Objective To investigate the associations of plasma aluminum and manganese levels and the environmental risk score (ERS) with the incidence of cardiovascular diseases (CVD) in the rural population of southern Xinjiang Uygur Autonomous Region, thereby providing new insights for the prevention of CVD in this population.

Methods  A typical sampling method was used to enroll 12 794 rural residents ≥ 18 years in a regiment of a division in southern Xinjiang Uygur Autonomous Region from August to September 2016 for the baseline survey, and follow-up surveys were carried out in July 2019, July 2020, August 2021, and June 2022. During the follow-up period, 761 newly diagnosed CVD cases were enrolled in the CVD group. In addition, 761 participants without CVD and with the age differences ≤ 3 years from the CVD cases were enrolled in the control group for a nested case-control study. Plasma levels of aluminum and manganese were measured by inductively coupled plasma mass spectrometry. An adaptive elastic net model was employed to construct the ERS. Conditional multivariate logistic regression was employed to assess the associations of plasma aluminum and manganese levels and ERS with the risk of CVD. Restricted cubic splines were used to explore the dose-response relationship between ERS and CVD risk.

Results  The plasma aluminum and manganese levels in the CVD group were 31.46 (22.82, 46.13) μg/L and 2.71 (1.85, 4.14) μg/L, respectively. Corresponding levels in the control group were 29.97 (21.98, 42.46) μg/L and 2.37 (1.66, 3.70) μg/L, respectively. The CVD group had higher plasma aluminum and manganese levels than the control group. After adjusting for age, sex, educational background, smoking, alcohol consumption, estimated glomerular filtration rate (eGFR), obesity, hypertension, diabetes, dyslipidemia, and family history of CVD, the conditional multivariate logistic regression showed that elevated plasma levels of both aluminum (OR = 1.17, 95%CI: 1.04–1.32) and manganese (OR = 1.18, 95%CI: 1.05–1.32) were associated with an increased risk of CVD. When the plasma manganese level was used as a categorical variable, the risk of CVD in the high plasma manganese group was 1.58 times of that in the low plasma manganese group (95%CI: 1.26–1.97). Moreover, the plasma aluminum level had a positive linear dose-response relationship with the risk of CVD (P_overall = 0.022, P_nonlinear = 0.112). Following inclusion of the linear term for aluminum concentration, linear term for manganese concentration, quadratic term for aluminum, quadratic term for manganese, and the aluminum-manganese interaction term into the adaptive elastic net model, the linear term for manganese (β = 0.02, 95%CI: –0.05–0.10), the quadratic term for aluminum (β = 0.01, 95%CI: 0.00–0.02), and the aluminum-manganese interaction term (β = 0.01, 95%CI: –0.01–0.03) were retained. ERS was calculated by summing the products of the metal terms retained in the adaptive elastic net model and their corresponding regression coefficients. An increase in ERS was associated with an elevated risk of CVD (OR = 1.20, 95%CI: 1.08–1.35). Furthermore, a positive linear dose-response relationship was observed between ERS and CVD risk (P_overall = 0.021, P_nonlinear = 0.751).

Conclusions  Elevated plasma levels of aluminum and manganese in the rural population and an elevated ERS in southern Xinjiang Uygur Autonomous Region were associated with an increased risk of CVD.