Study design and participants

A cross-sectional study was conducted using baseline data collected from the Fuqing cohort study from July 2020 to June 2021 in Gaoshan town. The Fuqing cohort study is a prospective multi-purpose research program conducted by the Fujian Medical University and the local government of Fuqing City, located in Fujian province, Southeast China. The Fuqing cohort study investigates the natural history and risk factors of chronic non-communicable diseases, including cancer, diabetes, and fatty liver, among the Chinese population residing in the Southeast coastal region of China. Detailed information on sociodemographics (age, sex, level of education, and employment status), lifestyle factors (smoking, alcohol drinking, and tea drinking), history of chronic diseases, and history of taking medicine was collected through face-to-face questionnaires by trained interviewers. In addition, most participants underwent physical examination and provided biological samples. The inclusion criteria for this study were as follows: (1) aged 35-75 years; (2) residents of Fuqing city. The exclusion criteria were as follows: (1) participants with a history of cancer or type 1 diabetes mellitus; (2) participants with hypoglycemia at baseline. A total of 6790 participants were initially included. In addition, we excluded 177 participants with missing values for waist circumference, blood parameters, or glycemic status indicators. Finally, a total of 6,613 subjects were included in the final analysis. This study was approved by the ethical committee of Fujian Medical University (Approval number, 2017–07 and 2020–58), and the study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and all participants provided written informed consent.

Definition of central adiposity

Anthropometric measurements including waist and hip circumference were obtained using standardized methods by trained examiners. Waist circumference was measured in centimeters with a nonstretchable tape held at the midpoint between the lower margin of the least palpable rib and the top of the iliac crest, and hip circumference was measured around the widest portion of the buttocks. WHR was calculated as waist circumference/hip circumference. According to the World Health Organization (WHO) cut-off points for substantially increased risk of metabolic complications, central adiposity was defined as WHR ≥ 0.90 in men and ≥ 0.85 in women [15].

Measurement of peripheral WBC counts

Venous blood sample of each participant was obtained. WBC including neutrophil, lymphocyte, monocyte, eosinophil, and basophil were measured using standard laboratory procedures (fully automated blood cell analyzer, Xs-1000i, Sysmex, Osaka, Japan). Total WBC was defined as the sum of all five types of white blood cells.

Definition of normal glucose, prediabetes, and diabetes

According to the WHO criteria [16], participants were classified into three groups: normoglycemia, prediabetes, and diabetes. Prediabetes was defined as fasting blood glucose 110 mg/dL (6.1 mmol/L) to 125 mg/dL (6.9 mmol/ L), and oral glucose tolerance test: two hour blood glucose <140 mg/dL (7.8 mmol/L), or fasting blood glucose <126 mg/dL (7.0 mmol/L), and oral glucose tolerance test: two-hour blood glucose 140 mg/dL (7.8 mmol/L) to 199 mg/dL (11.0 mmol/L) in participants without a history of diabetes. Diabetes was defined as a previous physician diagnosis of diabetes, or current use of diabetes medication, or having one or more of the following: (1) fasting blood glucose ≥126 mg/dL (7.0 mmol/L); (2) oral glucose tolerance test: two-hour blood glucose ≥200 mg/dL (11.1 mmol/L).

Assessment of covariates

Demographic data, including age, sex, education, occupation, as well as, lifestyle behaviors (smoking, alcohol drinking, tea drinking), status of chronic disease (hypertension and hyperlipidemia), and history of taking lipid-lowering drugs were collected via standardized questionnaires at baseline. Hypertension was defined as one or more of the following: (1) self-reported or doctor-diagnosed hypertension, (2) current treatment with antihypertensive agents, (3) systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg. Hyperlipidemia was defined as one or more of the following: total cholesterol (TC) ≥ 6.2 mmol/L, triglyceride (TG) ≥ 2.3 mmol/L, high-density lipoprotein cholesterol (HDL-c) <1.0 mmol/L, or low-density lipoprotein cholesterol (LDL-c) ≥4.1 mmol/L.

Sensitivity analysis

Sensitivity analysis was performed by excluding participants who used antibiotic medication for infection in the past 6 months, and participants with abnormal white blood cell count (<2.5% or > 97.5% of the distribution) to estimate the association between WBC counts and glycemic status, and the mediation effect of WBC counts in the association between WHR and glycemic status.

Statistical analysis

Demographic and clinical characteristics were first presented for the normal glycemic group, prediabetic, and diabetic groups. Categorical variables were expressed as numbers (percentages) and tested for association with glycemic status using Chi-squared tests. Continuous variables were presented as medians and interquartile ranges (IQR) and tested using Kruskal-Wallis rank-based tests, owing to their non-normal distributions.

In the analysis investigating the effect of WHR, or WBC counts on the glycemic status, we use age, sex, education, occupation, smoking, alcohol drinking, tea drinking, hypertension, hyperlipidemia, and lipid-lowering drugs as covariates. Separate logistic regression models were used to evaluate the associations of WHR or WBC counts with prediabetes and diabetes. Furthermore, Spearman’s rank correlation analysis was conducted to quantify the relationship between WHR and WBC counts. Statistical analyses were conducted by using Stata (version 16.0).

Mediation analysis was also conducted to explore whether the effect of WHR on diabetes was mediated by white blood cells, which include neutrophils, lymphocytes, monocytes, eosinophils, and basophils. Both simple and multiple mediation analyses were employed to assess the roles of different WBC counts in the association between WHR and diabetes. In the mediation analysis, we decomposed the total effects into natural direct and indirect effects [17, 18] using CMAverse [19], an R package that provides a suite of functions for causal mediation analysis. The 95% confidence intervals (CIs) for indirect effects were evaluated using bias-corrected CIs from 10000 bootstrapping samples (Rstudio version 4.2.2).