Depression and anxiety screening improves cardiovascular risk prediction in population-based sample of 4,750 russian individuals

Introduction

Cardiovascular diseases (CVD) are the most serious public health problem, remaining the leading cause of death in the adult population. A great number of CVD risk scales are successfully used in clinics, yet they are tuned to work the best for patients of older age and often are population specific.

Methods

We used the data from a longitudinal epidemiologic study of 4,750 individuals aged 25–64 years from European part of Russia recruited in 2012–2013 to predict the 10-year risk of cardiovascular events in Russian population and evaluate performance for common clinical risk scales. The cohort was divided into two groups: “cases” – individuals with a cardiovascular event registered during the follow up period (2013–2019) as a study group (N=106) and “controls” – individuals with no cardiovascular events registered during follow up as a control group (N=4,644). All individuals with previous history of CVD (ischemic heart disease, stroke or heart attack) prior to 2012 were assigned high risk in all scales.

The case and control groups were split by age: below and above 40 to reflect the properties of cardiovascular risk scales to work best for age over 40.

We assessed CVD risk scores using Framingham 2008 [1], ASCVD 2013, both were calculated with mean parameters values calculated from our cohort and from the original study [2], SCORE 2003, 2017 and 2019 [3] and MOSP - the recalibrated for Russian population scale of the SCORE 2017 [4]

Results

For patients over 40 years old, cardiovascular risk scales showed similar performance, with ASCVD normalized to the original study's mean parameter values returning the best prediction scores. Expectedly, for the younger group of patients, cardiac risk scales do not have notable predictive power.

Further, we thought to identify additional factors discriminating young individuals at higher risk of CVD. We downsampled the control cohort to include only samples with age, weight, height, LDL, total cholesterol and systolic blood pressure matching those in a case cohort (<40 years of age).

186 phenotypic features were tested and 4 out of top 10 (p<0.05) associated features included questions related to depression, anxiety or current stress level from HADS scale, though not significant after multiple hypothesis correction, this suggests that younger individuals with high scores in depression/anxiety screenings might be at higher risk of CVD. Depression/anxiety features were added to cardiovascular risk scales as binary predictor components to improve prediction quality (Table 1).

Conclusions

We identified depression and anxiety screening questionnaires as a valuable predictor for cardiovascular events in the younger population improving the quality of predictions of traditional clinical scales.

Funding Acknowledgement

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Ministry of Science and Higher Education of the Russian Federation Table 1. ROC analysis

Clinical and genetic markers of prehypertension in North-Western Russian population

Introduction

Prehypertension is a medical classification for patients with above normal blood pressure (BP) but insufficient for confirming hypertension diagnosis. Condition is often asymptomatic as it gradually develops over the years, yet recent meta-analyses suggest that prehypertension is a significant risk factor for stroke and other cardiovascular diseases.

Methods

We analyzed phenotypic data from 879 (age 25–64) individuals without hypertension from a population based-sampling cohort of St. Petersburg region in the North-West of Russia to identify clinical risk factors associated with prehypertension condition. All patients were divided in two groups – optimal BP (systolic BP <120 mmHg and diastolic BP <80; N=426) and prehypertension (systolic BP in range 120–140 mmHg and diastolic BP in range 80–90 and not on antihypertensive therapy; N=453).

Results

Phenotypic analysis with linear regression was corrected for age, sex, smoking status, BMI and levels of LDL and HDL cholesterol (Figure 1). Interestingly, blood level of insulin was significantly associated with prehypertension status along with insulin resistance index, however, presence of diabetes diagnosis in medical history was not significant. Phenotypes associated with prehypertension suggest that prehypertension is often developing along with hyperinsulinemia.

Finally, we estimated polygenic risk scores (PRS) for hypertension using UK biobank GWAS summary statistics and confirmed that prehypertension is more frequent at earlier age in patients with higher genetic susceptibility (Figure 1, Figure 2).

Conclusions

Hyperinsulinemia and genetic susceptibility to hypertension are strong risk factors for prehypertension.

Funding Acknowledgement

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): This work was financially supported by the Ministry of Science and Higher Education of the Russian Federation (Agreement No. 075-15-2020-901) to Al.K. Figure 1. Clinical markers of prehypertensionFigure 2. Genetic markers of prehypertension