Objective Despite an upsurge in the incidence of atherosclerotic cardiovascular diseases (ASCVD) among young adults, the attributable risk of recreational substance use among young patients has been incompletely evaluated. We evaluated the association of all recreational substances with premature and extremely premature ASCVD.
Methods In a cross-sectional analysis using the 2014–2015 nationwide Veterans Affairs Healthcare database and the Veterans wIth premaTure AtheroscLerosis (VITAL) registry, patients were categorised as having premature, extremely premature or non-premature ASCVD. Premature ASCVD was defined as having first ASCVD event at age <55 years for men and <65 years for women. Extremely premature was defined as having first ASCVD event at age <40 years while non-premature ASCVD was defined as having first ASCVD event at age ≥55 years for men and ≥65 years for women. Patients with premature ASCVD (n=135 703) and those with extremely premature ASCVD (n=7716) were compared against patients with non-premature ASCVD (n=1 112 455). Multivariable logistic regression models were used to study the independent association of all recreational substances with premature and extremely premature ASCVD.
Results Compared with patients with non-premature ASCVD, patients with premature ASCVD had a higher use of tobacco (62.9% vs 40.6%), alcohol (31.8% vs 14.8%), cocaine (12.9% vs 2.5%), amphetamine (2.9% vs 0.5%) and cannabis (12.5% vs 2.7%) (p<0.01 for all comparisons). In adjusted models, the use of tobacco (OR 1.97, 95% CI 1.94 to 2.00), alcohol (OR 1.50, 95% CI 1.47 to 1.52), cocaine (OR 2.44, 95% CI 2.38 to 2.50), amphetamine (OR 2.74, 95% CI 2.62 to 2.87), cannabis (OR 2.65, 95% CI 2.59 to 2.71) and other drugs (OR 2.53, 95% CI 2.47 to 2.59) was independently associated with premature ASCVD. Patients with polysubstance use had a graded response with the highest risk (~9-fold) of premature ASCVD among patients with use of ≥4 recreational substances. Similar trends were observed among patients with extremely premature ASCVD. Gender interactions with substance use were significant (p-interaction <0.05), with recreational substance use and premature ASCVD showing stronger associations among women than in men with premature ASCVD.
Conclusions All subgroups of recreational substances were independently associated with a higher likelihood of premature and extremely premature ASCVD. Recreational substance use confers a greater magnitude of risk for premature ASCVD among women. A graded response relationship exists between increasing number of recreational substances used and higher likelihood of early-onset ASCVD.
- peripheral vascular disease
- smoking cessation
- coronary artery disease
- cardiac risk factors and prevention
Statistics from Altmetric.com
- peripheral vascular disease
- smoking cessation
- coronary artery disease
- cardiac risk factors and prevention
Despite substantial improvements in the prevention and treatment of atherosclerotic cardiovascular diseases (ASCVD), recreational substance use remains one of the key modifiable risk factors.1 The role of recreational substances such as alcohol, tobacco and illicit drugs in the development and progression of clinical ASCVD has been well defined.2–5 The pathophysiological mechanisms behind this association include an overall hyperadrenergic state,3 6 increased oxidative stress,4 endothelial dysfunction7 and direct toxic effects of drugs such as methamphetamine.8 Users of such drugs experience higher rates of adverse cardiovascular events compared with non-users with a similar cardiovascular risk profile.9 10 Furthermore, current data suggest a slow yet sustained increase in the use of such substances globally.11 12
ASCVD, which comprises ischaemic heart disease (IHD), ischaemic cerebrovascular disease (ICVD) and peripheral arterial disease (PAD), is a disease no longer restricted to older age. Epidemiological data have shown an upsurge in the incidence of all major realms of ASCVD in young and middle-aged adults.13 Although the global epidemic of metabolic syndrome has been blamed for the rising incidence of premature ASCVD,14 the attributable risk of recreational substances among young adults remains incompletely evaluated. A substantial portion of the current evidence is limited due to its focus on a single domain of ASCVD,15 a single recreational substance or synergism with alcohol or tobacco. Furthermore, prevailing data on this subject largely focus on substance use in men with early-onset ASCVD, while the gender-specific risk among women has not been completely explored.
The primary aim of our analysis was to conduct a comprehensive exploration of all major subgroups of recreational substances and their association across all domains of premature ASCVD. We also sought to evaluate the prevalence and association of recreational substances among patients with extremely premature ASCVD. Finally, we aimed to evaluate gender differences and the gender-specific risk of premature and extremely premature ASCVD among patients with recreational substance use.
Cohort development and definitions
Using the nationwide Veterans Affairs (VA) Healthcare System administrative and clinical datasets, we identified patients with a diagnosis of ASCVD aged ≥18 years. We included 1 248 158 patients receiving primary care services throughout the VA Healthcare System between 1 October 2014 and 30 September 2015 (VA fiscal year 2015). We defined the most recent primary care provider (PCP) visit in fiscal year 2015 as the index PCP visit, which was used to anchor our analyses. Additional details regarding cohort creation have been previously described.16 ASCVD was defined as a history of IHD, ICVD or PAD, which was ascertained using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis and procedural codes along with Current Procedural Terminology codes.17 Based on a manual chart review of 200 patients, the positive predictive value of our algorithm for the identification of ASCVD was 95%.18
Using this ASCVD cohort, we developed the Veterans wIth premaTure AtheroscLerosis (VITAL) nationwide registry by using patients’ age cut-offs.19 20 In accordance with cardiovascular guidelines,21 we defined premature ASCVD as men and women who experienced their first ASCVD event at age <55 years and <65 years, respectively; non-premature ASCVD as men and women with their first ASCVD event at age ≥55 years and ≥65 years, respectively; and extremely premature ASCVD as patients who experienced their first ASCVD event at age <40 years. Patients with missing gender information or date of birth were excluded from our analyses along with patients with limited life expectancy (hospice care in the preceding 12 months or a history of metastatic cancer in the last 5 years).22
Patient and public involvement
Patients were not directly involved in this research.
Study outcomes and measures
We used ICD-9-CM codes (see online supplemental table 1) to identify recreational substances including tobacco, alcohol and illicit drugs. Illicit drugs included cocaine, amphetamine, cannabis and other drugs. These ICD-9-CM codes have been previously validated with a specificity of >90% among the VA population.23 The use of these recreational substances served as our exposure variables which were assessed across patients with premature ASCVD, extremely premature ASCVD and non-premature ASCVD. The subgroup of patients with non-premature ASCVD served as the reference group while patients with premature and extremely premature ASCVD served as the outcome variables. We also conducted subgroup analysis of patients with premature IHD, premature PAD and premature ICVD.
We used the VA clinical data sources to identify patient demographics. Patients’ medical history including hypertension, diabetes mellitus, hypercholesterolaemia, IHD, myocardial infarction, ICVD and PAD was identified using Current Procedural Terminology codes, ICD-9-CM diagnosis and procedure codes. We also computed the Diagnosis Cost Group Relative Risk Score (DCG-RRS), which is a well-validated surrogate marker of a patient’s overall illness and accordingly healthcare cost burden.24 A DCG-RRS score of 1 signifies a patient with an average illness burden while a score above or below 1 signifies an illness burden which is higher or lower than average, respectively.
Various patient level variables were assessed across patients with premature and non-premature ASCVD. A χ2 test was used to assess categorical variables. We evaluated the frequency of use of recreational substances across patients with premature and non-premature ASCVD and created multivariable logistic regression models to assess the association between use of each of the recreational substances and premature ASCVD. We adjusted our initial regression models (partially adjusted models) for the following covariates: patient’s gender, race, obesity, hypertension, diabetes, hypercholesterolaemia, presence of IHD and mean DCG-RRS. We further adjusted our regression models (fully adjusted models) for the presence of other recreational substance use categories. For example, in fully adjusted models for tobacco use, we adjusted for alcohol use and illicit drug use. Similar analytical methodology was used to assess the association between each of the recreational substances and each domain of premature ASCVD (premature IHD, premature PAD or premature ICVD).
Additionally, we evaluated the burden of polysubstance use by evaluating the frequency of only 1, only 2, only 3, and ≥4 recreational substances across patients with premature and non-premature ASCVD. Multivariable logistic regression models adjusting for the aforementioned covariates were constructed to evaluate a graded response relationship between polysubstance use and premature ASCVD. In addition, we evaluated combinations of different illicit drugs and their association with premature ASCVD. Subsequently, we conducted similar analyses among patients with extremely premature ASCVD. The control group for all analyses consisted of patients with non-premature ASCVD.
We also used multiplicative interaction models to assess interactions between gender and use of recreational substances. Furthermore, we used the aforementioned statistical methodology to conduct subgroup analyses stratified by gender in order to assess the role of gender in the association between recreational substance use and premature or extremely premature ASCVD.
SAS version 9.1.3 (SAS Institute, Cary, North Carolina, USA) and Stata version 14 (StataCorp, College Station, Texas, USA) were used to perform our analyses. We obtained an informed consent waiver and our study protocol was approved by the institutional review boards at Baylor College of Medicine and Michael E DeBakey VA Medical Centre.
Baseline characteristics: premature ASCVD
We identified 135 703 patients with premature ASCVD included in the VITAL registry while the remaining 1 112 455 patients were identified as having non-premature ASCVD (table 1). Compared with patients with non-premature ASCVD, patients with premature ASCVD had a higher proportion of women (13.9% vs 0.7%), non-white population (33.1% vs 19.8%) and had a higher prevalence of obesity (55.8% vs 39.9%) and hypercholesterolaemia (47.4% vs 29.4%) (p<0.01 for all comparisons). The mean DCG-RRS was significantly higher among patients with premature ASCVD (2.18 vs 1.54, p<0.01).
Recreational substances and premature ASCVD
The use of all three categories of recreational substances and all illicit drugs was observed to be significantly higher among patients with premature ASCVD than in those with non-premature ASVCD (table 2). In fully adjusted regression models, the use of all recreational substances was independently associated with a higher odds of premature ASCVD (figure 1): tobacco use (OR 1.97, 95% CI 1.94 to 2.00), alcohol use (OR 1.50, 95% CI 1.47 to 1.52) and illicit drug use (OR 2.87, 95% CI 2.81 to 2.93). Among the illicit drug category, all individual drugs were independently associated with increased odds of premature ASCVD: cocaine use (OR 2.44, 95% CI 2.38 to 2.50), amphetamine use (OR 2.74, 95% CI 2.62 to 2.87), cannabis use (OR 2.65, 95% CI 2.59 to 2.71) and other drugs (OR 2.53, 95% CI 2.47 to 2.59). These results were consistent throughout our subgroup analyses of patients with premature IHD, premature PAD and premature ICVD. In fully adjusted regression models, the use of all recreational substances was independently associated with increased odds of (~1.3–3.0) premature IHD, premature PAD and premature ICVD (see online supplemental tables 2-4).
Additionally, patients with premature ASCVD had higher rates of polysubstance use (table 3). Based on our adjusted regression models, patients at each level of polysubstance use had a higher likelihood of premature ASCVD (figure 2). The likelihood of premature ASCVD increased in a graded response manner with increasing number of recreational substances used: use of only 1 substance (OR 2.05, 95% CI 2.02 to 2.08), only 2 substances (OR 3.45, 95% CI 3.38 to 3.52), only 3 substances (OR 6.38, 95% CI 6.18 to 6.58), and ≥4 substances (OR 8.85, 95% CI 8.63 to 9.08).
Even among patients without concomitant alcohol or tobacco use, the use of illicit drugs was associated with ~4 times increased risk of premature ASCVD (OR 4.17, 95% CI 3.88 to 4.48). We also found a graded response among polysubstance users without concomitant alcohol or tobacco use (see online supplemental table 5). Finally, across different combinations of illicit drugs (see online supplemental table 6), the combined use of cannabis, cocaine and amphetamine had the highest likelihood of premature ASCVD (OR 5.57, 95% CI 5.20 to 5.98).
Extremely premature ASCVD
A subset of 7716 patients identified from the VITAL registry as having extremely premature ASCVD were compared against the 1 112 455 patients with non-premature ASCVD. The baseline characteristics of patients with extremely premature ASCVD are shown in online supplemental table 7.
Compared with patients with non-premature ASCVD, patients with extremely premature ASCVD had a higher use of all recreational substances (see online supplemental table 8). The frequency of use for all illicit drugs was also higher among the extremely premature ASCVD population. After complete adjustment for all covariates, the use of each of the recreational substances was independently associated with ~1.5–3 times higher odds of extremely premature ASCVD (figure 1).
The frequency of polysubstance use was also observed to be higher among patients with extremely premature ASCVD (see online supplemental table 9). In fully adjusted regression models we observed a graded relationship between polysubstance use and its association with extremely premature ASCVD (1.66 higher odds with use of only 1 substance and 6.32 higher odds with use of ≥4 substances). Across various combinations of illicit drugs (online supplemental table 10), the combined use of cannabis and amphetamine demonstrated the highest likelihood of extremely premature ASCVD (OR 7.20, 95% CI 6.15 to 8.43).
Multiplicative interactions with female gender and the use of each of the recreational substances were significant among patients with premature ASCVD and those with extremely premature ASCVD (p-interaction <0.05 for all comparisons). Gender-based subgroup analysis showed that although all categories of recreational substances were independently associated with higher odds of premature ASCVD in both men and women, the magnitude of effect size was higher among women than men (figure 3; tables 4 and 5). Similar findings showing a higher magnitude of risk with recreational substances was also observed among women with extremely premature ASCVD compared with men (see online supplemental tables 11 and 12).
We have shown that, compared with patients with non-premature ASCVD, patients with premature and extremely premature ASCVD had a higher prevalence of all recreational substance use. The use of all forms of recreational substances was independently associated with ~1.5–2 times higher likelihood of premature and extremely premature ASCVD. Illicit drug use was associated with ~3 times higher likelihood of both premature and extremely premature ASCVD. Among all illicit drugs, amphetamines and cannabis had the greatest odds of early-onset ASCVD. A graded response relationship exists between increasing number of substances used and higher likelihood of early-onset ASCVD. Finally, we highlight that female users of recreational substances have a greater risk of premature and extremely premature ASCVD compared with their male counterparts.
To our knowledge, our results serve as one of the most comprehensive pool of evidence demonstrating associations of each individual class of recreational substance across all major domains of premature and extremely premature ASCVD. Our results are consistent with earlier observations demonstrating an association between the use of recreational substances and premature ASCVD.15 25 Furthermore, compared with patients with non-premature ASCVD, the penetration of all recreational substances was higher among patients with premature and extremely premature ASCVD. Therefore, the traditional practice of directing the majority of clinical focus towards optimisation of traditional risk factors without equally addressing the use of recreational substances may result in a disservice to the young adult population. As noted in prior literature,26 tobacco was the most prevalent recreational substance in our cohort. However, despite widespread use of tobacco, illicit drug use was associated with the highest likelihood of early-onset ASCVD.
We also demonstrated synergism between different recreational substances. In our partially adjusted regression models we demonstrated a larger effect size of recreational substances on premature ASCVD, which was subsequently attenuated (although remained statistically significant) on adjustment for other classes of recreational substances. Given the high prevalence of tobacco use among illicit drug users, prior investigations have hypothesised a synergistic effect between tobacco use and other illicit drugs such as cocaine and cannabis.27 However, we found that all categories of illicit drugs were associated with early-onset ASCVD, independent of tobacco or alcohol use. The association of these recreational substances with early-onset ASCVD, independent of traditional atherosclerotic risk factors or concomitant use of other drugs, emphasises the importance of facilitating cessation among young adult users.
Our gender-based subgroup analyses showed that, although the use of recreational substances was associated with an increased risk of early-onset ASCVD across both genders, women remain at a substantially higher risk for premature and extremely premature ASCVD than men. This elevated risk among women is evident across all categories of recreational substances. Female users were noted to have a 2–7 times higher odds of premature or extremely premature ASCVD compared with male users who had 1–3 times higher odds. Prior investigations have shown gender-based disparities in screening and prevention of substance abuse disorders particularly in women.28 Societal beliefs, misconceptions and patient/clinician gender discordance are some reasons which have been attributed to the prevailing gender disparity in screening and treatment measures. These reasons may hinder appropriate screening and surfacing of the true prevalence of substance use among women. The prevailing gender-based disparity in screening, in conjunction with female users being at a higher risk for premature and extremely premature ASCVD, raises a significant public health concern. Therefore, the results from our study serve as supporting evidence for clinicians to be more vigilant towards women for screening measures and has implications for aggressive and early treatment efforts to prevent the onset of premature ASCVD.
We demonstrated a graded response relationship between the increasing number of substances used and higher odds of premature and extremely premature ASCVD. Although intuitive, the cumulative risk of polysubstance abuse in the setting of early-onset ASCVD has not been thoroughly investigated in prior studies.29 Our results show a linear increase in the likelihood of early-onset ASCVD with higher number of recreational substances used. Compared with the ~1.5–2 times higher odds with use of only one recreational substance, patients who consumed ≥4 substances had a ~6–8-fold increase in likelihood of early-onset ASCVD. Furthermore, approximately one-third of young and middle-aged adults with ASCVD had polysubstance use (>1 substance use). The high prevalence of polysubstance abuse in this population along with the magnitude of increased risk of early-onset ASCVD implies a significantly elevated population attributable risk. Accordingly, such information regarding polysubstance use can serve to mitigate the residual risk of ASCVD despite controlling traditional risk factors with medication and lifestyle changes.
With increasing use and legalisation of recreational substances such as cannabis, it is essential for clinicians to recognise the penetration and associated cardiovascular risk of these substances among young adults. Although the current American College of Cardiology/American Heart Association guidelines recommend drug screening for all patients with acute myocardial infartion, similar recommendations among patients with ICVD or PAD and policy-driven quality improvement initiatives may improve screening rates. Our findings also support the need for aggressive interventions in implementation and accessibility of drug cessation programmes as another key element of the primary prevention of ASCVD.
Our study has several limitations. Due to the inherent constraints of clinical datasets and the observational design of our study, we were unable to ascertain and adjust for additional confounders such as dose and duration of recreational substance use. The socioeconomic background of patients was not included in our study, which has been independently associated with the risk of ASCVD.30 Our analyses did not distinguish between smoking versus smokeless tobacco, which may have a variable effect on the risk of premature ASCVD. Patients receiving medically prescribed amphetamines (eg, for attention deficit hyperactivity disorder) were not excluded from our analyses and thereby could have influenced the true proportion of patients using amphetamines as a recreational substance. Although our regression models were adjusted for various comorbid conditions such as hypertension and diabetes, the optimal control of these comorbidities or adherence with medical therapy was not assessed, which may influence the risk of incident ASCVD. We did not identify patients with myocardial infarction without obstructive coronary artery disease (MINOCA), which may have influenced our results given the well-known association of recreational substances with various elements of MINOCA such as coronary vasospasm, coronary artery dissection or thromboembolism. Finally, given our white male predominant cohort, the generalisability of our findings to other ethnic minorities and non-VA population may be limited.
We show that patients with premature and extremely premature ASCVD have higher use of all recreational substances. The use of all subgroups of recreational substances was independently associated with a higher likelihood of early-onset ASCVD, despite adjustment for traditional atherosclerotic risk factors and concomitant use of other recreational substances. Women using recreational substances have greater odds of premature and extremely premature ASCVD than men. There was a graded response relationship between the number of substances used and the likelihood of early-onset ASCVD. Apart from optimisation of metabolic risk factors, young adults and especially women with premature ASCVD should be screened thoroughly for the use of recreational substances, appropriately risk-stratified, and directed towards multidisciplinary pathways (pharmacological and non-pharmacological) to achieve and maintain cessation.
What is already known about this subject?
The role of recreational substances such as alcohol, tobacco and illicit drugs in the development and progression of clinical ASCVD has been well defined. However, most of the data pertains to the effects of recreational substances on the incidence of acute myocardial infarction and patients with ischaemic heart disease.
What does this study add?
This study provides a comprehensive analysis of all subgroups of recreational substances and their association with all three domains of premature ASCVD (ischaemic heart disease, ischaemic cerebrovascular disease and peripheral arterial disease). Furthermore, this study provides these associations among patients with premature and those with extremely premature ASCVD. More importantly, this study also highlights the higher risk for premature and extremely premature ASCVD among women with recreational substance use compared with men. Lastly, this study demonstrates a graded risk relationship between polysubstance use and premature ASCVD.
How might this impact on clinical practice?
The associations between subgroups of recreational substances and premature as well as extremely premature ASCVD, as demonstrated in this study, would allow clinicians to be more vigilant regarding screening this population and risk-stratifying young patients with ASCVD. Additionally, recognition of these strong associations would allow clinicians to be aggressive about providing young patients with drug cessation resources.
Contributors DM: Conception, design and interpretation of data; drafting and revising of manuscript; final approval of the manuscript submitted. DR: Analysis and interpretation of data; final approval of the manuscript submitted. MAlR, CK, NK, ZS, HJ, CB, LAP: Interpretation of data; revising of manuscript; final approval of the manuscript submitted. SSV: Conception, design and interpretation of data; drafting and revising of manuscript; final approval of the manuscript submitted.
Funding This work was supported by a Department of Veterans Affairs Health Services Research & Development Service Investigator Initiated Grant (IIR 16–072), an American Heart Association Beginning Grant-in-Aid (14BGIA20460366), the American Diabetes Association Clinical Science and Epidemiology award (1-14- CE-44), and the Houston VA Health Services Research & Development Center for Innovations grant (CIN13-413). Support for VA/CMS data provided by the Department of Veterans Affairs, VA Health Services Research and Development Service, VA Information Resource Center (Project Numbers SDR 02–237 and 98–004).The opinions expressed reflect those of the authors and not necessarily those of the Department of Veterans Affairs or the US government.
Competing interests DM, DR, CK, MAlR, ZS, HJ, LAP: None. NK: Serves on the advisory board of Medicines company and Regeneron. CB: Grant/Research Support - all significant (all paid to institution, not individual): Akcea, Amgen, Esperion, Novartis, Regeneron, Sanofi-Synthelabo, NIH, AHA, ADA. Consultant: Abbott Diagnostics, Akcea, Amarin, Amgen, Arrowhead, Astra Zeneca, Corvidia, Denka Seiken, Esperion, Intercept, Matinas BioPharma, Merck, Novartis, Regeneron, Sanofi-Synthelabo. SSV: Honorarium, American College of Cardiology (Associate Editor for Innovations, acc.org); Steering Committee, Patient and Provider Assessment of Lipid Management (PALM registry) at the Duke Clinical Research Institute (no financial remuneration).
Patient consent for publication Not required.
Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information. The Veterans wIth premaTure AtheroscLerosis (VITAL) registry.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.