LAUSR

Background: This study aimed to evaluate the association of serum L-carnitine with first stroke and explore potential effect modifiers. Methods: This is a nested, case-control study drawn from the China Stroke Primary Prevention Trial among rural Chinese adults with hypertension, including 557 first stroke cases and 557 age-matched, sex-matched, treatment group–matched, and residence-matched controls. Serum L-carnitine was measured by liquid chromatography with tandem quadrupole mass spectrometry. Multiple conditional logistic regression models were used to evaluate the association between L-carnitine and first stroke. Results: The mean level of serum L-carnitine in the stroke population was 4.7 μg/mL, which was significantly lower than that of the control group (5.7 μg/mL). When L-carnitine was assessed as quintiles, compared with the reference group (quintile 1, <3.3 μg/mL), the odds of stroke were 0.62 (95% CI, 0.39–1.00) in quintile 2, 0.66 (95% CI, 0.40–1.10) in quintile 3, 0.47 (95% CI, 0.28–0.81) in quintile 4, and 0.50 (95% CI, 0.30–0.84) in quintile 5. The trend test was significant (P=0.01). When quintiles 2 to 5 were combined, the adjusted odds ratio of first stroke was 0.58 (95% CI, 0.38–0.87) compared with quintile 1. Similar associations were found for ischemic stroke and hemorrhagic stroke. In subgroup analysis, a significant L-carnitine–stroke association was only observed in the normal folate group (P interaction, 0.039) and in the MTHFR CC genotype group (P interaction, 0.047). Conclusions: In this study of rural Chinese adults with hypertension, serum L-carnitine had an inverse but nonlinear association with first stroke. Folate status and the MTHFR C677T variant were significant effect modifiers of the association.