Person:

Rexrode, Kathryn

Background: Magnesium is associated with lower risk of sudden cardiac death, possibly through antiarrhythmic mechanisms. Magnesium influences endothelial function, inflammation, blood pressure, and diabetes, but a direct relation with coronary heart disease (CHD) risk has not been established. Methods and Results: We prospectively examined the association between dietary and plasma magnesium and risk of CHD among women in the Nurses' Health Study. The association for magnesium intake was examined among 86 323 women free of disease in 1980. Information on magnesium intake and lifestyle factors was ascertained every 2 to 4 years through questionnaires. Through 2008, 3614 cases of CHD (2511 nonfatal/1103 fatal) were documented. For plasma magnesium, we conducted a nested case–control analysis, with 458 cases of incident CHD (400 nonfatal/58 fatal) matched to controls (1:1) on age, smoking, fasting status, and date of blood sampling. Higher magnesium intake was not associated with lower risk of total CHD (P‐linear trend=0.12) or nonfatal CHD (P‐linear trend=0.88) in multivariable models. However, magnesium intake was inversely associated with risk of fatal CHD. The RR comparing quintile 5 to quintile 1 of magnesium intake was 0.61 (95% CI, 0.45 to 0.84; P‐linear trend=0.003). The association between magnesium intake and risk of fatal CHD appeared to be mediated partially by hypertension. Plasma magnesium levels above 2.0 mg/dL were associated with lower risk of CHD, although not independent of other cardiovascular biomarkers (RR, 0.67; 95% CI, 0.44 to 1.04). Conclusions: Dietary and plasma magnesium were not associated with total CHD incidence in this population of women. Dietary magnesium intake was inversely associated with fatal CHD, which may be mediated in part by hypertension.

Background: Fetuin‐A may be involved in the etiology of coronary heart disease (CHD) through opposing pathways (ie, promoting insulin resistance and inhibiting ectopic calcification). We aimed to explicitly examine whether systemic inflammation, a factor leading to elevated vascular calcification, may modify the association between fetuin‐A and CHD risk. Method and Results During 16 years of follow‐up (1990–2006), we prospectively identified and confirmed 466 incident fatal or nonfatal CHD case in the Nurses' Health Study. For each case, 1 healthy control was selected using risk‐set sampling from 26 245 eligible participants. Cases and controls were matched for age, smoking status, fasting status, and date of blood draw. After multivariate adjustment for lifestyle factors, body mass index, diet, and blood lipids, fetuin‐A levels were not associated with CHD risk in the whole population: odds ratio (OR) (95% CI) comparing extreme quintiles of fetuin‐A was 0.79 (0.44 to 1.40). However, a significant inverse association was observed among participants with higher C‐reactive protein levels (Pinteraction=0.04). The OR (95% CI) comparing highest versus lowest quintiles of fetuin‐A was 0.50 (0.26 to 0.97; Ptrend=0.004) when C‐reactive protein levels were above population median (0.20 mg/dL), whereas among the remainder of the participants, the corresponding OR (95% CI) was 1.09 (0.58 to 2.05; Ptrend=0.75). Conclusions: In this population of US women, fetuin‐A levels were associated with lower CHD risk when C‐reactive protein levels were high, but null association was observed among participants with lower C‐reactive protein levels. This divergent pattern of association needs replication in future studies.

Background: Ambient air pollution exposures have been frequently linked to cardiovascular disease (CVD) morbidity and mortality. However, less is known about the populations most susceptible to these adverse effects. Methods and Results: We assessed the associations of long‐term particulate matter (PM) exposures with incident CVD in a nationwide cohort of 114 537 women in the Nurses’ Health Study, and performed analyses to identify subpopulations at the greatest risk. Residential address level time‐varying monthly exposures to PM 2.5, PM 10, and PM 2.5 to 10 microns in diameter were estimated from spatio‐temporal prediction models. In multivariable models, increases in all size fractions of PM were associated with small, but not statistically significant, increased risks of total CVD, coronary heart disease, and stroke. PM‐associated CVD risks were statistically significantly higher among women with diabetes as compared to those without (P‐for‐interaction <0.0001 for PM 10 and PM 2.5 and 0.007 for PM 2.5 to 10). For each 10 μg/m3 increase in 12‐month average PM 2.5, PM 2.5 to 10, and PM 10, the multivariable adjusted hazard ratios were 1.44 (95% CI: 1.23 to 1.68), 1.17 (95% CI: 1.05 to 1.30), and 1.19 (95% CI: 1.10 to 1.28) among women with diabetes. There were also suggestions of higher risks among older (≥70 years) women, the obese, and those living in the Northeast and South. Smoking status and family history did not consistently modify the association between PM and CVD, and risks were most elevated with exposures in the previous 12 months. Conclusions: In this nationwide cohort, women with diabetes were identified as the subpopulation most sensitive to the adverse cardiovascular health effects of PM.

Background: Clinical practice focuses on the primary prevention of cardiovascular (CV) disease (CVD) through the modification and pharmacological treatment of elevated risk factors. Prediction models based on established risk factors are available for use in the primary prevention setting. However, the prevention of risk factor development through healthy lifestyle behaviors, or primordial prevention, is of paramount importance to achieve optimal population‐wide CV health and minimize long‐term CVD risk. Methods and Results: We developed a lifestyle‐based CVD prediction model among 61 025 women in the Nurses’ Health Study and 34 478 men in the Health Professionals Follow‐up Study, who were free of chronic disease in 1986 and followed for ≤24 years. Lifestyle factors were assessed by questionnaires in 1986. In the derivation step, we used the Bayes Information Criterion to create parsimonious 20‐year risk prediction models among a random two thirds of participants in each cohort separately. The scores were validated in the remaining one third of participants in each cohort. Over 24 years, there were 3775 cases of CVD in women and 3506 cases in men. The Healthy Heart Score included age, smoking, body mass index, exercise, alcohol, and a composite diet score. In the validation cohort, the risk score demonstrated good discrimination (Harrell's C‐index, 0.72; 95% confidence interval [CI], 0.71, 0.74 [women]; 0.77; 95% CI, 0.76, 0.79 [men]), fit, and calibration, particularly among individuals without baseline hypertension or hypercholesterolemia. Conclusions: The Healthy Heart Score accurately identifies individuals at elevated risk for CVD and may serve as an important clinical and public health screening tool for the primordial prevention of CVD.