Skip to main content
Download PDF

The association between creatinine to body weight ratio and the risk of progression to diabetes from pre-diabetes: a 5-year cohort study in Chinese adults

BMC Endocrine Disorders volume 23, Article number: 266 (2023) Cite this article

A Correction to this article was published on 02 January 2024

This article has been updated

Abstract

Objective

Evidence on the association between the creatinine to body weight (Cre/BW) ratio and the risk of pre-diabetes to diabetes development remains limited. Our study aimed to examine the association between the Cre/BW ratio and incident diabetes in pre-diabetic patients.

Methods

This retrospective cohort study included 24,506 pre-diabetic participants who underwent health checks from 2010 to 2016 in China. We used the Cox proportional-hazards regression model to explore the relationship between baseline Cre/BW ratio and diabetes risk in pre-diabetes patients. Using a Cox proportional hazards regression with cubic spline function and smooth curve fitting (cubical spline smoothing), we were able to determine the non-linear relationship between them. We also carried out a number of subgroup and sensitivity analyses.

Results

The age range of the participants included in this study was 20–99 years, with a majority of 16,232 individuals (66.24%) being men. The mean baseline Cre/BW ratio was 1.06 (SD 0.22) umol/L/kg. 2512 (10.25%) participants received a diabetes final diagnosis over a median follow-up period of 2.89 years. After adjusting for covariates, the Cre/BW ratio had a negative association with incident diabetes in participants with pre-diabetes, per umol/L/kg increase in Cre/BM ratio was accompanied by a 55.5% decrease in diabetes risk (HR = 0.445, 95%CI 0.361 to 0.548). The Cre/BW ratio and risk of diabetes had a non-linear connection, with 1.072 umol/L/kg serving as the ratio's inflection point. The HR were 0.294 (95%CI:0.208–0.414) and 0.712 (95%CI:0.492–1.029), respectively, on the left and right sides of the inflection point. The sensitivity analysis demonstrated the robustness of these results. Subgroup analyses indicated that the Cre/BW ratio was strongly associated with the risk of diabetes among participants who were younger than 50 years, as well as among those with diastolic blood pressure (DBP) < 90 mmHg and triglyceride (TG) < 1.7 mmol/L. In contrast, among participants 50 years of age or older, those with DBP ≥ 90 mmHg, and those with TG ≥ 1.7 mmol/L, the relationship between the Cre/BW ratio and the risk of diabetes was attenuated.

Conclusion

This study demonstrates a negative, non-linear relationship between the Cre/BW ratio and the risk of diabetes among the Chinese population with pre-diabetes. From a therapeutic standpoint, it is clinically meaningful to maintain the Cre/BW ratio levels above the inflection point of 1.072 umol/L/kg.

Peer Review reports

Introduction

Diabetes mellitus (DM) is a worldwide epidemic that has been linked to significant financial expenditures. China has the most considerable percentage of diabetic patients worldwide as of 2013. By 2035, it is expected to have the greatest impact [1]. According to the International Diabetes Federation (IDF), the global prevalence of diabetes among 20–79-year-olds was expected to be 10.5% (536.6 million individuals) in 2021 and will increase to 12.2% (783.2 million) in 2045 [2]. Diabetes increases the risk of cardiovascular disease and is associated with relatively specific microvascular problems affecting the eyes, nerves, and kidneys [3,4,5]. DM with diabetic complications is becoming a crucial public health issue. Therefore, only by fully understanding the risk factors of diabetes can we effectively and timely prevent and screen diabetes.

Pre-diabetes, which includes impaired fasting glucose (IFG) and impaired glucose tolerance, is the transition from normal glucose metabolism to diabetes (IGT) [6]. A recent national cross-sectional survey revealed that 35.7% of Chinese adults had pre-diabetes [7]. In those with pre-diabetes, the annual chance of getting diabetes is 5–10%, and up to 70% of people eventually develop the disease [8]. Interventions initiated at the pre-diabetes stage are more effective and prudent than those initiated after the beginning of diabetes because they can stop or delay pre-diabetes progression to diabetes [9, 10]. Considering that patients with pre-diabetes are more likely to develop diabetes than the general population, it is even more urgent to explore the risk factors for progression from pre-diabetes to diabetes and to intervene promptly.

Skeletal muscle is one of insulin's primary target organs and plays an important role in maintaining glucose homeostasis [11, 12]. According to a recent study, decreasing muscle mass was closely associated with insulin resistance (IR) [13]. Skeletal muscle atrophy and abnormal muscle protein metabolisms have been seen in people with type 2 diabetes mellitus (T2DM). Creatinine (Cre) is the only phosphate creatine metabolite found in the body's skeletal muscles. Cre, despite being a measure of renal function, is affected by muscle size because it is produced by muscle mass. Due to the stability of total skeletal muscle mass, the Cre concentration is also reasonably steady [14]. Serum creatinine (Scr) is regarded as a low-cost and simple-to-measure index for assessing the condition of skeletal muscle [15]. Recent research has revealed that the intriguing new biomarker of Cre to body weight ratio (Cre/BW) is strongly associated with incident DM [16, 17] and non-alcoholic fatty liver disease (NAFLD) [18, 19] in general populations. Furthermore, recent research found that the Cre/BW ratio is closely related to the incidence of diabetes in general Chinese participants [20]. However, the relationship between the Cre/BW ratio and DM has not been studied in pre-diabetes participants, a population that is prone to progress to diabetes. Therefore, in order to ascertain the connection between the Cre/BW ratio and the risk of developing diabetes from pre-diabetes, we carried out a retrospective cohort analysis utilizing available data.

Methods

Study design

Using information from the database provided by China Rich Healthcare Group, we carried out a retrospective cohort research. The Cre/BW ratio at baseline was the most interesting independent variable in the present study. The dependent variable was diabetes diagnosed during follow-up (dichotomous variable: 0 = non-DM, 1 = DM).

Data source

The data used in the study is from a publicly available database. The raw data was taken from the DATADRYAD database (https://datadryad.org/stash) for free provided by Chen, Ying et al. (2018), Data from: Association of body mass index and age with incident diabetes in Chinese adults: a population-based cohort study, Dryad, Dataset, https://doi.org/10.5061/dryad.ft8750v. Researchers could use this data for secondary studies without breaching the rights of the authors in accordance with Dryad's terms of service [21].

Study population

To mitigate the impact of selection bias, a non-selective and consecutive collection of participant data was conducted at 32 sites across 11 cities in China (including Beijing, Suzhou, Changzhou, Shanghai, Nantong, Shenzhen, Chengdu, Nanjing, Guangzhou, Hefei, and Wuhan). This approach aimed to minimize the potential for over- or under-estimation of results, which is a common issue associated with selection bias in studies. To protect the anonymity of the participants, their identification information was encoded using non-traceable codes. The China Rich Healthcare Group's electronic medical record system was used to retrieve the data. The Rich Healthcare Group Review Board approved the original study, and the informed consent was waived because of the retrospective nature of this study [22].

The original study comprised 685,277 people who were at least 20 years old and had at least two health checks. The original study removed 473,744 subjects in total. Ultimately, 211,833 people were included in the initial research analysis. The following were the exclusion criteria for the original study: (i) absence of fasting plasma glucose (FPG) value, weight, sex, and height at baseline; (ii) extreme body mass index (BMI) values (< 15 kg/m2 or > 55 kg/m2); (iii) fewer than two years of follow-up times; (iv) diagnosis of diabetes at enrollment; and (v) uncertain diabetes status at follow-up (18). We first excluded 185,815 participants who were outside the FPG 5.6–6.9 mmol/l range at baseline in the current study. Pre-diabetes was characterized as an FPG level between 5.6 and 6.9 mmol/L by the American Diabetes Association 2021 guidelines [23]. We further excluded participants with missing Scr information (n = 1352) and those with abnormal and extreme Cre/BW ratio (greater or less than three standard deviations from the mean) (n = 170) [24]. The secondary analysis eventually comprised 24,506 subjects. Figure 1 depicted the process of participant selection.

Fig. 1
figure 1

Flowchart of study participants. Figure 1 showed the inclusion of participants. In the original study, 211,833 participants were assessed for eligibility. We further excluded 187,327 participants. The final analysis included 24,506 subjects in the present study

Full size image

Variables

Cre to body weight ratio

We obtained the information on the Cre to body weight ratio at baseline and recorded it as a continuous variable. Defining the Cre/BW ratio was described as: Cre/BW ratio = serum creatinine divided by body weight. Notably, the unit of Cre was umol/L, and the unit of BW was kg.

Outcome measures

Our primary outcome variable was incident diabetes (dichotomous variable: 0 = non-DM, 1 = DM). At the follow-up evaluation, incident diabetes was diagnosed based on self-report or FPG ≥ 7.0 mmol/l [21]. Participants were censored at the time of diabetes diagnosis or their last visit, whichever came first. The duration of follow-up was five years.

Covariates

We chose covariates for our study based on previous research and clinical experience [16, 17, 20, 21]. The following variables were therefore used as covariates based on the principles outlined above (1) continuous variables: systolic blood pressure (SBP), height, age, Blood urea nitrogen(BUN), diastolic blood pressure (DBP), triglyceride(TG), FPG, total cholesterol (TC), alanine aminotransferase(ALT), high-density lipoprotein cholesterol (HDL-c), aspartate aminotransferase(AST), low-density lipoprotein cholesterol(LDL-c); (2) categorical variables: smoking status, gender, drinking status, and family history of diabetes.

Variable measurement

Each time they visited the health check center, each participant was given a thorough questionnaire that inquired about their lifestyle, family history of chronic illness, demographic factors, and medical history. Weight, blood pressure, and height were measured by trained staff. Participants were weighed in light clothing with no shoes and measured to nearest 0.1 kg. Height was measured accurately to within 0.1 cm. Blood pressure was measured using mercury sphygmomanometers. Fasting venous blood samples were collected after fasting for at least 10 h at each appointment. HDL-c, AST, Scr, FPG, TC, BUN, TG, ALT, and LDL-c were measured on an autoanalyzer (Beckman 5800) [21].

Missing data processing

In observational research, missing data were common and could not be prevented entirely [25]. In our study, the number of participants with missing data of SBP, DBP, TC, TG, HDL-c, LDL-c, ALT, AST, BUN, smoking status, and drinking status was 6(0.024%), 6(0.024%), 394(1.61%), 394(1.61%), 9299(37.95%), 8683(35.43%), 133(0.54%), 13,822(56.40%), 1559(6.36%), 16,184(66.04%), and 16,184(66.04%), respectively. To reduce the variation brought on by missing variables, which made it impossible to accurately depict the statistical effectiveness of the target sample throughout the modeling phase, this study used multiple imputations [26]. The imputation model included age, sex, height, TG, AST, DBP, BUN, TC, ALT, LDL-c, smoking status, HDL-c, drinking status, SBP, and family history of diabetes. The procedures for analyzing missing data used missing-at-random (MAR) assumptions [25].

Statistical analysis

Quartiles and deciles of the Cre/BW ratio stratified the participants. In the case of continuous variables, baseline characteristics were presented as mean ± standard deviation (SD) (Gaussian distribution) or median (range) (Skewed distribution), and as percentages for categorical variables. We used three kinds of statistical tests to detect the differences among different Cre/BW ratio groups: χ2 for categorical variables, One-Way ANOVA for normal distribution, or Kruskal-Whallis H for skewed distribution. We employed the Kaplan–Meier method to compute the survival estimates and time-to-event variables. We utilized the log-rank test in pre-diabetes populations to compare the probability of diabetes-free survival among Cre/BW ratio groups [27].

To assess covariate colinearity, the variance inflation factor (VIF) was calculated [28]. VIF = 1/(1-R2). Where R2 was the R-squared value derived from a linear regression equation, this variable was the dependent variable, and all other variables were independent. If the VIF was greater than 5, then the variables would be considered collinear and could not be included in the multiple regression model (Table S1).

Following collinearity screening, we used the univariate and multivariate Cox proportional-hazards regression to assess three alternative models for the relationship between the Cre/BW ratio and the risk of diabetes. As for the model I, it was the nonadjusted model with no covariates adjusted. As for model II, it was the minimally-adjusted model with only sociodemographic variables adjusted, including SBP, gender, DBP, age, family history of diabetes, height, drinking, and smoking status. Model III was the fully-adjusted model with covariates presented in Table 1, including SBP, gender, FPG, age, height, BUN, DBP, TG, ALT, HDL-c, AST, family history of diabetes, LDL-c, drinking, and smoking status. In cases where the Hazard ratio (HR) increased by 10% or more after covariances were introduced to the model, we recorded the effect sizes (HR) and 95% confidence intervals (CIs) and adjusted them [29]. Additionally, it referred to the outcomes of the collinearity screening. We ultimately decided not to include TC in the multivariate Cox proportional-hazards regression equation since the results of the collinearity screening indicated that it was collinear with other variables (Table S1).

Table 1 The baseline characteristics of participants
Full size table

Methods based on Cox proportional-hazards regression models were often accused of being unsuitable for dealing with non-linear models. As a result, we applied the Cox proportional hazards regression model with cubic spline functions and the smooth curve fitting (penalized spline method) to address the non-linearity between the Cre/BW ratio and diabetes. Following the detection of non-linearity, the inflection point was computed using a recursive technique, and two-piecewise Cox proportional-hazards regression models were conducted on both sides of the inflection point. The optimal model for characterizing the risk associated with the Cre/BW ratio and diabetes was determined using a log-likelihood ratio test [30].

For subgroup analysis, a stratified Cox proportional-hazards regression model was applied to the various subgroups (gender, BMI, age, SBP, TG, DBP, drinking and smoking status, and family history of diabetes). Firstly, continuous variable age (< 30, ≥ 30 to < 40, ≥ 40 to < 50, ≥ 50 to < 60, ≥ 60 to < 70, ≥ 70 years), BMI (< 18.5, ≥ 18.5 to < 24, ≥ 24 to 28, ≥ 28 kg/m2), SBP(< 140, ≥ 140 mmHg), DBP(< 90, ≥ 90 mmHg), TG(< 1.7, ≥ 1.7 mmol/L) were converted to a categorical variable based on the clinical cut point [31,32,33]. Secondly, we adjusted each stratification for all factors in addition to the stratification factor itself (SBP, gender, FPG, age, height, BUN, DBP, TG, ALT, HDL-c, AST, family history of diabetes, LDL-c, drinking and smoking status). Lastly, a likelihood ratio test for models with and without interaction terms was used to test for interactions [34, 35].

We conducted a sensitivity analysis of our findings to determine their reliability. To confirm the results of the Cre/BW ratio as a continuous variable and investigate the possibility of non-linearity, the Cre/BW ratio was converted into a categorical variable in accordance with the deciles. Obesity and weight gain are associated with an increased risk of developing diabetes [36]. Smoking and alcohol consumption are related to an increased risk of T2DM [37]. When exploring the Cre/BW ratio and incident diabetes association in other sensitivity analyses, we excluded participants with a smoking and drinking history. We also excluded drinking and smoking status from the multivariate model as sensitivity analysis. Smoking and drinking status did not have complete data in about 70% of the cases, and might not be useful as covariates to adjust in the model. Furthermore, the continuity covariate was also inserted into the equation (model IV) as a curve using cubic splines to ensure the robustness of the results [38]. We also calculated E-values to examine the possibility of unmeasured confounding between the Cre/BW ratio and the risk of diabetes [39]. The STROBE statement was used to write all the results [29].

R (http://www.r-project.org, The R Foundation) and EmpowerStats (http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA) were used to execute the statistical analyses. A P-value of < 0.05 determined statistical significance in all cases.

Results

Characteristics of participants

Table 1 listed the baseline characteristics of the participating participants. The age range of the participants included in this study was 20–99 years, with a majority of 16,232 individuals (66.24%) being men. The mean baseline Cre/BW ratio was 1.06 (SD 0.22) umol/L/kg. During a median follow-up period of 2.89 years, 2512 patients (10.25%) were diagnosed with diabetes. We assigned adults into subgroups using Cre/BW ratio quartiles (Q1: < 0.90 umol/L/kg, Q2: 0.90–1.04 umol/L/kg, Q3: 1.04–1.20 umol/L/kg, Q4: ≥ 1.20 umol/L/kg). When compared with the Q1(< 0.90 umol/L/kg) group, the values or proportions of age, HDL-c, BUN, Scr, males, current or ever smokers, and current or ever drinkers increased significantly in the Q4(Cre/BW ratio ≥ 1.20 umol/L/kg) group. In contrast, the opposite results were detected in covariates in terms of SBP, LDL-c, BMI, height, weight, FPG, TG, AST, DBP, TC, ALT, females, never smokers, never drinkers, and family history of diabetes.

According to Fig. 2, the Cre/BW ratio levels had a normal distribution ranging from 0.377 to 1.802 umol/L/kg; the average was 1.062 umol/L/kg. Participants were divided into two groups based on whether participants progressed to diabetes from pre-diabetes during the follow-up. As shown in Fig. 3, the distributions of the Cre/BW ratio in the non-diabetes group were higher than those in the diabetes group. Men were more likely to progress to diabetes regardless of age group when age groups were stratified by 10 intervals (Fig. 4). Furthermore, the incidence of diabetes increased with age in both females and males.

Fig. 2
figure 2

Distribution of the Cre/BW ratio. Figure 2. It presented a normal distribution ranging from 0.377 to 1.802 umol/L/kg, and the average was 1.062 umol/L/kg

Full size image
Fig. 3
figure 3

Data visualization of the Cre/BW ratio of all participants from diabetes and non-diabetes groups. Figure 3 revealed that the Cre/BW ratio distribution level was lower in the diabetic group. In contrast, the ratio of Cre to BW was considerably larger in the non-diabetic group

Full size image
Fig. 4
figure 4

Diabetes incidence rate of age stratification by 10 intervals. Figure 4 demonstrated that, regardless of age group, male individuals had a higher incidence of diabetes than female subjects in age stratification by ten intervals. It was also shown that the prevalence of diabetes increased with age in both men and women

Full size image

The incidence rate of diabetes in participants with pre-diabetes

Table 2 revealed that 2512 (10.25%) participants progressed to diabetes from pre-diabetes during a median follow-up time of 2.89 years. Incidence rates for all people combined were 3.47 per 100 person-years. Particularly, the four Cre/BW ratio groups had cumulative incidences of 4.62, 3.62, 3.07, and 2.57 per 100 person-years, respectively. The incidence rate of total diabetes and each Cre/BW ratio group was 10.25%(9.87%-10.63%), 13.68%(12.82%-14.54%), 10.71%(9.93%-11.48%), 9.09%(8.37%-9.81%), and 7.52%(6.86%-8.18%), respectively. Participants with a high Cre/BW ratio had lower incidence rates of diabetes compared to the group with the lowest Cre/BW ratio (P < 0.0001 for trend) (Fig. 5).

Table 2 Incidence rate of incident diabetes
Full size table
Fig. 5
figure 5

Incidence of diabetes according to the deciles of the Cre/BW ratio. Figure 5. Participants in the high Cre/BW ratio group had a lower diabetes incidence than the lower Cre/BW ratio group (P < 0.0001 for trend)

Full size image

The results of univariate analyses using the Cox proportional-hazards regression model

The univariate analysis was conducted on the available data, showing that the factors in terms of height and Scr were not connected with diabetes. Still, age, weight, BMI, SBP, DBP, FPG, TC, TG, LDL-c, HDL-c, ALT, AST, BUN, current smokers, current drinkers were positively associated to diabetes, and Cre/BW ratio, females, ever drinkers were negatively related with diabetes (See Table 3 for detail).

Table 3 Univariate proportional hazard analysis (n = 24,506)
Full size table

Figure 6 depicted the Kaplan–Meier survival curves for diabetes-free survival probability stratified by Cre/BW ratio group. Between the Cre/BW ratio groups, there were substantial differences in the likelihood of surviving without developing diabetes (log-rank test, P < 0.0001). Diabetes-free survival probabilities increased as the Cre/BW ratio increased, which indicated that those with the highest Cre/BW ratio faced the lowest risk of diabetes.

Fig. 6
figure 6

Kaplan–Meier event-free survival curve. Figure 6. Kaplan–Meier event-free survival curve. The probability of diabetes-free survival differed significantly between the Cre/BW ratio groups (log-rank test, P < 0.0001). The probability of diabetes-free survival gradually increased with the rising Cre/BW ratio, indicating that the group with the highest Cre/BW ratio had the lowest risk of developing diabetes

Full size image

Results from a multivariate Cox proportional-hazards regression model

The authors created three models using the Cox proportional-hazards regression model to examine the relationship between the Cre/BW ratio and diabetes in the subjects with pre-diabetes. In the unadjusted model (Model I), a 61.3% reduction in the risk of diabetes was associated with an increase of 1 unit (umol/L/kg) in the Cre/BW ratio (HR = 0.387, 95%CI 0.321 to 0.466). Statistical significance was found in the results. When only demographic factors were considered in the minimally-adjusted model (Model II), the risk of diabetes fell by 76.5% for each extra umol/L/kg of the Cre/BW ratio (HR = 0.235, 95%CI 0.191 to 0.288). The model's results on the relationship between diabetes and the Cre/BW ratio were statistically significant. In the fully-adjusted model (Model III), each additional umol/L/kg of the Cre/BW ratio was accompanied by a 55.5% decrease in diabetes risk (HR = 0.445, 95%CI 0.361 to 0.548). As shown by the distribution of confidence intervals, the relationship between the Cre/BW ratio and diabetes obtained by the model was reliable (Table 4).

Table 4 Relationship between Cre/BW ratio and the incident diabetes in different models
Full size table

Besides, we analysed the relationship between log (Cre), log (weight) and the risk of diabetes separately (Table S2). We found that weight was positively associated with the risk of progression to diabetes in prediabetes, whereas Cre had a negative association with the risk of progression to diabetes. The results suggest that both Cre and weight contributed to our analysis of the relationship between Cre/BW ratio and diabetes.

Sensitivity analysis

The robustness of our findings was examined using a number of sensitivity analyses. The Cre/BW ratio was first converted into a categorical variable (based on deciles), which was then included in our model after being categorical-transformed. After the Cre/BW ratio was converted into a categorical variable, the results showed that the trend of HR change was not completely consistent between the groups. The above results suggest that the relationship between the Cre/BW ratio and the risk of diabetes progression is not completely linear and that there is a possibility of a non-linear relationship between the two (Table 4).

In addition, to include the continuity covariate as a curve in the equation, we used cubic splines. This usually remained consistent with the fully adjusted model, according to the Model IV result in Table 4 (HR = 0.521, 95%CI: 0.420–0.646, P < 0.0001). Besides, E-values were calculated to assess sensitivity to unmeasured confounding. 3.92 was the E-value. The E-value was higher than the relative risk of unmeasured confounders and the Cre/BW ratio, implying that unmeasured or unknown confounders had little influence on the connection between the Cre/BW ratio and incident diabetes.

Furthermore, the authors excluded participants with BMI ≥ 28 kg/m2 (N = 20,431) for the sensitivity analysis. A negative association of the Cre/BW ratio with diabetes risk was also observed after adjusting for confounding factors (HR = 0.515, 95%CI:0.399 to 0.665) (Table 5). We also excluded participants with a history of smoking. The results showed that after adjusting SBP, gender, FPG, age, height, BUN, DBP, TG, ALT, HDL-c, AST, family history of diabetes, LDL-c, and drinking status, the Cre/BW ratio was still negatively associated with diabetes (HR = 0.475, 95% CI:0.370 to 0.612) (Table 5). For sensitivity analyses, we also excluded persons with a history of drinking. We still got similar results (HR = 0.453, 95% CI:0.359 to 0.571).

Table 5 Relationship between Cre/BW ratio and diabetes in different sensitivity analyses
Full size table

Since smoking and alcohol status had about 70 percent of missing data, these data might not be suitable as covariates. In other sensitivity analyses, we excluded drinking and smoking status from the multivariate model. It was still similar to the previous results (HR = 0.446, 95% CI:0.362 to 0.549). The results obtained from sensitivity analysis indicated the well-robustness of our findings (Table 5).

Cox proportional hazards regression model with cubic spline functions to address non-linearity

We discovered that the Cre/BW ratio and incident diabetes relationship in pre-diabetic patients was also non-linear using Cox proportional hazards models with cubic spline functions and smooth curve fitting (Fig. 7). We used a basic Cox proportional-hazards regression model to fit the data based on the sensitivity analysis, and we determined the best-fitting model by calculating log-likelihood ratios. In our analysis, we discovered that the P value for the log-likelihood ratio test was less than 0.05. We first determined the inflection point to be 1.072 umol/L/kg using a recursive technique. We then used two-piecewise Cox proportional-hazards regression models to calculate the HR and CI on either side of the inflection point. At the inflection point on the left side, the effect size and 95%CI were 0.294 (0.208, 0.414), respectively. At the inflection point on the right side, they were 0.712 (0.492, 1.029), respectively (Table 6).

Fig. 7
figure 7

The non-linear relationship between the Cre/BW ratio and the risk of diabetes. Figure 7. We used a Cox proportional hazards regression model with cubic spline functions to evaluate the relationship between the Cre/BW ratio and diabetes risk. The relationship between the Cre/BW ratio and diabetes showed an L-shaped curve with an inflection point of 1.072 umol/L/kg

Full size image
Table 6 The result of the two-piecewise Cox regression model
Full size table

To explore the non-linear relationship between the Cre/BW ratio and diabetes in pre-diabetic and non-pre-diabetic participants, we further investigated the non-linear relationship between the Cre/BW ratio and diabetes in people without pre-diabetes. It was found that the relationship between the Cre/BW ratio and diabetes remained non-linear in the population without pre-diabetes (Fig. S1). After controlling for covariates, the point of inflection for the Cre/BW ratio was 1.068 umol/L/kg. At the inflection point on the left side, the HR and 95%CI were 0.030 (0.018, 0.049), respectively. At the inflection point on the right side, they were 0.650 (0.382, 1.106), respectively (Table S3).

The results of subgroup analyses

The grouping of variables such as Gender, BMI, SBP, family history of diabetes, smoking, and alcohol use status did not affect the relationship between Cre/BW ratio and the risk of progression to diabetes in prediabetes (Table 7). In contrast, significant interactions were detected in variables such as age, DBP, and TG.

Table 7 Effect size of the Cre/BW ratio on incident diabetes in prespecified and exploratory subgroups
Full size table

Specifically, a stronger association between the Cre/BW ratio and diabetes was observed in the participants who were younger than 30 years (HR = 0.036,95%CI:0.005–0.245) and 30–50 years (HR = 0.192,95%CI:0.097–0.380; HR = 0.325,95%CI:0.196–0.539), as well as among those with TG < 1.7 mmol/L(HR = 0.373,95%CI:0.281–0.495) and DBP < 90 mmHg (HR = 0.373, 95%CI: 0.281–0.495). In contrast, the association of the Cre/BW ratio with the risk of incident diabetes was attenuated among participants who were above 50 years of age (HR = 0.545, 95%CI: 0.376–0.791; HR = 0.689,95%CI:0.462–1.206; HR = 0.786,95%CI:0.476–1.298), and among those with DBP ≥ 90 mmHg (HR = 0.675,95%CI:0.437–1.041) and TG ≥ 1.7 mmol/L (HR = 0.570, 95%CI:0.430–0.755).

Discussion

The purpose of our retrospective cohort study was to investigate the relationship between the Cre/BW ratio and diabetes in pre-diabetic patients. We discovered that increasing the Cre/BW ratio was associated with a considerably lower risk of diabetes. Furthermore, a threshold effect curve was discovered, and different associations between the Cre/BW ratio and diabetes risk were discovered on both sides of the inflection point. In addition, as potential moderators of the relationship between the Cre/BW ratio and pre-diabetes, age, TG, and DBP were found to be significant, as significantly stronger associations were observed in participants with age < 50 years, TG < 1.7 mmol/L, and DBP < 90 mmHg, while significantly weaker associations were detected in age > 50 years, TG ≥ 1.7 mmol/L, DBP ≥ 90 mmHg.

In a Japanese study, the actual 5-year transition rate from pre-diabetes to diabetes was 8.5% [40]. In American pre-diabetic patients aged 70–79, the 7-year incidence of developing diabetes was 10.6% [41]. In the current investigation, over a median follow-up period of 2.89 years, the incidence of diabetes among patients with pre-diabetes was 10.25%. Variations in the age and ethnicity of the participants may bring about these variations in the incidence of diabetes among these pre-diabetic patients. It is important to note that pre-diabetic people have a high chance of acquiring diabetes, according to all research. Therefore, it is crucial to actively look for additional risk factors for developing diabetes from pre-diabetes.

The Cre/BW ratio was found to be inversely associated with the risk of diabetes in a recent Japanese study (HR = 0.84, 95% CI 0.80–0.88 for men and 0.88, 0.81–0.96 for women) in the general population [16] after adjusting age, FPG, drinking status, exercise, and smoking status. A Chinese study also found an inverse association between the Cre/BW ratio and the risk of diabetes in the general population (HR = 0.268, 95%CI 0.229–0.314) after adjusting confounding variables. They also found a negative non-linear connection between the Cre/BW ratio and incident diabetes [20]. By Cox proportional hazards regression model analysis, our study found that the decline in the Cre/BW ratio was strongly associated with an increased risk of diabetes, which was consistent with previous findings. At the same time, it is essential to point out that, unlike the above two studies, our study focused on pre-diabetic adults at higher risk of progression to diabetes. For this population, it is of great significance to delay its progression to diabetes and even propose to switch to normoglycemia with aggressive intervention. This is the first study to examine the relationship between Cre/BW ratio and diabetes risk in patients with pre-diabetes. Identifying the Cre/BW ratio as a protective factor for developing diabetes in pre-diabetes and clarifying the association between them will benefit diabetes prevention in patients with pre-diabetes. Also, the HRs of the relationship between creatinine body weight ratio and diabetes were different in our study from that of Chen et al. [20].

Concurrently, our sensitivity analysis found that the relationship remained stable among participants who never smoked or drank and those who did not have a BMI > 28 kg/m2. We also found that the negative association between the Cre/BW ratio and diabetes remained stable when the multiple regression equation did not adjust smoking and drinking status. The efforts indicated above supported the stability of the association between the Cre/BW ratio and diabetes risk.

The following are some potential explanations for the connection between the Cre/BW ratio and the progression from pre-diabetes to diabetes. Although there are no published studies on Cre/BW ratio and progression from pre-diabetes to diabetes, it is clear that insulin resistance is strongly associated with progression from pre-diabetes to diabetes [8, 42, 43]. It is generally known that muscle mass absorbs 80%-90% of blood glucose when the body is in a hyperinsulinemia-euglycemia state [44]. The ability to absorb glucose from the blood can be decreased with less muscle mass [45]. Moreover, insulin resistance, a major pathogenic mechanism of diabetes, may mediate the association between low muscle mass and incident diabetes [46]. The proportion of muscle mass to body weight is crucial since both an increase in muscle mass and fat mass could result in weight gain [47]. According to numerous research, weight-adjusted appendicular skeletal muscle mass has been linked to IR and cardiometabolic risk factors [12, 48,49,50,51]. Moreover, weight-adjusted appendicular skeletal muscle mass is associated with incident diabetes [14], metabolic syndrome [52], and NAFLD [53, 54]. This is due to the fact that a low skeletal muscle mass index (SMI) is linked to an increase in visceral fat, which in turn is connected to diabetes [55]. Cre is an alternative skeletal muscle mass marker; its levels positively correlate with skeletal muscle mass [56, 57]. Consequently, the weight-adjusted SMI was positively correlated with the Cre/BW ratio. Therefore, we propose the possibility that the relationship between Cre/BW ratio and progression from pre-diabetes to diabetes is mediated through insulin resistance.

Furthermore, the present investigation discovered for the first time a non-linear link between the Cre/BW ratio and the risk of diabetes in pre-diabetic patients. To explain the non-linear association, this study used a two-piecewise Cox proportional hazards regression model. The findings revealed a non-linear relationship and saturation effect between the Cre/BW ratio and pre-diabetic patients' probability of developing diabetes. After controlling for covariates, the point of inflection for the Cre/BW ratio was 1.072 umol/L/kg. A 1 umol/L/kg rise in the Cre/BW ratio was related to a 70.6% reduction in the risk of diabetes when the ratio was below 1.072 umol/L/kg. As we mentioned above, when the Cre/BW ratio was < 1.072 umol/L/kg, as the Cre/BW ratio increases, the skeletal muscle mass increases accordingly, so that the risk of developing insulin resistance or diabetes decreases accordingly. However, at the same time, we also found that the risk of diabetes did not reduce significantly with the increase of the Cre/BW ratio, when the Cre/BW ratio was above 1.072 umol/L/kg. The possible reason is that although Cre is an alternative marker of skeletal muscle mass, at the same time, it is also an indicator to reflect the state of renal function. Creatinine levels were influenced by renal functional status. Our previous study found that a decline in the glomerular filtration rate (GFR) was strongly associated with an increased risk of diabetes [58]. Chronic kidney disease (CKD) was a crucial and independent predictor of diabetes, according to another population-based cohort study in Taiwan (adjusted HR 1.204; 95% CI 1.11, 1.31) [59]. A recent study also showed that a decrease in GFR was associated with an increase in IR in 4680 American participants without diabetes [60]. This might be the reason when the Cre/BW ratio was ≥ 1.072 umol/L/kg, the risk of diabetes no longer decreased accordingly with the Cre/BW ratio increased. These results indicate that the Cre/BW ratio may be a viable predictor for individuals who wish to prevent diabetes by losing weight. Therefore, patients with pre-diabetes can try to control the Cre/BW ratio above 1.072 umol/L/kg through various treatment modalities, then the risk of progression to diabetes may be significantly reduced. Excellent clinical relevance can be derived from the identification of a non-linear association between the Cre/BW ratio and diabetes in pre-diabetic patients. In patients with pre-diabetes, it promotes clinical consultation and offers a reference for decision-making that is optimized for diabetes prevention. In addition, although the inflection points for the Cre/BW ratio were relatively similar in those with and without pre-diabetes (1.072 and 1.068 umol/L/kg), there were significant differences in the effect values for the relationship between Cre/BW ratio and diabetes on both sides of the inflection points. A 1 umol/L/kg rise in the Cre/BW ratio was related to a 70.6% reduction in the risk of diabetes in participants with pre-diabetes when the ratio was below 1.072 umol/L/kg. However, a 1 umol/L/kg rise in the Cre/BW ratio was related to a 97% reduction in the risk of diabetes in participants without pre-diabetes when the ratio was below 1.068 umol/L/kg. This result further suggests that patients with pre-diabetes are at high risk of progression to diabetes and that reducing the risk of developing diabetes by intervening in the Cre/BW ratio is not as significant as in non-prediabetic patients. Therefore, patients with pre-diabetes must be more aggressively sought and controlled for associated risk factors to reduce the risk of progression from pre-diabetes to diabetes.

We have outlined some of our study's advantages below. (1) Our study's overall sample size was large, which is a strength. (2) To the best of our knowledge, this is the first study to investigate the relationship between the Cre/BW ratio and the risk of diabetes in Chinese pre-diabetic individuals. (3) In this study, non-linearity is investigated and further explained. This study significantly improves upon earlier research. (4) In this study, missing data were handled using multiple imputations. Using multiple imputations could increase statistical power and reduce any potential bias brought on by the absence of covariate information. (5) Since this is an observational study, it is vulnerable to potential confounding. We minimized residual confounding by using strict statistical adjustment. (6) A series of sensitivity analyses were carried out throughout this study to ensure the accuracy of the findings (conversion of target-independent variable form, subgroup analysis, using cubic splines to insert the continuity covariate into the equation as a curve, calculating E-values to investigate the possibility of unmeasured confounding, and reanalyzing the association between the Cre/BW ratio and diabetes among participants who never smoked or drank, and among those who did not have a BMI > 28 kg/m2).

Our research has the following flaws that must be addressed: First, because the design of this study is observational, we cannot determine the exact causal relationship. Second, as with all observational studies, even while known potential confounders such as TG, age, FPG, and BUN were controlled, uncontrolled or unmeasured confounders may still have been present. In spite of this, the authors estimated the E-value to quantify the influence of unmeasured confounders and concluded that they were unlikely to account for the results. Third, diabetes was defined as an FPG ≥ 7.00 mmol/L and/or self-reported diabetes throughout the follow-up period rather than a 2-h oral glucose tolerance test or glycosylated hemoglobin level, which may underestimate the incidence of diabetes. Besides, there is a general issue with the outcome variable, in that it is not purely objective and people with certain characteristics may be followed up more closely and hence diagnosed earlier. In addition, diabetes's type remained unknown. Since T2DM accounts for more than 90% of all diabetes cases in China, it is the most prevalent kind of diabetes [61]. Our findings are, therefore, representative of T2DM patients. Fourth, the information from the original study did not include cases of diabetes during the first 2 years of follow-up. For such a large sample of participants, possible deaths during follow-up and diabetes observed in the first 2 years were inevitable. In the future, we can consider designing our studies and collecting the observed cases of diabetes throughout the follow-up period. Fifth, what readers need to know is whether they can use the Cre/BW ratio to predict subsequent diabetes. Therefore, a score based the ratio and other covariates is necessary. In the future we can collect our database and build a prediction model for progression of pre-diabetes to diabetes by Cre/BW ratio and other indicators. Sixth, we only measured the Cre/BW ratio at baseline and did not account for changes in the Cre/BW ratio over time. As a result, future studies should include as many variables as possible, including information on changes in the Cre/BW ratio during the study period. Sixth, the decile analysis may be confusing, since it is only when adding the blood covariates does the relationship of risk to Cre/BW ratio become non-monotonic. As the confounding factors change, the relationship between the Cre/BW ratio and the risk of progression to diabetes from pre-diabetes may change accordingly. In future studies, we will collaborate with other researchers to design cohort studies and collect more variables to observe changes in the relationship between the Cre/BW ratio and the risk of progression to diabetes from pre-diabetes by adjusting for various confounders. Finally, in this study, pre-diabetes was only recognized by the researchers in the subjects with impaired fasting glucose levels, which could have resulted in a missed diagnosis [62].

Conclusion

This study indicates a negative and non-linear association between the Cre/BW ratio and incident diabetes in pre-diabetic Chinese individuals. A saturating effect exists between the Cre/BW ratio and diabetes risk in pre-diabetic patients. When the Cre/BW ratio is < 1.072 umol/L/kg, there was a significant negative association with the risk of progression from pre-diabetes to diabetes. This result will offer clinicians a reference for aggressive weight loss and muscle mass increase in pre-diabetic patients. From a treatment perspective, it makes sense to increase the Cre/BW ratio above the inflection point by aggressive intervention to aggressive muscle mass gain and weight loss.

Availability of data and materials

The data used in the study is from a publicly available database. Data could be downloaded from the 'DATADRYAD' database (https://datadryad.org/stash).

Change history

Abbreviations

Cre/BW:

Creatinine to body weight

DM:

Diabetes mellitus

Cre:

Creatinine

IDF:

International Diabetes Federation

NAFLD:

Non-alcoholic fatty liver disease

HTG:

Hypertriglyceridemia

IGT:

Impaired glucose tolerance

SMI:

Skeletal muscle mass index

SBP:

Systolic blood pressure

Scr:

Serum creatinine

T2DM:

Type 2 diabetes mellitus

ALT:

Alanine aminotransferase

BMI:

Body mass index

TG:

Triglyceride

DBP:

Diastolic blood pressure

VIF:

Variance inflation factor

GFR:

Glomerular filtration rate

LDL-c:

Low-density lipid cholesterol

HDL-c:

High-density lipoprotein cholesterol

CKD:

Chronic kidney disease

AST:

Aspartate aminotransferase

FPG:

Fasting plasma glucose

TC:

Total cholesterol

BUN:

Blood urea nitrogen

IFG:

Impaired fasting glucose

GAM:

Generalized additive models

HR:

Hazard ratios

Ref:

Reference

CI:

Confidence intervals

MAR:

Missing-at-random

IR:

Insulin resistance

SD:

Standard deviation

References

  1. Zimmet PZ, Magliano DJ, Herman WH, Shaw JE. Diabetes: a 21st century challenge. Lancet Diabetes Endo. 2014;2(1):56–64.

    Google Scholar 

  2. Sun H, Saeedi P, Karuranga S, Pinkepank M, Ogurtsova K, Duncan BB, Stein C, Basit A, Chan J, Mbanya JC, et al. IDF Diabetes Atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183: 109119.

    PubMed  Google Scholar 

  3. Punthakee Z, Goldenberg R, Katz P. Definition, classification and diagnosis of diabetes, prediabetes and metabolic syndrome. Can J Diabetes. 2018;42(Suppl 1):S10–5.

    PubMed  Google Scholar 

  4. Nathan DM. The diabetes control and complications trial/epidemiology of diabetes interventions and complications study at 30 years: overview. Diabetes Care. 2014;37(1):9–16.

    PubMed  CAS  Google Scholar 

  5. Peters SA, Huxley RR, Woodward M. Diabetes as risk factor for incident coronary heart disease in women compared with men: a systematic review and meta-analysis of 64 cohorts including 858,507 individuals and 28,203 coronary events. Diabetologia. 2014;57(8):1542–51.

    PubMed  Google Scholar 

  6. Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, Zinman B. Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care. 2007;30(3):753–9.

    PubMed  CAS  Google Scholar 

  7. Wang L, Gao P, Zhang M, Huang Z, Zhang D, Deng Q, Li Y, Zhao Z, Qin X, Jin D, et al. Prevalence and ethnic pattern of diabetes and prediabetes in China in 2013. Jama-J Am Med Assoc. 2017;317(24):2515–23.

    Google Scholar 

  8. Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for diabetes development. Lancet. 2012;379(9833):2279–90.

    PubMed  PubMed Central  Google Scholar 

  9. Perreault L, Pan Q, Mather KJ, Watson KE, Hamman RF, Kahn SE. Effect of regression from prediabetes to normal glucose regulation on long-term reduction in diabetes risk: results from the diabetes prevention program outcomes study. Lancet. 2012;379(9833):2243–51.

    PubMed  PubMed Central  CAS  Google Scholar 

  10. Prediabetes and the potential to prevent diabetes. Lancet. 2012;379(9833):2213.  https://doi.org/10.1016/S0140-6736(12)60960-X.

  11. Zierath JR, Krook A, Wallberg-Henriksson H. Insulin action and insulin resistance in human skeletal muscle. Diabetologia. 2000;43(7):821–35.

    PubMed  CAS  Google Scholar 

  12. Lim S, Kim JH, Yoon JW, Kang SM, Choi SH, Park YJ, Kim KW, Lim JY, Park KS, Jang HC. Sarcopenic obesity: prevalence and association with metabolic syndrome in the Korean Longitudinal Study on Health and Aging (KLoSHA). Diabetes Care. 2010;33(7):1652–4.

    PubMed  PubMed Central  CAS  Google Scholar 

  13. Moon SS. Low skeletal muscle mass is associated with insulin resistance, diabetes, and metabolic syndrome in the Korean population: the Korea National Health and Nutrition Examination Survey (KNHANES) 2009–2010. Endocr J. 2014;61(1):61–70.

    PubMed  CAS  Google Scholar 

  14. Son JW, Lee SS, Kim SR, Yoo SJ, Cha BY, Son HY, Cho NH. Low muscle mass and risk of type 2 diabetes in middle-aged and older adults: findings from the KoGES. Diabetologia. 2017;60(5):865–72.

    PubMed  Google Scholar 

  15. Hjelmesæth J, Røislien J, Nordstrand N, Hofsø D, Hager H, Hartmann A. Low serum creatinine is associated with type 2 diabetes in morbidly obese women and men: a cross-sectional study. BMC Endocr Disord. 2010;10:6.

    PubMed  PubMed Central  Google Scholar 

  16. Hashimoto Y, Okamura T, Hamaguchi M, Obora A, Kojima T, Fukui M. Creatinine to body weight ratio is associated with incident diabetes: population-based cohort study. J Clin Med. 2020;9(1):227.

    PubMed  PubMed Central  CAS  Google Scholar 

  17. He J. Creatinine-to-body weight ratio is a predictor of incident diabetes: a population-based retrospective cohort study. Diabetol Metab Syndr. 2022;14(1):7.

    PubMed  PubMed Central  CAS  Google Scholar 

  18. Okamura T, Hashimoto Y, Hamaguchi M, Obora A, Kojima T, Fukui M. Creatinine-to-bodyweight ratio is a predictor of incident non-alcoholic fatty liver disease: a population-based longitudinal study. Hepatol Res. 2020;50(1):57–66.

    PubMed  CAS  Google Scholar 

  19. Lin J, Zheng J, Lin X, Chen Y, Li Z. A low creatinine to body weight ratio predicts the incident nonalcoholic fatty liver disease in nonelderly Chinese without obesity and dyslipidemia: a retrospective study. Gastroent Res Pract. 2020;2020:4043871.

    Google Scholar 

  20. Chen Z, Zou Y, Yang F, Ding XH, Cao C, Hu H, Wang X. Relationship between creatinine to body weight ratios and diabetes mellitus: a Chinese cohort study. J Diabetes. 2022;14(3):167–78.

    PubMed  PubMed Central  CAS  Google Scholar 

  21. Chen Y, Zhang XP, Yuan J, Cai B, Wang XL, Wu XL, Zhang YH, Zhang XY, Yin T, Zhu XH, et al. Association of body mass index and age with incident diabetes in Chinese adults: a population-based cohort study. BMJ Open. 2018;8(9): e21768.

    Google Scholar 

  22. Geleris J, Sun Y, Platt J, Zucker J, Baldwin M, Hripcsak G, Labella A, Manson DK, Kubin C, Barr RG, et al. Observational study of hydroxychloroquine in hospitalized patients with Covid-19. New Engl J Med. 2020;382(25):2411–8.

    PubMed  CAS  Google Scholar 

  23. American Diabetes Association. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2021. Diabetes Care. 2021;44(Suppl 1):S15–33.

    Google Scholar 

  24. Zhang N, Hu X, Zhang Q, Bai P, Cai M, Zeng TS, Zhang JY, Tian SH, Min J, Huang HT, et al. Non-high-density lipoprotein cholesterol: High-density lipoprotein cholesterol ratio is an independent risk factor for diabetes mellitus: Results from a population-based cohort study. J Diabetes. 2018;10(9):708–14.

    PubMed  CAS  Google Scholar 

  25. White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med. 2011;30(4):377–99.

    PubMed  Google Scholar 

  26. Groenwold RH, White IR, Donders AR, Carpenter JR, Altman DG, Moons KG. Missing covariate data in clinical research: when and when not to use the missing-indicator method for analysis. Can Med Assoc J. 2012;184(11):1265–9.

    Google Scholar 

  27. Robson ME, Tung N, Conte P, Im SA, Senkus E, Xu B, Masuda N, Delaloge S, Li W, Armstrong A, et al. OlympiAD final overall survival and tolerability results: Olaparib versus chemotherapy treatment of physician’s choice in patients with a germline BRCA mutation and HER2-negative metastatic breast cancer. Ann Oncol. 2019;30(4):558–66.

    PubMed  PubMed Central  CAS  Google Scholar 

  28. Wax Y. Collinearity diagnosis for a relative risk regression analysis: an application to assessment of diet-cancer relationship in epidemiological studies. Stat Med. 1992;11(10):1273–87.

    PubMed  CAS  Google Scholar 

  29. Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, Poole C, Schlesselman JJ, Egger M. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Ann Intern Med. 2007;147(8):W163–94.

    PubMed  Google Scholar 

  30. Rothenbacher D, Rehm M, Iacoviello L, Costanzo S, Tunstall-Pedoe H, Belch J, Söderberg S, Hultdin J, Salomaa V, Jousilahti P, et al. Contribution of cystatin C- and creatinine-based definitions of chronic kidney disease to cardiovascular risk assessment in 20 population-based and 3 disease cohorts: the BiomarCaRE project. BMC Med. 2020;18(1):300.

    PubMed  PubMed Central  CAS  Google Scholar 

  31. Chen Z, Hu H, Chen M, Luo X, Yao W, Liang Q, Yang F, Wang X. Association of Triglyceride to high-density lipoprotein cholesterol ratio and incident of diabetes mellitus: a secondary retrospective analysis based on a Chinese cohort study. Lipids Health Dis. 2020;19(1):33.

    PubMed  PubMed Central  CAS  Google Scholar 

  32. Karpov Y, Khomitskaya Y. PROMETHEUS: an observational, cross-sectional, retrospective study of hypertriglyceridemia in Russia. Cardiovasc Diabetol. 2015;14:115.

    PubMed  PubMed Central  Google Scholar 

  33. Zhao WH, Zhang J, You Y, Man QQ, Li H, Wang CR, Zhai Y, Li Y, Jin SG, Yang XG. Epidemiologic characteristics of dyslipidemia in people aged 18 years and over in China. Zhonghua Yu Fang Yi Xue Za Zhi. 2005;39(5):306–10.

    PubMed  Google Scholar 

  34. Mullee A, Romaguera D, Pearson-Stuttard J, Viallon V, Stepien M, Freisling H, Fagherazzi G, Mancini FR, Boutron-Ruault MC, Kühn T, et al. Association Between Soft Drink Consumption and Mortality in 10 European Countries. JAMA Intern Med. 2019;179(11):1479–90.

    PubMed  Google Scholar 

  35. Keidel D, Anto JM, Basagaña X, Bono R, Burte E, Carsin AE, Forsberg B, Fuertes E, Galobardes B, Heinrich J, et al. The role of socioeconomic status in the association of lung function and air pollution-a pooled analysis of three adult ESCAPE cohorts. Int J Environ Res Public Health. 2019;16(11):1901.

    PubMed  PubMed Central  CAS  Google Scholar 

  36. Kahn SE, Hull RL, Utzschneider KM. Mechanisms linking obesity to insulin resistance and type 2 diabetes. Nature. 2006;444(7121):840–6.

    PubMed  CAS  Google Scholar 

  37. Lee SF, Li CP. Personality as a predictor of HbA1c level in patients with type 2 diabetes mellitus. Medicine. 2021;100(27): e26590.

    PubMed  PubMed Central  CAS  Google Scholar 

  38. Zhu F, Chen C, Zhang Y, Chen S, Huang X, Li J, Wang Y, Liu X, Deng G, Gao J. Elevated blood mercury level has a non-linear association with infertility in U.S. women: data from the NHANES 2013–2016. Reprod Toxicol. 2020;91:53–8.

    PubMed  CAS  Google Scholar 

  39. Haneuse S, VanderWeele TJ, Arterburn D. Using the E-value to assess the potential effect of unmeasured confounding in observational studies. JAMA J AM Med Assoc. 2019;321(6):602–3.

    Google Scholar 

  40. Heianza Y, Hara S, Arase Y, Saito K, Fujiwara K, Tsuji H, Kodama S, Hsieh SD, Mori Y, Shimano H, et al. HbA1c 5·7-6·4% and impaired fasting plasma glucose for diagnosis of prediabetes and risk of progression to diabetes in Japan (TOPICS 3): a longitudinal cohort study. Lancet. 2011;378(9786):147–55.

    PubMed  CAS  Google Scholar 

  41. Lipska KJ, Inzucchi SE, Van Ness PH, Gill TM, Kanaya A, Strotmeyer ES, Koster A, Johnson KC, Goodpaster BH, Harris T, et al. Elevated HbA1c and fasting plasma glucose in predicting diabetes incidence among older adults: are two better than one? Diabetes Care. 2013;36(12):3923–9.

    PubMed  PubMed Central  CAS  Google Scholar 

  42. Stumvoll M, Goldstein BJ, van Haeften TW. Type 2 diabetes: principles of pathogenesis and therapy. Lancet. 2005;365(9467):1333–46.

    PubMed  CAS  Google Scholar 

  43. DeFronzo RA, Abdul-Ghani MA. Preservation of β-cell function: the key to diabetes prevention. J Clin Endocr Metab. 2011;96(8):2354–66.

    PubMed  CAS  Google Scholar 

  44. Duncan BB, Schmidt MI, Pankow JS, Ballantyne CM, Couper D, Vigo A, Hoogeveen R, Folsom AR, Heiss G. Low-grade systemic inflammation and the development of type 2 diabetes: the atherosclerosis risk in communities study. Diabetes. 2003;52(7):1799–805.

    PubMed  CAS  Google Scholar 

  45. Someya Y, Tamura Y, Suzuki R, Kaga H, Kadowaki S, Sugimoto D, Kakehi S, Funayama T, Furukawa Y, Takeno K, et al. Characteristics of glucose metabolism in underweight japanese women. J Endocr Soc. 2018;2(3):279–89.

    PubMed  PubMed Central  CAS  Google Scholar 

  46. Srikanthan P, Karlamangla AS. Relative muscle mass is inversely associated with insulin resistance and prediabetes. Findings from the third National Health and Nutrition Examination Survey. J Clin Endocr Metab. 2011;96(9):2898–903.

    PubMed  CAS  Google Scholar 

  47. Hashimoto Y, Osaka T, Fukuda T, Tanaka M, Yamazaki M, Fukui M. The relationship between hepatic steatosis and skeletal muscle mass index in men with type 2 diabetes. Endocr J. 2016;63(10):877–84.

    PubMed  CAS  Google Scholar 

  48. Kim TN, Park MS, Lee EJ, Chung HS, Yoo HJ, Kang HJ, Song W, Baik SH, Choi KM. Comparisons of three different methods for defining sarcopenia: an aspect of cardiometabolic risk. Sci Rep-UK. 2017;7(1):6491.

    Google Scholar 

  49. Furushima T, Miyachi M, Iemitsu M, Murakami H, Kawano H, Gando Y, Kawakami R, Sanada K. Comparison between clinical significance of height-adjusted and weight-adjusted appendicular skeletal muscle mass. J Physiol Anthropol. 2017;36(1):15.

    PubMed  PubMed Central  Google Scholar 

  50. Takamura T, Kita Y, Nakagen M, Sakurai M, Isobe Y, Takeshita Y, Kawai K, Urabe T, Kaneko S. Weight-adjusted lean body mass and calf circumference are protective against obesity-associated insulin resistance and metabolic abnormalities. Heliyon. 2017;3(7): e347.

    Google Scholar 

  51. Scott D, Park MS, Kim TN, Ryu JY, Hong HC, Yoo HJ, Baik SH, Jones G, Choi KM. Associations of low muscle mass and the metabolic syndrome in Caucasian and Asian middle-aged and older adults. J Nutr Health Aging. 2016;20(3):248–55.

    PubMed  CAS  Google Scholar 

  52. Park BS, Yoon JS. Relative skeletal muscle mass is associated with development of metabolic syndrome. Diabetes Metab J. 2013;37(6):458–64.

    PubMed  PubMed Central  Google Scholar 

  53. Osaka T, Hashimoto Y, Okamura T, Fukuda T, Yamazaki M, Hamaguchi M, Fukui M. Reduction of fat to muscle mass ratio is associated with improvement of liver stiffness in diabetic patients with non-alcoholic fatty liver disease. J Clin Med. 2019;8(12):2175.

    PubMed  PubMed Central  CAS  Google Scholar 

  54. Shida T, Oshida N, Oh S, Okada K, Shoda J. Progressive reduction in skeletal muscle mass to visceral fat area ratio is associated with a worsening of the hepatic conditions of non-alcoholic fatty liver disease. Diabet Metab Syndr Obes. 2019;12:495–503.

    CAS  Google Scholar 

  55. Fontana L, Eagon JC, Trujillo ME, Scherer PE, Klein S. Visceral fat adipokine secretion is associated with systemic inflammation in obese humans. Diabetes. 2007;56(4):1010–3.

    PubMed  CAS  Google Scholar 

  56. Baxmann AC, Ahmed MS, Marques NC, Menon VB, Pereira AB, Kirsztajn GM, Heilberg IP. Influence of muscle mass and physical activity on serum and urinary creatinine and serum cystatin C. Clin J Am Soc Nephrol. 2008;3(2):348–54.

    PubMed  PubMed Central  CAS  Google Scholar 

  57. Nishida K, Hashimoto Y, Kaji A, Okamura T, Sakai R, Kitagawa N, Osaka T, Hamaguchi M, Fukui M. Creatinine/(cystatin C × body weight) ratio is associated with skeletal muscle mass index. Endocr J. 2020;67(7):733–40.

    PubMed  CAS  Google Scholar 

  58. Mo Z, Hu H, Du X, Huang Q, Chen P, Lai L, Yu Z. Association of evaluated glomerular filtration rate and incident diabetes mellitus: a secondary retrospective analysis based on a Chinese cohort study. Front Med-Lausanne. 2021;8: 724582.

    PubMed  Google Scholar 

  59. Lin CY, Hsieh MC, Kor CT, Hsieh YP. Association and risk factors of chronic kidney disease and incident diabetes: a nationwide population-based cohort study. Diabetologia. 2019;62(3):438–47.

    PubMed  Google Scholar 

  60. Pham H, Robinson-Cohen C, Biggs ML, Ix JH, Mukamal KJ, Fried LF, Kestenbaum B, Siscovick DS, de Boer IH. Chronic kidney disease, insulin resistance, and incident diabetes in older adults. Clin J Am Soc Nephro. 2012;7(4):588–94.

    CAS  Google Scholar 

  61. Weng J, Ji L, Jia W, Lu J, Zhou Z, Zou D, Zhu D, Chen L, Chen L, Guo L, et al. Standards of care for type 2 diabetes in China. Diabetes-Metab Res. 2016;32(5):442–58.

    Google Scholar 

  62. Yip W, Sequeira IR, Plank LD, Poppitt SD. Prevalence of pre-diabetes across ethnicities: a review of Impaired Fasting Glucose (IFG) and Impaired Glucose Tolerance (IGT) for classification of dysglycaemia. Nutrients. 2017;9(11):1273.

    PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Due to the fact that this is a secondary analysis, the majority of the data and method description are drawn from the following study: Chen Y, et al. Association of body mass index and age with incident diabetes in Chinese adults: a population-based cohort study. BMJ Open. 2018 Sep 28;8(9):e021768. https://doi.org/10.1136/bmjopen-2018-021768. All authors of this study are grateful for their contributions.

Funding

This study were supported by Shenzhen Key Medical Discipline Construction Fund (SZXK009), Shenzhen Municipal Science and Technology Innovation Commission for Sustainable Development (KCXFZ20201221173213036), Shenzhen High-level Hospital Construction Fund.

Author information

Author notes

  1. Tong Li and Changchun Cao have contributed equally to this work.

Authors and Affiliations

  1. Department of Nephrology, Shenzhen Second People’s Hospital, Shenzhen, Guangdong Province, 518000, China

    Tong Li

  2. Department of Nephrology, The First Affiliated Hospital of Shenzhen University, Shenzhen, Guangdong Province, 518000, China

    Tong Li

  3. Department of Rehabilitation, Shenzhen Dapeng New District Nan’ao People’s Hospital, Shenzhen, Guangdong Province, 518000, China

    Changchun Cao

  4. Department of Rheumatology, Shenzhen Second People’s Hospital, Shenzhen, Guangdong Province, 518000, China

    Xuan Xuan

  5. Department of Rheumatology, The First Affiliated Hospital of Shenzhen University, Shenzhen, Guangdong Province, 518000, China

    Xuan Xuan

  6. Department of Geriatrics, Shenzhen Second People’s Hospital, No.3002 Sungang Road, Futian District, Shenzhen, Guangdong Province, 518000, China

    Wenjing Liu, Xiaohua Xiao & Cuimei Wei

  7. Department of Geriatrics, The First Affiliated Hospital of Shenzhen University, Shenzhen, Guangdong Province, 518000, China

    Wenjing Liu, Xiaohua Xiao & Cuimei Wei

Authors
  1. Tong Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Changchun Cao
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xuan Xuan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Wenjing Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xiaohua Xiao
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Cuimei Wei
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Tong Li, Changchun Cao, Wenjing Liu and Xuan Xuan conceived the research, drafted the manuscript, and performed statistical analysis. Tong Li, Cuimei Wei and Xiaohua Xiao revised the manuscript and designed the study. All authors have read and approved the final manuscript.

Corresponding authors

Correspondence to Xiaohua Xiao or Cuimei Wei.

Ethics declarations

Ethics approval and consent to participate

The Rich Healthcare Group Review Board reviewed and approved studies involving human participants, and retrospective information was retrieved. It was conducted in accordance with the ethical principles of the Declaration of Helsinki. The data are anonymous, and the Rich Healthcare Group Review Board waived the requirement for informed consent due to the study's observational nature, as reported elsewhere [22].

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The original version of this article was revised: The funding information has been revised after publication.

Supplementary Information

Additional file 1:

Table S1. Collinearity diagnostics steps. Table S2. Relationship between Cre, weight and the incident diabetes in different models. Table S3. The result of the two-piecewise Cox regression model in participants without prediabetes. Figure S1. The non-linear relationship between the Cre/BW ratio and the risk of diabetes in participants without prediabetes.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, T., Cao, C., Xuan, X. et al. The association between creatinine to body weight ratio and the risk of progression to diabetes from pre-diabetes: a 5-year cohort study in Chinese adults. BMC Endocr Disord 23, 266 (2023). https://doi.org/10.1186/s12902-023-01518-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12902-023-01518-9

Keywords

BMC Endocrine Disorders

ISSN: 1472-6823

Contact us