In the cross-sectional study, based on a nationally representative sample from NHANES, we demonstrated significant association between serum copper level and increased prevalence of stroke in general US adults. The adjusted OR of self-report stroke for participants in the highest tertile was 2.36(95% CI: 1.01,5.52) compared to the subject with the lowest tertile. A postive linear relationship between serum copper and stroke further reinforced this associaiton, with per SD increment of serum copper increased the OR of stroke by 44% (95%CI: 1.11–1.86). Stratified analyses stratified by sex, age, BMI, diabetes, hypertension, eGFR, TC, smoking status, drinking and sensitivity analysis generated confirmatory results. It is the first nationally representative study that suggest the postive association between serum copper and stroke.
To date, several literatures have reported the association between serum copper and stroke. Our finding accorded with previous studies. Hu et al.[19] performed a nested case-control study with 2510 subjects derived from the Chinese community-based population (80.32% hypertensives) to show that plasma copper levels were associated with the risk of first stroke (Per SD increment, OR:1.17; 95% CI: 1.07–1.28). Using data from the CSPPT, another nested case-control study observed a positive association between baseline plasma copper and the risk of first stroke in Chinese hypertensive adults (Per SD increment, OR:1.20; 95% CI: 1.03–1.39)[25]. In a nested case-control study based on retired employees of Dongfeng Motor Corporation, a lower risk of stroke in subjects with serum copper levels in the highest tertile compared with serum copper in the lowest tertile was found(OR:1.53; 95% CI: 1.20–1.96)[26].
However, several reports have also yielded some conflicting results. A casecontrol study based on NHANES 2013–2018 dataset illustrated that increase copper intake was associatied with lower prevalence of self-report stroke(highest versus the lowest quintiles,OR: 0.6, 95%CI: 0.43–0.83)[27].Furthermore, no association was observed between serum copper and the risk of stroke in a nested case-control study based on 68 participants(OR: 2.26, 95%CI: 0.65–7.88)[20]. One case-control study with 141 patients with acute ischemic stroke and 69 healthy people from the northeastern region of Poland also showed no associations between the risk of ischemic stroke and serum copper[28]. The possible reasons for conflicting reslts might derive from copper subtype, sample size, characteristics of observational study, and adjustment of confounders. Our study suggested that serum copper is linearly associated with prevalence of self-report stroke. Our dataset was based on a large nationally sample, which free of selection bias and representative of diverse. Furthermore, various confounding factor was adjusted, including demographics, sociological characteristics, lifestyle and comorbidities.
Basic research also supported our results. A mice experiment demonstate that chronic intake copper in drinking water resulted in impairment endothelial progenitor cells and aggravated cerebral ischemic injury[29]. Several study demonstated that copper can catalyse the oxidative modification of LDL in vitro and also participate in the oxidation of LDL within the arterial wall which is a key event in human atherogenesis[30, 31].In addition, literatures reported that copper-homocysteine complexes was identified in vitro and endothelial cells elicited remarkable changes in relation to atherogenic activities when culture with copper-homocysteine complexes[32–35]. Further experiments are needed to elucidate the mechanism responsible for the association between copper and stroke in vitro and in vivo.
This research has several strengths. Firstly, it is the first study that illustrated the association between serum copper and stroke in US adults. Secondly, this study was based on NHANES, a national survey, with large sample size and good generalization of the civilian, non-institutionalized US adult population. Third, we showed serum copper is linearly associated with self-report stroke and stratified analyses show a stable result, which showed the robustness of our results.
Our study has several limitations. Firstly, we cannot draw a causal relationship between serum copper and stroke from the current cross-sectional study. Second, stroke events were determined as self-reported questionable, recall bias was inventible. Third, stroke outcomes in our study did not differentiate between hemorrhagic stroke and ischemic stroke. We can’t determine the association between serum copper and the subtype of stroke. Fourth, despite adjusting lots of confounding variables, there was still uncontrolled confounding because of unmeasured differences in behaviors or other factors.