Objective  To explore causal relationship between fasting plasma glucose (FPG) and blood pressure (BP) in rural normotensive population and to provide evidences for early prevention and control of hypertension and diabetes risk factors.

  Methods  We established a cardiovascular disease study cohort consisting of 4 145 normotensive residents aged ≥ 35 years selected with stratified multistage random cluster sampling in three counties of Liaoning province from January to August 2013 and conducted a follow-up survey among the residents from August 2015 to January 2016. The causal relationship between BP and FPG was assessed with cross-lagged panel analysis and further verified with multivariate linear regression model.

  Results  For all the participants, the mean follow-up time was 28.79 ± 3.49 months; the FPG at the baseline and follow-up survey were 5.51 ± 0.83 and 5.48 ± 1.07 mmol/L; the systolic blood pressure (SBP) were 124.05 ± 9.62 and 125.12 ± 14.63 mm Hg and the diastolic blood pressure (DBP) were 75.17 ± 7.24 and 75.88 ± 9.33 mm Hg at the baseline and follow-up survey, respectively. The cross-lagged panel analysis revealed that baseline FPG preceded the follow-up SBP (ρ₂ = 0.063, P < 0.001) and DBP (ρ₂ = 0.029, P = 0.031). The results of multivariate linear regression model further validated the findings resulted from the cross-lagged panel analysis that the baseline FPG affected follow-up SBP (β = 1.535, P < 0.001) and DBP (β = 0.393, P = 0.015). The mean values in the changes of SBP and DBP corresponding to the first to the fourth quartile of change values of FPG were – 0.79 and – 0.25 mm Hg, 0.74 and 0.49 mm Hg, 1.17 and 0.91 mm Hg, and 3.11 and 1.68 mm Hg, respectively; and the changes in the FPG value were positively correlated with the change of SBP and DBP (both P < 0.001).

  Conclusion  The relationship between FPG and blood pressure is unidirectional causation in rural normotensive population and the increase in blood glucose level precedes the increase in blood pressure.