Association Between Dietary Protein Intake and Abdominal Obesity in Rural Adults: A Cross-Sectional Study

Objectives: Dietary protein intake has been suggested to be inversely associated with measures of adiposity, but evidence regarding its association with abdominal obesity remains limited. This study examined the association between different levels of dietary protein intake and prevalent abdominal obesity in a rural adult population, aiming to provide descriptive evidence for this specific rural population. Methods: This was a cross-sectional analysis of data from a previous nutritional epidemiological survey conducted among 515 rural adults. Dietary intake was assessed with a semi-quantitative food frequency questionnaire combined with a 3-day 24-h dietary recall. Anthropometric measurements comprised body weight, height, waist circumference and hip circumference. Participants were divided into quartiles of protein intake: Q1 (<0.74 g/(kg·d)), Q2 (0.74–0.97 g/(kg·d)), Q3 (0.97–1.23 g/(kg·d)) and Q4 (≥1.23 g/(kg·d)). Abdominal obesity was defined as a waist circumference ≥ 90 cm in men and ≥85 cm in women. Continuous variables were summarized as medians (interquartile range) and compared using the Kruskal–Wallis H or Mann–Whitney U test; categorical variables were expressed as frequencies (percentages) and compared using the χ2 test. Multivariable logistic regression was used to examine the association, with adjustment for potential confounders. Results: With increasing protein intake (Q1 to Q4), body weight, body mass index (BMI), waist circumference, hip circumference and waist-to-hip ratio showed significant decreasing trends (p < 0.05), whereas total energy and macronutrient intake and the consumption of several food groups increased significantly (p < 0.05). Protein intake was lower in the abdominal obesity group (n = 374) than in the normal waist circumference group (0.88 vs. 1.19 g/(kg·d); p < 0.001). In the logistic regression models, higher protein intake was inversely associated with the odds of abdominal obesity, with a graded association across quartiles (trend p < 0.001). In the unadjusted model, Q4 had lower odds than Q1 (odds ratio [OR] = 0.128, 95% confidence interval [CI]: 0.065–0.250); after adjustment for energy intake the association was stronger (Q4: OR = 0.008, 95% CI: 0.003–0.023), and after further adjustment for age, sex, energy intake, lifestyle and dietary factors it remained (Q4: OR = 0.006, 95% CI: 0.002–0.018, p < 0.001). In sensitivity analyses using absolute protein intake (g/day), protein as a percentage of energy and a continuous per-SD model, these associations were markedly attenuated and none remained statistically significant (all p > 0.05). Conclusions: Although protein intake expressed as g/(kg·d) appeared to be inversely associated with prevalent abdominal obesity in this rural sample, with a graded pattern across quartiles, this association was largely explained by mathematical coupling arising from body-weight normalization. Using metrics independent of body weight, no significant association remained; these findings should therefore be interpreted as associations rather than causal effects, and prospective studies using alternative protein metrics are needed.