The Deqing Cohort Study On Type 2 Diabetes Of Adults In Rural Communities Of China

[Objective] The objectives of this study are:to understand the distribution and determinants of hyperglycemia; to observe the cumulative incidence of type 2 diabetes mellitus (T2DM) among sub-population; to estimate the risk of T2DM at Fasting blood glucose (FBG) of 5.6-6.0mmol/l; to analyze the association between environmental factors, genetic polymorphisms of ENPP1 and Sirtl genes and T2DM; and to assess the interactions of environmental-gene and gene-gene to the risk of T2DM in a rural adult population in Deqing, China,[Methods] A baseline survey was carried out in 4 rural communities of Deqing County, Zhejiang Province during the period of September 2006 to May 2007. Subjects at the age of 18-64 years were face-to-face interviewed by trained health workers using a structured questionnaire through household visits. Informed consent was received from each participant before the interview. Information on demographics, education and occupation, marriage status, life styles including smoking, drinking, tea consumption, exercises, diet, and health status and health service utilization of subjects were collected from the interview. Physical examinations including height, weight, waistline, hipline, blood pressure, B-ultrasound and EEG were offered to all the participants free of charge. A total of 5 ml blood was taken for the tests of FBG, HBsAg, etc. The rest of blood specimen was stored in -20℃for further uses. Subjects were considered to have impaired fasting glucose (IFG) if FBG ranges at 5.6mmol/l-6.9mmol/l, and to have diabetes mellitus (DM) if FBG is 7.0 mmol/l or more or receiving antidiabetics. Based on this cross-sectional study, a population cohort was established. Subjects of the cohort were followed up during May 2008 to August 2009 with the similar procedure of the baseline study. Nested in this cohort, a community-based case-control study for T2DM was conducted. The matching ratio of case to controls were 1 to 3. The eligiblity for T2DM cases were Han patients who were newly diagnosed between September 2006 and August 2009 according to the critia of WHO 1999; while controls were randomly sampled from those subjects having norm-glucose in the same community, at similar age (±2 years old), same gender, Han, and bing not a relative to the cases. Polymerase Chain Reaction (PCR) and Restriction Fragment Length Polymorphism (RFLP) methods were used to detect polymorphisms of ENPP1 K121Q and Sirt1 rs3758391 SNP. All data were double entered into a database using Epidata 3.1, and analyzed through SPSS 16.0.χ2 test was used for noncontinuours varibles, Student's t-test or ANOVA for continuous varibles. Multinomial logistic regression and Cox regression models were employed to analyse the associations between environmental factors, genetic polymorphisms, and the risk of T2DM, as well as the possible interactions of gene-environment and gene-gene, which were measured by odds ratio (OR) or hazard ratio (HR), adjusted OR (aOR) or HR (aHR) and 95% confidence interval (95%CI). Gender specific standardised incidence and prevalence were calculated using China 2000 Census Population.[Results] Totally,5898 persons aged from 18 to 64 years were recuited and eligible into this study after the completement of the questionnaire. Among them,4,506 subjects were tested for FBG (76.2%) with the average age of 46.1±10.0 years old and the mean FBG of 4.7±0.9 mmol/l. Of 4506 subjects, men were 45.1%, the crude prevalences of IFG and DM were 5.4% and 2.2%(the standardized prevalences were 4.2% and 2.1%, respectively), and they were 4.8% and 2.5% for men,5.8% and 2.0% for women, respectively. The prevalence of hyperglycemia increased with age regardless of gender (χ2trend=43.18, p<0.001). The average ratio of IFG/DM was approximately 2.5, and such ratio was even high as 5.0 for those under age of 35 years.The results of the baseline study showed:among 4506 subjects with the age of 18-64 years, compared to subjects with medium household income, those in the high level of household income had significantly higher risks of IFG (aOR=1.80,95%CI:1.15-2.82) and those in the low level of household income had significantly higher risks of DM (aOR=1.85,95% CI:1.01,3.38); the prevalence of hypertension was 24.8% (the standardized prevalence was 17.8%), only 21.6% of hypertensive patients was aware, and hypertension was associated with IFG and DM (aOR=1.50,95%CI:1.10-2.07 and aOR=1.74,95%CI:1.09-2.79, respectively) after the adjustment of age, gender, smoking, drinking, BMI and regular exercises; the prevalence of anemia was 51.6% (the standardized prevalence was 43.5%), and non anemia was associated positively with DM (OR=1.8,95%CI:1.15-2.79) but, did not reach the significance after the adjustment of above possible confounders; the prevalence of snoring was 33.9%(the standardized prevalence was 53.8%), and snoring was associated positively with DM (OR=1.6,95%CI:1.04-2.47), but did not reach the significance after the adjustment of above possible confounders; the prevalence of HBsAg was 7.3%(the standardized prevalence was 8.0%), and HBsAg was associated positively with IFG (OR=1.7, 95%CI:0.92-3.16), but did not reach the significance after the adjustment of above possible confounders; the prevalence of tea drinking was 58.4%, and it was associated inversely with DM (aOR=0.52,95%CI:0.28-0.98) after the adjustment of above possible confounders; the prevalence of smoking was 30.5%(the standardized prevalence was 31.5%), and smoking was associated positively with IFG (OR=1.23, 95%CI:0.66-2.29), but did not reach the significance after the adjustment of above possible confounders; the prevalence of alcohol drinking was 22.2%(the standardized prevalence was 21.8%), which was not associated significantly with IFG or DM after the adjustment of above possible confounders; the average of BMI was 22.7±11.6 kg/m2, and subjects of BMI>=24 kg/m2 had aORs:1.8(95%CI:1.34,2.37) and 3.7(95%CI:2.42,5.65) for IFG and DM, respectively; in recent a year before the interview, the prevalence of illness and visiting doctor in 2 weeks was 6.5% and 2.3%, respectively, and subjects had an average cost of 1043.65±4753.54 Yuan and 9404.37±17195.01 Yuan per visiting and per hospitalization respectively as well as one tenth of their incoming per year was paid for illness, and DM patients had higher those measures.During May 2008 to June 2009,57.2%(1908) of targeted 3337 subjects excluded DM at baseline were followed up with the average of 1.92±0.34 person years varied from 1.24 to 2.85 person-years and the total 3654.75 person-years. The followings were found:the 2-year cumulative incidence (CI) of DM was 1.9%(the standardized incidence was 1.1%), and it was 1.8% for men and 1.9% for women, respectively; the 2-year CI of IFG among the subjects with FBG<5.6mmol/l at baseline was 4.7%(the standardized incidence was 3.6%), and it was 3.9% for men and 5.2% for women, respectively; and the incident density (ID) of DM was 9.8/1000 person years, and it was 9.2/1000 person years for men and 10.1/1000 person years for women, respectively, which suggested that the incident rate of DM was higher in rural Deqing.The findings from this prospective cohort demonstrated:after the adjustment of age, gender, education level, occupation, hypertension, smoking, alcohol drinking, regular exercise and FBG at baseline (not including the variable for analysis), among the adults of rural Deqing, subjects with IFG at baseline had higher risk of DM than those without IFG (77.7/1000 person years vs.5.4/1000 person years, aHR-14.26,95%CI: 5.87-34.63), and the risk of DM increased 7 times by a raise of lmmol/l (aHR=7.85, 95%CI:4.40-14.01); subjects with FBG of 5.6-6.0mmol/l had the ID of 39.5/1000 person years, and compared to those with FBG<5.6mmol/l, they had 6.6 times in the risk of developing DM (aHR=6.59,95%CI:2.08-20.82); with the comparison of those without hypertension at baseline, subjects with hypertension had a higher ID and risk of DM (21.9/1000 person years vs.4.9/1000 person years, aHR=3.92,95%CI: 1.61-9.62), and the risk of DM increased 3%-5% by an elevated 1mmHg, but significance disappeared after the adjustment; subjects of overweight/obesity at baseline had a higher ID of DM than those with normal BMI (17.2/1000 person years vs.7.3/1000 person years, aHR=2.20,95%CI:1.11-4.38), and after adjustment of age and gender and more possible confounders, there was a increase of 14%(aHR=1.14, 95%CI:1.04-1.25) and 5% (aHR=1.05,95%CI:0.94-1.17) in risk of DM by 1kg/m2 more, respectively; subjects receiving more education years had higher risk of DM than those receiving less (10.8/1000 person years vs.9.3/1000 person years, aHR=5.08,95%CI:1.82-14.20) and the risk of subjects with lower self-reported family economic level was 3.6 times as that of those with self-reported average level (aHR=3.56,95%CI:1.36-9.32); the traditional Chinese diet pattern had lower risk of DM than those at family eating a meatless diet (7.8/1000 person years vs.14.3/1000 person years, aHR=0.31,95%CI:0.12-0.83) and the similar protective effects of personal diet habit was observed for DM (8.9/1000 person years vs.9.8/1000 person years, aHR=0.33,95%CI:0.13-0.80); smokers had a no significantly higher ID of DM than non-smokers (11.1/1000 person years vs.9.3/1000 person years), and similar findings was obtained among subjects with or without alcohol drinking or with the more categories of alcohol consumptions; the ID of subjects drinking tea everyday was higher than that of those not drinking tea everyday (108.2/1000 person years vs. 7.5/1000 person years) and no significant association was found between tea drinking and incident DM; also, there were lack of evidences in the effects of snoring, anemia and HBsAg on incident DM.The results from 1:3 matched community-based case-control study showed that there was a marginal association between ENPP1 K121Q and type 2 diabetes mellitus (T2DM), i.e. subjects with the genotype of QQ had an aOR of 6.07 (95%CI: 0.81-45.28, p=0.079) with the comparison of those genotyped KK, and the multiple interaction on DM between smoking and KQ/QQ genotype (ORint=4.04,95%CI: 1.12-14.59, p= 0.033 and aORint=4.12,95%CI:1.07-15.91, p= 0.040 with the adjustment of age, sex, BMI and alcohol drinking) was found. Further with the stratification of smoking, the association between KQ/QQ genotype and DM was found in smokers (OR=3.26,95%CI:0.93-11.46, p=0.066) and not in nonsmokers (OR=0.83,95%CI:0.43-1.59), which suggested that more study on such interaction need be called. No significant association between Sirtl rs3758391 SNP and DM and no multiple gene-gene interaction on DM between Sirtl rs3758391 and ENPP1 K121Q were found.[Conclusions] The prevalence and incident rate of DM and IFG were relatively high in the population of 18-64 years old in Deqing in the context of Chinese rural population. A serious situation of exposure to possible risk factors of diabets such as hypertension, smoking, alcohol drinking arid lack of regular exercise has been identified in this population. IFG, hypertension, and overweight/obesity should be considered as independent risk factors to T2DM. It has also been found that people with better educated, lower household income have increased the risk of T2DM, and the traditional Chinese diet is probably a protective factor to the occurrurance of T2DM. The cutoff value of IFG at 5.6mmol/l is ideal in predictive high risk population for the prevention of T2DM in rural China. ENPP1 K121Q was marginally associated with T2DM and there might be the effect modification from smoking to the risk of T2DM in rural Chinese population, and further studies are required to get more evidences.