2, 4, 6, 8, and 10 years after diagnosis of prediabetes were 4.12% (95% CI: 3.44, 4.94), 9.11% (95% CI: 8.03, 10.33), 14.71% (95% CI: 13.22, 16.35), 21.98% (20.02, 24.10) and 27.99% (95% CI: 25.57, 30.58), respectively. Participants with FPG = 110-125 mg/dL had a significant higher progression rate to DM than participants who had FPG = 100-109 mg/dL (chi-square from Log-rank test = 43.82, P-value < 0.001). Results from multivariate Cox regression analysis show that current smokers, obesity defined by body mass index ≥27.5 kg/m2 and FPG = 110-125 mg/dL were independently associated with risk of developing diabetes. Hazard ratios of current smokers, obesity and FPG = 110-125 mg/dL were 1.65 (95% CI: 1.08, 2.52), 1.79 (95% CI: 1.38, 2.33) and 1.86 (95% CI: 1.56, 2.23), respectively. Results from multivariate linear regression analysis found that later chronotype measured by mid sleep time on free day was significantly associated with HbA1C level (β-coefficient =0.014 [95% CI: 0.0002, 0.029, P-value =0.047]). In addition, patients with higher modified PSQI score had significantly lower HbA1C level than patients with lower modified PSQI score (β-coefficient = -0.011 [95% CI: -0.019, -0.003, P-value =0.010]). However, sleep duration were not significantly associate with HbA1C level. Serum uric acid had the direct effect on HbA1c level. Increased 1 mg/dL of serum uric acid was significantly associated with 0.02% (95% CI: 0.008, 0.033) increased HbA1c level. In addition, serum uric acid also had the indirect effect on HbA1c via waist circumference.
