Association of dietary patterns with the newly diagnosed diabetes mellitus and central obesity: a community based cross-sectional study

Diabetes mellitus (DM), particularly type 2 diabetes (T2DM), has risen markedly in China, with adult prevalence increasing from 0.67% in 1980 to 10.9% in 2013. T2DM is closely linked to insulin resistance (IR) and is more prevalent among individuals with overweight, obesity, and especially central obesity. Visceral adipose tissue, compared with subcutaneous fat, releases more proinflammatory and proatherogenic factors, exacerbating oxidative stress and IR. Dietary patterns have shifted in China with rapid socioeconomic changes, and Western-style dietary patterns have been implicated in weight gain and increased body fat. Traditional single-nutrient approaches are limited; dietary pattern analysis considers combined food effects. Prior studies suggest vegetables/fruits and low-fat dairy may reduce IR, whereas Western patterns increase glucose intolerance. The study aimed to identify major dietary patterns among Chinese adults aged 40–65 years and evaluate their associations with body fat distribution (particularly visceral fat) and newly diagnosed DM.

Prior research indicates that diet quality and patterns influence IR, obesity, and T2DM risk. Vegetables and fruits, low-fat dairy, and moderate alcohol consumption have been associated with reduced IR, while Western-style patterns high in energy and fat and low in fiber increase glucose intolerance. Few studies have examined dietary patterns and body fat distribution (visceral vs subcutaneous). Evidence links fast food and sugar-sweetened beverages to obesity and T2DM, whereas higher fruit/vegetable intake is linked to lower T2DM risk and improved adiposity metrics. Findings for red meat, rice, eggs, and fish are mixed: red meat and high–glycemic index rice have been associated with higher T2DM risk in some cohorts, but null associations also exist, potentially due to nutrients like zinc. Associations of eggs and fish with T2DM vary by population and preparation methods, suggesting geographical and cultural heterogeneity.

Design and participants: Community-based cross-sectional survey conducted March–May 2010 in Caihe communities, Hangzhou, Zhejiang, China. Included 1432 Han Chinese adults aged 40–65 years. Exclusions: prior stroke or ischemic heart disease; previously diagnosed DM. Ethical approval obtained; informed consent collected. Demographics/lifestyle: Questionnaire captured smoking (current, former, never), alcohol use (drinker defined as ≥1/week over last 4 weeks), physical activity (light/middle/heavy), education, and income. Dietary assessment: Validated semi-quantitative FFQ (81 items) administered by trained dietitians assessed intake over previous 4 weeks. Frequency categories ranged from never to ≥2 times/day. Eighty-one items were grouped into 21 predefined food groups based on nutrient profiles and culinary use. Intakes were calculated using China Food Composition Table (2009). Laboratory measurements: Participants fasted after 20:00 and underwent OGTT (75 g). Blood samples at 0 and 2 h measured glucose, TG, TC, HDL-c, LDL-c (autoanalyzer), HbA1c (HPLC), insulin (radioimmunoassay). HOMA-IR = [fasting insulin (mU/L) × fasting glucose (mmol/L)]/22.5. Anthropometry and imaging: BMI, waist circumference (WC), hip circumference, waist-to-hip ratio (WHR). Body fat percentage by bioelectrical impedance (Tanita TBF-300). Blood pressure measured as mean of three readings. Abdominal adiposity assessed by MRI (SMT-100; spin-echo TR-500, TE-200; 256×256 matrix) at L4–L5 to quantify visceral fat area (VFA) and subcutaneous fat area (SFA). Definitions: Newly diagnosed DM per WHO criteria: fasting plasma glucose ≥7.0 mmol/L and/or 2-h glucose ≥11.1 mmol/L on OGTT in those without prior DM. Central obesity defined as VFA ≥80 cm². Statistical analysis: Continuous variables presented as mean ± SD (normal) or median (range) with log-transformation for skewed variables (glucose, insulin, HOMA-IR, TG, VFA, SFA, total energy intake). Sweet-fast pattern scores underwent log(x+10) transformation. Group comparisons by Student’s t-test (continuous) and Pearson’s χ² (categorical). Dietary patterns derived via principal components factor analysis among 21 food groups with varimax (orthogonal) rotation. Factor retention based on eigenvalues ≥1.5, scree plot, and interpretability; items retained if absolute loading ≥0.25. Four patterns were selected, explaining 12.33%, 12.29%, 8.41%, and 6.07% of variance, respectively. Associations between pattern scores and HOMA-IR and VFA assessed using partial correlation and linear regression adjusting for age, smoking, drinking, education, income, total energy intake, physical activity, and BMI for HOMA-IR; for VFA models, BMI was not listed among adjustments (age, smoking, drinking, education, income, total energy intake, physical activity). Pattern scores were categorized into tertiles for multivariable logistic regression estimating odds ratios (ORs) and 95% CIs for DM and central obesity. Multivariable models for DM adjusted for age, smoking, drinking, education, income, total energy intake, physical activity, and BMI; for central obesity adjusted for age, smoking, drinking, education, income, total energy intake, and physical activity. Analyses used SPSS v16.0; two-sided p < 0.05 considered significant.

- Sample: 575 males and 857 females; prevalence of newly diagnosed DM: 10.6% males, 6.1% females. DM cases had higher WC, WHR, body fat percentage, FPG, 2h-PG, FINS, HOMA-IR, HbA1c, HDL-c (lower in some cases), and VFA than controls. - Dietary patterns identified: 1) Vegetables-fruits (vegetables, beans, mushrooms, tubers, fruits, coarse cereals, seaweeds, wheat, nuts, dairy); 2) Rice-meat (red meat, white rice, poultry, organ meat, condiments, eggs, beans); 3) Seafood-eggs (eggs, seafood, dairy, nuts, fruits, tea; inverse beverages); 4) Sweet-fast food (fast foods, beverages, desserts, alcoholic beverages; inverse white rice, dairy). Variance explained by each pattern: 12.33%, 12.29%, 8.41%, and 6.07%, respectively (total ≈39.1%). - Correlations with insulin resistance (HOMA-IR), adjusted: Vegetables-fruits inversely associated (males R = -0.250, p < 0.001; females R = -0.152, p < 0.001). Sweet-fast positively associated (males R = 0.162, p = 0.002; females R = 0.225, p < 0.001). Rice-meat and seafood-eggs not significant. - Correlations with visceral fat area (VFA), adjusted: Vegetables-fruits inversely associated (males R = -0.111, p = 0.029; females R = -0.099, p = 0.017). Sweet-fast positively associated in males (R = 0.200, p < 0.001) but not females (R = 0.012, p = 0.775). Rice-meat and seafood-eggs not significant. - Odds of DM across tertiles (multivariable adjusted): Vegetables-fruits T3 vs T1 associated with lower DM risk: males OR 0.28 (95% CI 0.12–0.64; trend p = 0.005); females OR 0.28 (95% CI 0.11–0.72; trend p = 0.005). Sweet-fast T3 vs T1 associated with higher DM risk in males OR 2.58 (95% CI 1.23–5.88; trend p = 0.044); not significant in females (OR 0.73, 95% CI 0.29–1.85). Rice-meat and seafood-eggs patterns not significantly associated with DM in either sex. - Odds of central obesity across tertiles (multivariable adjusted): Vegetables-fruits T3 vs T1 associated with lower central obesity risk in males OR 0.50 (95% CI 0.29–0.86; trend p = 0.039); not significant in females (OR 1.48, 95% CI 0.93–2.38). Sweet-fast T3 vs T1 associated with higher central obesity risk in males OR 2.85 (95% CI 1.67–4.86; trend p < 0.001); not significant in females (OR 1.04, 95% CI 0.66–1.64). Rice-meat and seafood-eggs patterns not significantly associated with central obesity.

The study addressed whether common dietary patterns among middle-aged Chinese adults are associated with insulin resistance, newly diagnosed DM, and central adiposity. The vegetable-fruits pattern was consistently linked to lower HOMA-IR, lower visceral fat, and reduced odds of DM (both sexes) and central obesity (males). These findings align with proposed mechanisms: higher dietary fiber and lower glycemic index/load attenuate postprandial glycemia and improve insulin sensitivity; fruits and vegetables provide antioxidants that may mitigate oxidative stress related to IR. Conversely, the sweet-fast food pattern, characterized by energy-dense items with high glycemic load and low fiber/micronutrients, was linked to higher HOMA-IR and VFA and elevated odds of DM and central obesity in males, consistent with prior evidence connecting fast foods and sugary beverages with metabolic dysregulation. No associations were observed for rice-meat or seafood-eggs patterns with DM or central obesity, echoing mixed literature on red meat, refined rice, eggs, and fish in East Asian populations. Preparation methods, accompanying foods, and overall dietary context may account for heterogeneity. Sex differences were evident: associations were stronger or only present in males for central obesity and for the sweet-fast pattern’s risks. Potential explanations include sex-specific physiology, hormonal status (e.g., peri-/postmenopausal changes in women), and differing dietary exposures, though the study was not powered to definitively explain these differences. Overall, the findings suggest that promoting a vegetables-fruits-rich pattern and limiting sweet-fast foods may reduce IR, visceral adiposity, and incident DM in this demographic.

This community-based cross-sectional study identified four major dietary patterns in middle-aged Chinese adults and demonstrated that a vegetables-fruits pattern is associated with lower insulin resistance, visceral adiposity, and odds of newly diagnosed DM (both sexes) and central obesity (males), whereas a sweet-fast food pattern shows adverse associations, particularly in males. Rice-meat and seafood-eggs patterns were not significantly associated with DM or central obesity. These results support dietary guidance emphasizing vegetables, fruits, legumes, whole/coarse grains, nuts, and dairy while reducing fast foods, desserts, and sugary beverages to mitigate DM and central obesity risks. Future research should include longitudinal and interventional studies to establish causality, explore mechanisms underlying sex-specific effects, and assess generalizability across different Chinese regions and age groups.

- Cross-sectional design precludes causal inference between dietary patterns and DM or central obesity. - Factor analysis involves subjective decisions (food grouping, number of factors, rotation), which may influence identified patterns. - Potential residual confounding despite adjustment for multiple covariates. - Study population limited to middle-aged Han Chinese from a single urban area (Hangzhou, Zhejiang), limiting generalizability to other regions or age groups. - Possible low consumption of fast foods/desserts in this cohort may attenuate or bias associations. - Limited power for sex-stratified analyses and to investigate menopausal status effects among women.