To the best of our knowledge, this is a large prospective study of creatinine and diabetes, demonstrating that low serum creatinine is an independent risk factor for new T2DM in Chinese adults with a median follow-up of 3.1 years, compared with high serum creatinine at baseline. Independent of established risk factors for type 2 diabetes, such as BMI, fasting blood glucose (FPG), and a genetic history of diabetes. The relationship was also stable across various stratified analyses.
The only metabolite of phosphocreatine in skeletal muscle is creatinine. Under stable conditions, creatine in skeletal muscle forms creatinine at a relatively constant rate, mainly through non-enzymatic dehydration, and is released into the blood for excretion in urine. Because this blood flesh anhydride and total muscle inside body concern closely, not easy to suffer food effect8,11. Studies have shown that muscle mass is closely related to creatinine (correlation coefficient ≥ 0.7)15,26.Thus, when renal function is stable and protein intake is normal, creatinine may also be an inexpensive and readily available marker of muscle mass compared to more advanced techniques such as computed tomography, magnetic resonance imaging, or dual-energy X-ray absorption to assess muscle mass12.
Low serum creatinine was reported as a risk factor for type 2 diabetes in three small cohort studies of men n = 857016, n = 331317, n = 3134318.The study population was single man, and the results limited extrapolation to the general population. Three large Asian population studies (n = 966719, n = 5758720and n = 4143921participants, also showed an association between serum creatinine levels and the risk of type 2 diabetes in both men and women. In our cohort study, n = 198739, we adjusted for more covariables (established risk factors), including age, sex, BMI, blood pressure, lipids, fasting blood glucose and genetic history of diabetes. We concluded that a more robust low level of creatinine is a risk factor for diabetes.
Current research suggests that low levels of creatinine are a risk factor for diabetes. The range of creatinine in the low-level group currently studied is at baseline: 0.4–0.616, < 0.719, 0.38–0.6917, 0.3–0.78 (27–70 umol/L)20, < 0.718, < 0.9121 mg/dL for men; < 0.5019,0.16–0.57 (15–51 umol/L)20, < 0.6821 mg/dL for women.The low level group in our study was < 0.80(71.8 umol/L) and < 0.58(51.6 umol/L) mg/dL respectively. We had nearly n = 50,000 participants in each group, and after adjusting for variables, the higher risk of diabetes in the lower creatinine group was very stable. We also performed a grouping analysis based on clinical and test thresholds (< 0.80(71.8umol/L) for men and < 0.58(51.6umol/L) mg/dL for women), lower creatinine was associated with a reduced incidence in the population, but adjusted for the corresponding variables, this population was at a higher risk of diabetes.
We also did a detailed stratified analysis. In age stratification at 45 years old, our study shows that low levels of creatinine are associated with a higher risk of diabetes in the young group. Contrary to previous studies, a cohort study of Japanese men18 found that low levels of creatinine are associated with a higher risk of diabetes in the elderly, which may be the reason for our large sample size. It also suggests that younger people may be looking at creatinine levels and muscle mass issues while looking at established risk factors for diabetes. The stratification of interest is that baseline fasting glucose is less than 6.1 mmol/ L and low creatinine is associated with a higher risk of diabetes, which needs to be confirmed by further studies.
Our findings have a biological basis based on existing evidence, although the underlying mechanism of the negative association between serum creatinine and type 2 diabetes remains to be described. Studies have shown that skeletal muscle insulin resistance, oxidative stress and inflammation are considered to be the main mechanism leading to the development of diabetes27,28. Insulin resistance is a characteristic of type 2 diabetes and may be the primary cause of most type 2 diabetes cases. Compared with obesity and lipid metabolism disorders leading to insulin resistance, skeletal muscle insulin resistance is relatively less studied. Insulin resistance refers to the reduced sensitivity of the target organs of insulin action (mainly liver, muscle and adipose tissue)28 to insulin action. Reduced skeletal muscle mass leads to reduced target tissue for insulin action, decreased sensitivity to insulin action, and skeletal muscle insulin resistance, resulting in reduced skeletal muscle glucose uptake and impaired glucose phosphorylation, resulting in impaired skeletal muscle mitochondrial function and metabolic disorders. The accumulation of metabolites (mainly diacylglycerol (DAGs))28 and a series of myoinflammatory factors29,30 (such as interleukin-631) participate in the impaired autophagy function of cells32, and then block insulin signal transduction in skeletal muscle33, and interact with each other in this process, resulting in the occurrence and progression of type 2 diabetes. Skeletal muscle undertakes most of the tasks of glucose metabolism, and its decreased ability of insulin stimulated glucose uptake is of great significance for systemic glucose homeostasis34. Studies have shown that skeletal muscle has impaired glucose uptake and phosphorylation in the early stage of diabetes35, and muscle fiber atrophy in skeletal muscle leads to the occurrence and progression of diabetes36. In addition, studies have shown that resistance exercise can improve skeletal muscle insulin resistance37, which may be used as an intervention to reduce the occurrence and progression of diabetes. Therefore, we speculate that skeletal muscle insulin resistance may be an important factor in the increased risk of new-onset diabetes. In general, future studies need to further study the potential mechanisms of serum creatinine and muscle mass on the occurrence of diabetes.
Our study also has some limitations. Firstly, although a series of covariables were adjusted in the regression model, residual confounding effects from unmeasured or unknown factors could not be excluded. Secondly, our study was a cohort with a large sample size, which could overcome some confounding effects. Second, our current study was conducted among Chinese adults, and further investigation is needed to determine whether the observed results can be extrapolated to other populations. Third, we do not have detailed information on diet, but our participants are generally healthy urban adults who follow the Traditional Chinese diet for three meals a day. Therefore, there is little difference between individuals38, and it is difficult to achieve complete consistency in diet in a large cohort study of the general population. Fourthly, we used the diagnostic criteria for diabetes without oral glucose tolerance test. Some studies have shown that the incidence of diabetes may be lower if fasting glucose is used to diagnose diabetes, however, it is noteworthy that low incidence of diabetes resulting from such errors would bias toward the null, thus, resulting in an underestimation of the association between creatinine and type 2 diabetes. Overall, our results need to be confirmed in future studies.
In conclusion, low serum creatinine is independently correlated with increased risk of incident diabetes in Chinese adults. It was also stable in various subgroups stratified.