Time trends in cardiovascular risk factors across levels of education in a general population

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Northern Norway Health Authority Background Favourable changes over time in population cardiovascular disease (CVD) risk has not benefitted all socioeconomic groups equally. Social inequality in health also exists in high-income countries with an egalitarian welfare system. Purpose We aimed to study CVD risk factors across educational groups over time in a Norwegian general population. Methods We used multivariable linear regression and generalised estimating equation models with age-adjusted means and proportions to examine CVD risk (smoking, physical activity level, obesity [body mass index ≥30 kg/m2], blood pressure, total- and low-density lipoprotein [LDL] cholesterol, and total risk of CVD [estimated 10-year risk with NORRISK 2]) across four levels of education: Primary/partly secondary up to 10 years of schooling, upper secondary, tertiary <4 years, tertiary ≥4 years in women and men aged 40-79 years attending a population-based study with two consecutive surveys in 2007-2008 (N=11,941) and 2015-2016 (N=20,322). Results In women, the difference between the lowest and the highest education level in risk factors in 2007-2008 versus 2015-2016 was for smoking 22.0 vs 19.2 percentage points (pp), obesity 8.7 vs 8.8 pp, sedentary physical activity level 10.5 vs 11.5 pp (p=0.012), systolic blood pressure 5.5 vs 3.4 mmHg (p=0.001), total cholesterol 0.2 mmol/L vs no difference (p<0.001), LDL cholesterol 0.3 vs 0.1 mmol/L (p<0.001), and total CVD risk 0.9 vs 1.8 (p<0.001), respectively. In men, the corresponding numbers were for smoking 20.1 vs 16.8 pp, obesity 9.9 vs 12.9 pp, sedentary physical activity level 10 vs 11.5 pp (p=0.027), systolic blood pressure 1.7 vs 1.7 mmHg, total cholesterol 0.2 mmol/L vs no difference (p<0.001), LDL cholesterol 0.1 mmol/L vs no difference (p<0.001), and total CVD risk 1.2 vs 0.9, respectively. No difference across educational levels and time regarding blood pressure or total- or LDL cholesterol in users of antihypertensives and lipid-lowering drugs was found, except for an educational difference in the first but not the second time period in female antihypertensive users. Conclusion We found a clinically relevant educational gap in CVD risk factors, with more favourable levels with higher education. Over time, the educational gap increased in physical activity in both sexes, and in total CVD risk in women. Further, the educational gap declined in blood lipids in both sexes, and in systolic blood pressure in women. While medication use seems to contribute to weaken the educational gradient in blood pressure and lipid levels, a stronger emphasis on lifestyle interventions is needed to reduce the educational inequalities in behavioural CVD risk factors.