The inclusion criterion for subjects was age ranging from 35 years to 85 years old. The exclusion criteria were as follows: type 1 diabetes, recent acute disease, chronic inflammatory disease, infectious disease, and metabolic disease other than prediabetes and diabetes. Prediabetes and diabetes were diagnosed according to the diagnostic criteria[9]. The adult community residents (n = 1164, 584 men and 580 women) were recruited from Haikou City on Hainan Island from March 2011 to September 2011 using a multistage stratified cluster sampling design. The following clinical characteristics and information were recorded for each subject: age, gender, body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), fasting plasma glucose (FPG), and 2-hour plasma glucose(2 h PG) in the oral glucose tolerance test (OGTT). The subjects were assigned into four groups based on blood glucose level: normal fasting glucose (NFG) group (n = 282), impaired fasting glucose (IFG) group (n = 287), impaired glucose tolerance (IGT) group (n = 293), and T2DM group (n = 302). The age composition did not differ by more than 5 years, and the gender composition ratio did not differ by more than 5%. Physical examination and blood biochemical testing were conducted for all subjects. GRK5 rs10886471 (CA)n polymorphism experiments were also performed from October 2011 to March 2013 as follow-up tests. Our study was considered and approved by Hainan medical ehtics committee on January 2011. Our study began after all participants provided written informed consent.

Genomic DNA was extracted from the peripheral blood using a BloodGen Mini kit (CWBiotech, Beijing, China). Microsatellite polymorphism was identified via PCR and sequencing. The primers were designed to amplify the 320 bp region of GRK5 rs10886471. Information on the rs10886471 sequence is available online (http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=10886471#fasta). The forward primer was 5′-aagttcttccctgctagagaa-3′ and the reverse primer was 5′-ctctttttgttctaagtgaaaac-3′. PCR was performed under the following conditions: initial denaturation at 94°C for 5 min; followed by 33 cycles of denaturation at 94°C for 1 min, annealing at 53°C for 1 min, and extension at 72°C for 1 min; and a final extension at 72°C for 7 min. The reaction was performed at a final volume of 50 µl, which contained the basic reaction components. The PCR products were verified via 2.0% agarose gel electrophoresis and purified using a Quick Gel Extraction Kit (CWBiotech, Beijing, China). The purified PCR products were directly sequenced or ligated into a pGEM-T Easy Vector sequence (Shanghai Sangon Biotech Co. Ltd, China). The sequencing results were aligned with the intron region of the GRK5 gene from GenBank (NM_005308.2) and were analyzed using the BioEdit software. Standard procedures and the latest scientific test specifications were strictly followed. Two people independently counted the alleles and discrepancies between the two examiners were resolved through repeat examinations of the samples.

The microsatellite polymorphism was analyzed using the SSRHunter genetic profiler software. The (CA)_n allelic frequencies were estimated through direct gene counting. Polymorphism information content (PIC) was calculated using the PIC-Calc0.6 software. A Pearson's chi-square test was used to count the variables and an ANOVA was used for mean comparisons. Forward stepwise regression was used for multivariate logistic regression analysis to estimate the strength of the associations of GRK5 polymorphism with prediabetes and with T2DM. SPSS v17.0 was used for all statistical analysis. Differences with p values <0.05 were considered statistically significant, and all p values are two tailed.

Table 1 summarizes the clinical characteristics and biochemical results of the subjects. The four groups did not significantly differ in terms of age and gender (P>0.05). However, waist circumference, BMI, SBP, DBP, FPG, and 2 h PG increased with the abnormal increase in blood glucose (IFG and IGT groups) and continued to increase with the blood sugar until it reached T2DM levels. The clinical parameters significantly differed between the four groups (all P<0.05).

CA repeat sequences are abundant in the human genome. Numerous studies have revealed that intronic (CA)_n repeats could play a novel and generally important role in the splicing of enhancers or repressors during gene expression[10], [11], [12], [13]. Therefore, we extended the rs10886471 SNP analysis to study the short tandem repeat (STR) function. Genomic DNA from 1164 subjects was amplified via PCR and sequenced using primers specific for the rs10886471 studied region (about 320 bp) as shown in Figure 1 and 2).

The allelic frequencies are listed in Figure 3 and Table 2, (CA)₁₇ had the highest allelic frequency in each group, followed (CA)₁₆. However, the allelic frequencies of (CA)₁₆ and (CA)₁₇ were lower in the NFG group than in the IFG, IGT, and T2DM groups. The allelic frequencies of (CA)₁₅, (CA)₁₈, and (CA)₁₉ in the NFG group were higher than those in the IFG, IGT, and T2DM groups. The allelic frequency of (CA)₁₆ was significantly lower than that of (CA)₁₇, but significantly higher than those of (CA)₁₈ and (CA)₁₉ among the four groups (χ² = 16.190, P = 0.001; χ² = 10.221, P = 0.017; and χ² = 8.265, P = 0.041, respectively). The allelic frequency distributions of (CA)₁₅, (CA)₁₇, (CA)₁₈, and (CA)₁₉ did not significant differ among the four groups [(CA)₁₅ (χ² = 0.570, P = 0.903); (CA)₁₇ (χ² = 6.096, P = 0.107); (CA)₁₈ (χ² = 1.368, P = 0.713); (CA)₁₉ (χ² = 3.889, P = 0.274), respectively]. By contrast, the frequency of (CA)₁₆ increased with the abnormally increasing blood glucose (IFG and IGT groups) and continued to increase with blood sugar until it reached T2DM levels. The allelic frequencies of (CA)₁₆ in the IFG, IGT, and T2DM groups were much higher than in the NFG group (χ² = 12.300, P = 0.000; χ² = 13.672, P = 0.000; χ² = 14.476, P = 0.000, respectively).

The PIC values of the rs10886471 (CA)_n alleles in the NFG, IFG, IGT, and T2DM groups were 0.6146, 0.6233, 0.6291, and 0.6327, respectively. The PICs of the four groups all exceeded 0.5, which indicates that the GRK5 (CA)_n repeats exhibited genetic polymorphism. The PIC of each group did not significantly deviate from the Hardy–Weinberg equilibrium.

Logistic regression analysis was conducted on the alleles and the results are presented in Table 2. The NFG group was designated as the control group, whereas the three remaining groups were designated as the case groups and classified as dependent variables (NFG = 0, IFG = 1, IGT = 2, and T2DM = 3). Each allele was classified as an independent variable. The statistical significance of the inclusion criteria was set to P>0.05, whereas that for the exclusion criteria was set to P<0.10. Table 2 shows the association of intronic (CA)_n repeat polymorphisms with prediabetes and T2DM risk. Using the most common (CA)₁₇ allele as a reference for estimating the strength of the association, allele (CA)₁₆ was significantly associated with increased risk of prediabetes and T2DM [IFG, OR (95% CI) = 1.938 (1.289–2.915), P = 0.001; IGT, OR (95% CI) = 2.021(1.347–3.034), P = 0.001; T2DM, OR (95% CI)  = 2.012 (1.345–3.009), P = 0.001, respectively]. The other alleles were not significantly associated with abnormal blood glucose (P>0.05). The NFG group was designated as the control group, whereas the other three groups were designated as the case groups. All alleles without (CA)₁₆ that included (CA)₁₅, (CA)₁₇, (CA)₁₈, and (CA)₁₉ were classified as one multiple allele. The (CA)₁₆ allele was used as a reference. The multiple alleles were negatively correlated significantly with the IFG, IGT, and T2DM groups [IFG, OR (95% CI)  = 0.510 (0.347–0.750), P = 0.001; IGT, OR (95% CI)  = 0.492 (0.336–0.721), P = 0.001; T2DM, OR (95% CI)  = 0.484 (0.331–0.708), P = 0.000, respectively].