Reduced levels of serum creatinine were significantly associated with an increased risk of T2DM in men with creatinine below 1.20 mg/dl even after adjustment for age, BMI, SBP, DBP, and fasting plasma glucose. The highest category of serum creatinine levels (serum creatinine ≥ 1.10 mg/dl) was significantly associated with an increased risk of T2DM in women; however, this association disappeared after adjustment for age, BMI, SBP, DBP, and fasting plasma glucose.
The results for men with creatinine below 1.20 mg/dl in our study were consistent with those of previous studies in Japanese men. A study including 31,343 Japanese men without diabetes with a median observation of 7.7 years showed that low cumulative average serum creatinine levels were associated with an increased risk of diabetes after adjusting for age, smoking, BMI, hypertension, and dyslipidemia [14]. Among 8,570 Japanese men aged 40-55 years at entry who had fasting plasma glucose levels < 126 mg/dl and serum creatinine levels < 2.0 mg/dl during the 4-year follow-up period, low serum creatinine was associated with an increased risk of T2DM [15]. Because we did not exclude those with comorbidities such as cardiovascular disease or cancer at baseline or those older than 65 years, other confounding factors may affect the results regarding the association between serum creatinine levels and the risk of T2DM in men with serum creatinine above 1.2 mg/dl.
Our study showed that although there was a trend that reduced levels of serum creatinine were associated with an increased risk of T2DM among women with serum creatinine below 1.1 mg/dl, the association between the level of serum creatinine and the risk of incident diabetes was not significant in women. However, there were several studies showing that the inverse association between the level of serum creatinine and the risk of incident diabetes was consistent for both sexes. In a previous study including 9,667 Japanese individuals without diabetes or hypertension and with normal creatinine levels at baseline during the follow-up period (mean: 5.6 years), low serum creatinine levels independently predicted T2DM development in both men and women [16]. A Chinese cohort study including 41,439 participants (44.5% of those were women) who were ≥ 18 years (range 18-96) and did not have T2DM found that low serum creatinine levels were associated with an increased risk of T2DM after the exclusion of cardiovascular disease, cancer, and abnormally high serum creatinine levels (> 1.2 mg/dL for men and > 1.0 for women) for both men and women [17]. Because we included women with mildly elevated levels of serum creatinine (1.0-1.4 mg/dl) and did not exclude those with comorbidities such as hypertension, cardiovascular disease, or cancer, other confounding factors affecting serum creatinine may influence the results regarding the association between serum creatinine levels and the risk of incident diabetes in women.
The results regarding the association between the levels of serum creatinine and the risk of incident diabetes were different between men and women in our study. Additionally, a previous study in Korean subjects demonstrated that serum creatinine was more closely associated with the risk of T2DM in men than in women [11]. Because total muscle mass is known to be different by sex, the difference in muscle mass may affect the different results regarding the association between serum creatinine levels and the risk of T2DM between men and women. Women were reported to have lower skeletal muscle mass than men. The mean value of serum creatinine was reported to be higher in men than in women in a previous study [18].
The mechanism of the association between serum creatinine levels and the risk of incident diabetes is not clear. Several studies have demonstrated the close association between low muscle mass and dysglycemia. Among Korean subjects aged 65 or older, insulin resistance was higher in the obese group with relatively low muscle mass than in the obese group without low muscle mass [19]. Additionally, hyperinsulinemia, a compensatory response to maintain plasma glucose levels within normal ranges as an early predictor of insulin resistance, was significantly associated with loss of skeletal muscle mass in a cohort study of individuals without diabetes at the 4.6-year follow-up [20]. Increased muscle mass was associated with reduced insulin resistance and a decreased risk of diabetes [21]. The improvement of the amount of lean mass with nutritional supplements was associated with increased insulin sensitivity in elderly subjects with low muscle mass [22]. Insulin receptors in the muscle are known to play a key role in the regulation of glucose metabolism. Because skeletal muscle is the major site of insulin-mediated glucose uptake in the postprandial phase, the defect of skeletal muscle insulin resistance was suggested be the pathogenesis of the development of type 2 diabetes [23]. Increased total lean mass was associated with a decreased risk of incident diabetes for older normal-weight women [24]. Since myokines released by muscle fibers were reported to have systemic effects on the liver, adipose tissue, and pancreas function, lack of myokines such as interleukin-6 and myostatin due to reduced muscle mass may influence insulin resistance [25]. Additionally, insulin sensitizer medication use (metformin and/or thiazolidinediones) may attenuate muscle loss in men with impaired fasting glucose and diabetes [26]. Because insulin is a potent anabolic stimulus for skeletal muscle, it is possible that defects in insulin signaling can induce a reduction in muscle synthesis [27]. Additionally, glomerular hyperfiltration observed early in the natural history in patients with diabetes [28] could contribute to the association between low serum creatinine and the risk of incident diabetes mellitus. Further studies are needed to clarify the mechanism between the close association between serum creatinine levels and the incidence of T2DM.
Our study has several strengths. We used a larger data cohort than a previous study in the Korean population. Because the participants were recruited for health checkups at the national scale, it is reasonable to generalize the results of our study in Korea. Additionally, as women have less muscle mass than men, we analyzed subjects by sex separately.
There were some limitations. Although the measurement of serum creatinine was reliable, serum creatinine was measured in different laboratories in Korea. Other confounding factors, such as dietary factors, physical activity, family history of diabetes or comorbidities that may affect the amount of muscle mass and the development of incident diabetes, were not adjusted for. Furthermore, we used serum creatinine at baseline only, and we cannot assess the relationship between the change in serum creatinine and the risk of diabetes during the follow-up period. Because our study is an observational study, it was difficult to clarify the mechanism of the relationship between serum creatinine and incident diabetes. We did not classify diabetes as type 1 or type 2 diabetes because we did not check β-cell function or islet cell autoantibodies. Asian individuals have more fat with less skeletal muscle than other ethnic groups, including European and Pacific Island adults [29,30]. Furthermore, the risk of diabetes tended to be higher among Asian participants than among Caucasian subjects for the same categories of BMI [31]. Ethnic differences in body composition may contribute to the differences in the association between serum creatinine and the risk of incident diabetes, and it is difficult to generalize the results of our study in other ethnicities.