Association of Atherosclerosis Prevalence With Age, Race, and Traditional Risk Factors in Patients With Psoriasis

Abstract

This analysis of aggregate data from electronic medical records assesses the prevalence of atherosclerosis in African American and white patients with psoriasis.

Psoriasis is a chronic skin condition associated with systemic inflammation.¹ This inflammation may promote atherosclerosis, and multiple studies have shown an association between psoriasis and atherosclerotic cardiovascular disease (ASCVD).² Although this association scales with severity and duration of skin disease,^3,4 little is known about race- and age-specific factors. Furthermore, the association between psoriasis and atherosclerosis in the absence of traditional cardiovascular risk factors is controversial.⁵ Our dual-center study aims to elucidate whether certain characteristics of patients with psoriasis are disproportionally associated with ASCVD.

Methods

Data from patient databases at University of Chicago and Northwestern University hospitals were obtained using the i2b2 platform (i2b2 tranSMART Foundation), which allows for deidentified queries of aggregate electronic medical record data using International Classification of Diseases, Ninth Revision (ICD-9) or Tenth Revision (ICD-10) codes and demographics. This platform is designed for aggregate queries, providing counts or percentages that meet specified criteria. It does not provide or give access to individual-level data, so means and medians cannot be computed. The institutional review boards of University of Chicago and Northwestern University determined that study approval and patient written consent were not required because the study was not considered human subject research.

Psoriasis was defined by ICD-9 code 696 or ICD-10 code L40. Atherosclerotic cardiovascular disease was defined by any of ICD-9 codes 410 to 414 and 440, as well as ICD-10 codes I20 to I25 and I170. Myocardial infarction was defined by ICD-9 codes 410 and 412 or ICD-10 codes I21 to I22. Stroke was defined by ICD-9 code 434 or ICD-10 code I63. Traditional cardiovascular risk factors were assessed using ICD-9 codes 272.0 to 272.4 or ICD-10 codes E78.0 to E78.5 for hyperlipidemia; ICD-9 code 250 or ICD-10 codes E08 to E13 for diabetes mellitus; ICD-9 codes 401 to 405 or ICD-10 codes I10 to I15 for hypertension; ICD-9 codes V15.82 and 305.1 or ICD-10 code Z72.0, and Current Procedural Terminology codes 1032F and 1034F for current/past tobacco smoker. Statistical significance between groups was ascertained by χ² testing. The electronic medical record data spans 2008 to 2018; data queries and analyses were performed from May to November 2018.

Results

Medical records of 1 991 149 adults were included, 555 875 African American patients and 1 435 274 white patients, of whom 12 184 had a diagnosis of psoriasis. As supported by the data reported in detail in the Table, ASCVD prevalence was 2.4-fold higher in both races when psoriasis was present, and absolute ASCVD rates, including rates of stroke, were highest in African Americans with psoriasis (n = 2495). Most cardiovascular risk factors were more prevalent in patients of both races with psoriasis and in African American patients with psoriasis compared with white patients with psoriasis. In young African American patients (age range, 18-44 years), but not young white patients, psoriasis was associated with a 3.1-fold increase in ASCVD (Figure), and traditional cardiovascular risk factor burden was higher in African American patients than in white patients in this age group (diabetes, 9.0% vs 2.4%; smoking, 10.7% vs 5.2%; hypertension, 21.9% vs 7.2%; hyperlipidemia, 12.8% vs 7.3%; P < .001 for all factors). In patients of both races with psoriasis, older adults had comparable increased rates of ASCVD, regardless of sex.

Table. Baseline Characteristics and Prevalence of ASCVD, Myocardial Infarction, and Stroke in Patients With and Without Psoriasis.

Characteristic/Factor	AA Patients (n = 555 875), %		P Value	WH Patients (n = 1 435 274), %		P Value	P Value (AA vs WH Patients With Psoriasis)
	With Psoriasis	Without Psoriasis		With Psoriasis	Without Psoriasis
No. of patients	2495	553 380	NA	9689	1 425 585	NA	NA
Age, y
18-44	38.1	40.8	.001	31.4	32.7	.005	<.001
45-64	36.8	35.4	.08	38.8	35.2	<.001	.04
≥65	25.1	23.8	.07	29.9	32.1	<.001	<.001
Sex, % men	31.5	38.5	<.001	48.3	44.2	<.001	<.001
Cardiovascular risk factors
Diabetes	20.6	9.3	<.001	10.2	3.8	<.001	<.001
Smoking	20.1	7.0	<.001	10.0	2.9	<.001	<.001
Hypertension	45.7	21.7	<.001	28.8	11.1	<.001	<.001
Hyperlipidemia	29.4	9.7	<.001	27.8	8.8	<.001	.10
No risk factors	45.3	73.8	<.001	57.0	83.6	<.001	<.001
ASCVD	15.4	6.3	<.001	10.2	4.2	<.001	<.001
Stroke	4.5	1.9	<.001	1.7	0.7	<.001	<.001
Myocardial infarction	2.2	1.0	<.001	2.3	1.0	<.001	.74

Abbreviations: AA, African American; ASCVD, atherosclerotic cardiovascular disease; NA, not applicable; WH, white.

Figure. Prevalence Ratios of Atherosclerotic Cardiovascular Disease in Patients With vs Without Psoriasis.

Age, race, sex, and risk factor subgroups are shown. Bars represent the 95% CIs. AA indicates African American; ASCVD, atherosclerotic cardiovascular disease; WH, white.

In bivariate analysis, patients with psoriasis along with diabetes or hypertension had higher rates of ASCVD compared with patients with diabetes or hypertension who did not have psoriasis (Figure). In the presence of hyperlipidemia or smoking, patients with psoriasis did not have higher ASCVD rates than those without psoriasis. In patients older than 35 years without any traditional risk factors (smoking, hyperlipidemia, hypertension, diabetes), ASCVD prevalence was 1.7 times greater than in those with psoriasis (Figure).

Discussion

The association between psoriasis and ASCVD has been established.¹ Our study offers a nuanced view of this association within the context of age, race, and sex. Particular subgroups of patients with psoriasis are at greater risk of ASCVD than other groups; increased risk of ASCVD associated with psoriasis occurs at an earlier age in African American patients than in white patients. This finding could inform decisions to refer such patients for early cardiovascular risk assessment and reduction.

Although some research suggests that psoriasis raises cardiovascular risk solely because it is associated with higher rates of traditional cardiovascular risk factors, other research has supported psoriasis as an independent cardiovascular risk factor.⁵ Our study aligns with both mechanisms; accounting for traditional cardiovascular risk factors greatly reduces the ASCVD prevalence ratio, but patients with psoriasis and no documented risk factors still have a higher ASCVD prevalence than their counterparts without psoriasis. The psoriatic march hypothesis, which posits that psoriasis drives both systemic inflammation and development of traditional cardiovascular risk factors, matches well with our observations.⁶

This study’s limitations include the retrospective dual-center design, database-only analysis, and the lack of access to data from medical records to verify coding or perform individual-level multivariable analysis. Despite these limitations, our sample was sufficiently large and diverse to reveal insights into a complex association between psoriasis and atherosclerosis that varies by traditional risk factors and demographics.