Objective To understand the relationship between masticatory function and diabetes mellitus in middle-aged and older adults, along with the mediating role of dietary factors therein, thus providing a basis for research into the mechanisms by which oral health influences metabolic diseases and for the early intervention of diabetes mellitus.

Methods Data were collected from the China Development and Nutrition Health Impact Cohort Survey (2022–2023) and the China Health and Nutrition Survey across ten regions: Heilongjiang province, Liaoning province, Shandong province, Henan province, Hunan province, Hubei province, Jiangsu province, Shaanxi province, Guangxi Zhuang Autonomous Region, and Guizhou province. This study utilized comprehensive demographic and epidemiological data from 2 509 middle-aged and older adults aged ≥55 years, encompassing masticatory function, dietary intake, medical history, physical measurements, and blood biochemical test data. A multivariable unconditional logistic regression model was employed to analyze the relationship between masticatory function and the prevalence of diabetes mellitus. Mediating effect models were further employed to examine the mediating effects of different food categories.

Results Among the 2 509 middle-aged and older adults included in the analysis, 805 (32.08%) exhibited masticatory dysfunction, while 516 (20.57%) had diabetes mellitus. Among the entire cohort, middle-aged and older adults with normal masticatory function consumed higher amounts of other cereals and pulses, tubers, vegetables, fruits, eggs, beans and bean products, nuts, fish and shellfish, meat, and cooking oil than those with masticatory dysfunction. Male middle-aged and older adults with normal masticatory function consumed higher amounts of other cereals and pulses, tubers, fruits, meat, and cooking oil than those with masticatory dysfunction. Female middle-aged and older adults with normal masticatory function consumed higher amounts of fruits, eggs, legumes and their products, nuts, fish and shellfish, and meat than those with masticatory dysfunction (all P < 0.05). After adjusting for confounding factors including gender, age, educational attainment, place of residence, employment status, annual household income per capita, smoking status, alcohol consumption, total physical activity, presence of hypertension, presence of dyslipidaemia, body mass index, and intake of various food categories, the results from unconditional multivariable logistic regression analysis revealed that middle-aged and older adults with masticatory dysfunction exhibited a 28.8% higher likelihood of developing diabetes mellitus than those with normal masticatory function (OR = 1.288, 95%CI: 1.029–1.612). After stratification by gender, female middle-aged and older adults with masticatory dysfunction had a 39.8% higher likelihood of developing diabetes mellitus than those with normal masticatory function (OR = 1.398, 95%CI: 1.033–1.894). Mediating effect analysis revealed that fruit intake exerted a mediating effect on the relationship between masticatory function and diabetes mellitus prevalence in the entire cohort (ab = 0.005, P = 0.032) and among female middle-aged and older adults (ab = 0.006, P = 0.018). Legume and legume product intake also mediated the relationship between masticatory function and diabetes mellitus prevalence among female middle-aged and older adults (ab = 0.006, P = 0.045).

Conclusions Masticatory dysfunction in middle-aged and older adults is associated with an increased likelihood of diabetes mellitus, with fruit intake serving as an important mediating factor.