Characteristics and Demographics of Patients Younger than 50 with Atherosclerotic Cardiovascular Disease

Abstract

Background: Premature atherosclerosis (PreAS) is generally defined as a disease affecting those under the age of 50 and has an outsized impact on quality-adjusted life years. We sought to better understand what individuals are at the highest risk for PreAS by examining differences in demographics and comorbidities compared to traditional atherosclerosis (TradAS). Study Design: An Institutional Review Board (IRB) approved retrospective study was conducted using retrospective data from a large regional health system. Patients who received a diagnosis of cerebrovascular disease (CeVD), coronary artery disease (CAD) or peripheral arterial disease (PAD) between 2012 and 2023 were included. Results: The review identified 136,328 patients in which 17,008 or 13% presented with PreAS (diagnosed from age 18 up to, and including, age 50). Rates of comorbidities were as follows (PreAs/TradAS): hypertension 63%/86%, diabetes 29%/35%. hyperlipidemia 45%/67%, chronic kidney disease 15%/26%, tobacco use 52%/60% and substance use 25%/9%. Differences in race, ethnicity and gender were as follows (PreAS/TradAS): White 59%/80%, Black 22%/10% and Latinx 17%/6%; male 51%/55%, and female 49%/45%. Conclusions: Patients with PreAS had lower rates of diseases that typically progress with aging, including hypertension, hyperlipidemia, chronic kidney disease, and diabetes. Tobacco use was less prevalent in the PreAS group and there was a significantly higher rate of illicit substance use in the PreAS population. Race and ethnicity were notably different with Black and Hispanic patients representing a significantly larger proportion of those with PreAS relative to TradAS. Our findings suggest risk factors beyond those classically described may play key roles in causing patients to develop PreAS.

1. Introduction

Atherosclerosis (AS) symptoms primarily impact patients over the age of 50 (TradAS), though the underlying pathology progresses at varying rates across an individual’s lifespan. Clinical sequela of atherosclerotic disease can include myocardial infarction (MI), stroke and limb ischemia [1,2]. Atherosclerotic diagnoses include coronary artery disease (CAD), peripheral arterial disease (PAD) and cerebral vascular disease (CeVD), and represent a significant burden to healthcare systems, particularly in economically disadvantaged regions. Recent studies have shown a shift in the epidemiology of atherosclerotic disease with an increased prevalence in developing countries and in women [3].

Premature atherosclerosis (PreAS) is a category of patients who develop symptomatic disease at or before the age of 50. Though many risk factors for PreAS overlap with TradAS, there is evidence to support nontraditional risk factors predisposing to PreAS. Traditionally cited risk factors include age, family history and sex, in addition to smoking, diabetes, hypertension, dyslipidemia and chronic kidney disease. These are thought to also contribute to PreAS. However, there is also data to support less common risk factors as drivers of PreAS, such human immunodeficiency virus (HIV) infection, highly active antiretroviral therapy (HAART), chemotherapy, radiation, recreational substance use, inflammatory and autoimmune conditions, and obesity and metabolic syndromes [2,4,5]. Though PreAS only represents approximately 10% of patients with AS, it can have a disproportionate impact on quality of life for both afflicted individuals and their families [6]. While mortality and morbidity rates from atherosclerotic diseases have improved for older patient groups, rates for younger individuals with PreAS have remained relatively unchanged. Some data even suggests there has been an increase in CAD-associated mortality [7,8].

The purpose of this study was to better understand what individuals are at the highest risk for PreAS by examining differences in demographics and comorbidities compared to TradAS. To do this, we conducted a retrospective review of patients with atherosclerotic disease in a single health system that cares for a diverse urban and rural population.

2. Methods

A chart review was performed on electronic records for a regional health system in compliance with Health Insurance Portability and Accountability Act (HIPAA). After IRB review, to ensure the research was consistent with ethical principles and regulatory requirements, an IRB exemption was granted (ID# 2000034707). Data was obtained for diagnosis made from February 2012 to September 2023 with the assistance of the Joint Data Analysis Team. Inclusion criteria required patients to be over 18 and diagnosed with an International Classification of Diseases (ICD) diagnosis attributed to CAD, PAD and CeVD, including subcategory codes (Supplemental Table S1). The following variables were collected: demographic information (age, sex, race and ethnicity) and clinical characteristics (comorbidities, substance use, mortality and date of diagnosis of each category of disease). Comorbidities were based on a documented diagnosis with no standardized threshold values applied. ThePreAS group were patients who had a documented initial diagnosis of PAD, CeVD or CAD at or before 50 years of age. Demographics and comorbidities were then compared between patients with PreAS and TradAS (patients older than 50 years). Though there are variations in the literature, age 50 years is a commonly used threshold in defining PreAS [2]. Patient confidentiality was strictly maintained throughout the study, and all data was anonymized as soon as feasible during analysis.

Analysis of patient demographics, comorbidities, and mortality was performed with statistical software (Excel version 16, Microsoft, Redmond, WA and Stata version 18, StrataCorp, College Station, TX, USA). The manuscript was composed using a word processor (Word version 16, Microsoft, Redmond WA, USA). Continuous variables are presented as mean values. Categorical variables are summarized as frequencies and percentages. Comparative subgroup analysis was performed using two proportion z-tests. A p-value of <0.05 was considered statistically significant. To better convey effect size, Cohen’s h-value was calculated comparing PreAS to TradAS, with negative values indicating a larger TradAS percentage; h-values greater than 0.5 were shown in bold font to aid in identifying a medium or greater effect. Cohort age was characterized utilizing mean, median, standard deviation and interquartile range.

3. Results

A total of 136,328 patients diagnosed with PAD, CAD or CeVD were identified, in which 17,008 (12.5%) were diagnosed at or before age 50 (

Table 1. Demographics and comorbidities.

	Total	PreAS (≤50 y/o)	TradAS (>50 y/o)	Cohen’s h-Value
Demographics
Total number of patients	136,328	17,008	119,320
Percent of total patients *	100%	12.5%	87.5%	−1.70
Male *	54.1%	50.8%	54.6%	−0.08
Female *	45.9%	49.2%	45.4%	0.08
Mortality Rate *	21.0%	8.2%	22.8%	−0.41
Age at Diag Mean/Median	67/68	40/42	70/70
Age Std Dev/Interq Range	13.7/58–76	8.0/35–47	11.2/62–79
Comorbidities
CAD *	61.0%	42.4%	63.6%	−0.43
PAD *	38.4%	27.6%	39.9%	−0.26
CeVD *	45.2%	46.6%	45.0%	0.03
Atrial fibrillation *	26.3%	9.1%	28.7%	−0.52
Chronic Heart Failure *	34.8%	22.5%	36.6%	−0.31
Chronic Kidney Disease *	24.9%	15.4%	26.3%	−0.27
Dementia *	12.8%	1.4%	14.4%	−0.54
Diabetes *	33.9%	28.5%	34.7%	−0.13
HIV *	1.0%	1.7%	0.9%	0.07
Hyperlipidemia *	65.1%	45.4%	67.9%	−0.46
Hypertension *	83.2%	63.4%	86.0%	−0.53
Liver Disease *	16.3%	19.0%	15.9%	0.08
MI *	27.4%	24.9%	27.8%	−0.07
Malignancy *	30.1%	11.9%	32.7%	−0.51
Pulmonary Disease *	36.8%	32.2%	37.4%	−0.11
Rheumatological Disease +	6.9%	6.5%	7.0%	−0.02

* p < 0.001 when comparing PreAS to TradAS, ⁺ p < 0.05 when comparing PreAS to TradAS. CAD = coronary artery disease, PAD = peripheral artery disease, CeVD = cerebrovascular disease, HIV = human immunodeficiency virus, MI = myocardial infarction. h-values greater than 0.5 were shown in bold font to aid in identifying a medium or greater effect.

). Although males represented a majority in both groups, women made a significantly larger proportion of the PreAS group (49.2% vs. 45.4%). The population distribution by gender and age at diagnosis is shown in Figure 1. There was no difference in population distribution in either group. We did not observe any significant trends in the rate of premature diagnosis across the years of the study (2011–2023). The mean percent of total diagnosis across all years for premature CAD was 10% (SD = 1%), PAD was 6% (SD = 2%) and CeVD was 13% (SD = 2%).

A significant difference in age of diagnosis was noted with mean age for males of 66 (SD = 14) and 68 (SD = 15) for females. There was a significant difference between race, ethnicity and comorbidities with the mean age for the PreAS group of 40 (SD = 8), while the mean age for the TradAS group was 70 (SD = 11). The combined mean age for both groups was 67 (SD = 15). CeVD represented the most common presentation for PreAS (46.6%), while CAD was the most common in TradAS (63.6%). Classic comorbidities such as hypertension, diabetes, hyperlipidemia and chronic kidney disease were significantly less prevalent in PreAS. In addition, less commonly discussed comorbidities such as malignancy, atrial fibrillation, chronic heart failure, dementia and pulmonary disease were also more prevalent (Table 1).

When considering race and ethnicity, a significant difference between patients with PreAS and TradAS was also found (

Table 2. Comparison of race and ethnicity between PreAS and TradAS.

Race and Ethnicity	PreAS	TradAS	Cohen’s h-Value
White *#	59.0%	79.9%	1.26
Black or African American *#	22.4%	9.5%	−0.11
Hispanic or Latina/o/x *#	17.2%	6.2%	−0.17
Asian *#	2.8%	1.5%	0.02
Race Other *#	15.3%	8.9%	0.09

* p < 0.001 when comparing PreAS to TradAS and compared to White. ^# p < 0.001 when comparing PreAS to TradAS compared to all other races. h-values greater than 0.5 were shown in bold font to aid in identifying a medium or greater effect.

). Though White patients represented the majority in both populations, Black, Latinx and Asian patients accounted for a significantly larger proportion of patients in the PreAS population. Figure 2 shows the breakdown of CeVD, PAD and CAD in PreAS and TradAS by race and ethnicity. Significant differences were seen across all race and ethnicities when comparing PreAS and TradAS diagnosis. All race and ethnicity groups, except for White, represented a greater proportion of the PreAS than the TradAS group. Intergroup significance was also found when comparing race/ethnicity to a combined group of all other races and ethnicities within a category.

Table 3. Tobacco and substance use by race and ethnicity.

	Tobacco Use			Substance Use
	PreAs	TradAS	Cohen’s h-Value	PreAS	TradAS	Cohen’s h-Value
Total	50.2%	58.2%	−0.16 *	14.8%	4.3%	0.37 *
Male	55.2%	63.9%	−0.18 *	17.8%	5.4%	0.40 *
Female	45.1%	51.3%	−0.12 *	11.6%	3.1%	0.34 *
White	50.7%	59.6%	−0.18 *	13.2%	3.9%	0.35 *
Black or African American	53.7%	59.2%	−0.11 *	22.7%	9.6%	0.36 *
Hispanic or Latina/o/x	49.8%	50.7%	−0.02 *	14.1%	4.7%	0.33 *
Asian	29.9%	31.3%	−0.03 *	2.4%	0.6%	0.16 *
Race other	47.2%	50.3%	−0.06 *	11.7%	3.7%	0.31 *

* p < 0.001 when comparing PreAS to TradAS for both tobacco and substance use.

shows the rate of tobacco and substance use in PreAs and TradAS by race and ethnicity. Tobacco use was significantly lower in the patients with PreAS in all categories, while substance use was significantly higher in the PreAS group.

Mortality rates varied across different races and ethnicities (Figure 3). The mortality rate for Black patients was significantly higher relative to non-Black patients in both the PreAS and TradAS groups, and this was seen across all diagnoses. Latinx and Asian populations had significantly lower mortality rates when compared to the other ethnic groups. Asians had the lowest overall mortality across all atherosclerotic cardiovascular disease (ASCVD) diagnosis, except for patients with CeVD in the PreAS population.

4. Discussion

Risk factors and initial clinical presentations for ASCVD patients with PreAS are unique from those with TradAS [2,9]. Despite the growing prevalence of premature disease, there remains limited research and understanding of the demographic and socioeconomic determinants of PreAs and how best to mitigate these risk factors. We noted a prevalence of PreAS in our study, 12.5% of the patients with ASCVD, which was consistent with prior studies that reported between 10 and 20% of patients with premature disease [2,6]. The bell-shaped distribution of the age of first diagnosis of ASCVD in our cohort and the lag for women, shown in Figure 1, matches that reported previously. In a previous study, the prevalence of premature PAD in men was found to be 1.4%, which was lower compared to 1.9% in women [10]. We previously observed the predominance of CAD, with over 74% of patients having this as their only diagnosis, while 24% had disease diagnosed in two vascular beds and 7% had disease diagnosed in all three vascular beds (unpublished observations).

Our study shows that there was significant difference when comparing patients with PreAS and TradAS. CeVD was a more frequent initial diagnosis for the PreAS group while the prevalence of PAD was significantly lower in the PreAS population. This was an unexpected and novel finding. Likewise, the finding that the PreAS group was less likely to have lower extremity complaints as their initial presentation is an important observation that is suggested by other studies [9]. As expected, diseases that typically progress with aging, such as hypertension, chronic kidney disease, heart failure, malignancy, atrial fibrillation and diabetes, were significantly less prevalent in the PreAS population. Given the limited time frame of the study, it is difficult to know if patients with PreAS would experience similar rates of these comorbidities after the age of 50. In contrast, we noted that there was a significantly higher rate of liver disease and HIV/AIDS in the PreAS population. The association of liver disease with ASCVD is not clear [11]. However, nonalcoholic fatty liver disease and atherosclerosis share similar risk factors like dyslipidemia, inflammation and insulin resistance [12]. Inflammation is thought to promote the growth of atherosclerotic plaques and the progression of liver disease [13]. In addition, an association of liver fibrosis with carotid atherosclerosis, coronary artery calcification, aortic valve sclerosis and diastolic dysfunction has been reported [13]. HIV and some HAART medications have been found to promote atherosclerosis through dyslipidemia, and injury to macrophages, endothelial and smooth muscle cells [14]. The chronic inflammation of HIV is also thought to contribute [15].

The role of race and ethnicity in the prevalence of AS has been widely studied but not in the context of age [16]. One group studied American population-based cohorts and reported that 1.6% of individuals in the 40–49 year age group had PAD [10]. Premature PAD prevalence among men was lower compared with women, and racial disparities in premature PAD prevalence were apparent. Women, particularly African American women, and American Indian men had relatively higher disease prevalence [10]. Our study also found a relative increased percentage of women with PreAS relative to women with TradAS, suggesting a gender difference beyond premature PAD that warrants further investigation.

Our results show that Black and Latinx patients represented a significantly larger proportion of patients with PreAS than TradAS, and this appears to correlate with higher rates of tobacco and substance use. Previous studies have reported that Black Americans have higher rates of ASCVD than White and Hispanic Americans. This includes a higher risk of stroke, heart disease and sudden cardiac death [17]. Apart from the traditional risk factors, Black Americans are particularly affected by their socioeconomic status and stress. They are more likely to develop ASCVD at earlier ages than other groups, but are diagnosed later and with more advanced disease, less likely to receive optimal medical therapy, more likely to undergo amputations, and are more likely to die from major cardiovascular events [18].

Prior studies have suggested that Hispanic patients have lower ASCVD mortality compared to non-Hispanic Whites despite a disproportionately higher burden of ASCVD. However, recent evidence that disaggregated Hispanic subgroups revealed that ASCVD mortality increased in PreAS Hispanic adults compared to PreAS non-Hispanic white adults [19]. Individuals of South Asian ancestry are also at higher risk of early ASCVD, with higher proportional mortality rates from ASCVD compared with other Asian ethnic groups and non-Hispanic Whites [20]. In our study, although White patients represented the majority of patients in both groups, they were less likely to be diagnosed with PreAS. We also noted significant differences in mortality rates with Black and White patients having significantly higher mortality rates relative to Asian and Latinx patients. Such trends have been seen in more general demographic studies and further investigation is needed to determine if this finding is related to AS [21].

The socioeconomic conditions of the patients are clearly important. Adherence to treatment regimens are influenced by comorbid mental illness, lack of health literacy, financial strain, inadequate housing conditions, food insecurity and poor social support, all of which greatly impair ASCVD outcomes. Likewise, comparison of high-income and low/middle-income countries suggests that prevalence rates were higher for women as compared with men in the low-income countries [22]. The number of PreAS patients with PAD across the world increased by nearly 30%, which was primarily driven by an increasing number of patients in low-income countries.

We also noted a significant difference in tobacco and substance use between the PreAs and TradAS populations. This is consistent with previous large studies showing that active cigarette smoking is a strong risk factor for premature PAD [9]. Nonetheless, it is important to note that the comorbidities of smoking, diabetes, hypertension and dyslipidemia are commonly observed among patients with premature PAD [9]. These risk factors may increase PAD risk by acting both independently and synergistically when present together. A limitation of this dataset is that it did not differentiate type I and type II diabetes. A potential reflection of ongoing tobacco cessation efforts, there was a lower prevalence of tobacco use in the PreAS population across all race and ethnicities in our study [23]. However, there was a nearly three times higher rate of illicit substance use in the PreAS group relative to those with TradAs. Significant difference in the use of illicit substances across race and ethnicity has been noted in prior studies [24]. These illicit substances are known to increase ASCVD. In particular, both cocaine and methamphetamine can have adverse and potentially fatal effects on arteries and blood vessels, including elevated blood pressure, acute vasospasm and atherosclerotic cardiovascular disease [25,26]. Though we were able to examine illicit medication, we were not able to confirm optical medical management and compliance, another area which warrants further investigation.

Limitations of this study lie in its retrospective nature and dependence on the completeness and accuracy of medical record documentation, in particular ICD 10 coding. Thus, we rely on accurate diagnosis and documentation; however, current hard stops in the system may lead to unintentional diagnosis when attempting to obtain diagnostic tests and prescribe medications. As our data was limited to a single health system, it would not capture disease diagnosis that may have been made previously outside the health system, and similarly may exclude mortalities that occurred outside the system. Given the broad region from which patients present to this health system, we were not able to compare how differences in regional demographics may have affected our findings. In relying on diagnostic codes as representative of underlying atherosclerotic disease, we may have inadvertently included patients whose MI, strokes or PAD were due to non-atherosclerotic etiology. Furthermore, national demographics have shifted with greater racial and ethnic representation in the younger population [27] and this may confound the interpretation of our findings. Additional nontraditional risk factors that may have contributed to PreAS, such as chemotherapy, radiation, inflammatory, autoimmune conditions, genetic etiologies, obesity and metabolic syndrome, were not included in this study [2]. As these were not assessed in this study, they pose the risk of being unrecognized cofounders to our dataset. As the cause of death was not identified, our analysis of mortality was significantly limited in this dataset, making trends reflective of correlation rather than causative in relation to atherosclerosis.

Our study demonstrated significant differences in tobacco and substance use between the PreAS and TradAS groups, with notable variation among race and ethnicity. More probative studies to evaluate these findings relative to the local and regional population demographics are still needed, along with a better understanding of the impact of socioeconomic and genetic components. Our large sample size aided in identifying statistical significance, but some differences may not reflect a clinically relevant difference. Though further research is needed, these findings suggest that risk factors beyond those classically described may play key roles in causing patients to develop PreAS.

In summary, our study identifies key differences in comorbidities, race and ethnicity in patients with PreAS relative to TradAS. These findings help to better identify patients at highest risk for premature disease and ensure that prevention and treatment strategies are tailored to this unique subgroup of AS patients. There is new emerging evidence for variation in “vascular aging” that is not explored in this study [28]. Screening for the at-risk group of young patients may provide early interventions to mitigate their future risk for ASCVD.