Relation of Cannabis Use to Elevated Atherosclerotic Cardiovascular Disease Risk Score

Abstract

We hypothesized that cannabis use is associated with cardiovascular disease (CVD) risk factors. This could explain the reported link between cannabis and cardiovascular events including stroke and myocardial infarction. This analysis included 7,159 participants (age 37.8±12.4 years, 48.6% male, 61.5% Caucasian) from the National Health and Nutrition Examination Survey years 2011–2018. Cannabis use was defined by self-report. Participants with a history of stroke or myocardial infarction were excluded. Composite CVD risk was assessed using the ACC/AHA 10-year Atherosclerotic Cardiovascular Risk (ASCVD) Score. Participants were classified based on their ASCVD risk levels as: low (<5.0%), borderline (5.0–7.4%), intermediate (7.5–19.9%), and high (≥20.0%). Multinomial logistic regression was used to examine the association between cannabis use and ASCVD risk category using low-risk ASCVD category as the reference level. About 63.9% (n=4,573) of participants had ever used cannabis. Ever cannabis use was associated with 60% increased odds of high-risk ASCVD score [Odds Ratio (OR) and 95% Confidence Interval (95% CI): 1.60 (1.04–2.45), p-value=0.03]. We also observed a dose-response relationship between increased use of cannabis and a higher risk of ASCVD. Those reporting ≥2 uses per month had 79% increased odds of high-risk ASCVD score [OR (95%CI): 1.79 (1.10–2.92), p-value=0.02] and those reporting ≥1 use per day had 87% increased odds of high-risk ASCVD score [OR (95%CI): 1.87 (1.16–3.01), p-value<0.001]. In conclusion, cannabis use is associated with elevated CVD risk. Individuals using cannabis should be screened for CVD risk, and appropriate risk reduction strategies should be implemented.

Despite improvements in atherosclerotic risk factor modification and treatment of clinical atherosclerotic cardiovascular disease (ASCVD), recreational drug use remains one of the key modifiable risk factors^1,2. With a growing prevalence of cannabis use in the young population³ in the setting of widespread legalization, the importance of risk stratification is paramount. The connection between cannabis use and ASCVD has several proposed pathophysiological mechanisms including thrombosis⁴, vasospastic and tachycardia-mediated oxygen supply/demand mismatch^5,6, hyperadrenergic state^7,8, and oxidative stress/endothelial injury⁹, all of which are frequently implicated with atherosclerotic cardiovascular events. We hypothesized that cannabis use is associated with CVD risk factors. We tested this hypothesis using a composite of CVD risk factors assessed using the American College of Cardiology/American Heart Association 10-year ASCVD risk score in the National Health and Nutrition Examination Survey (NHANES).

Methods:

The NHANES is a survey of the U.S. population. Data in NHANES were collected from 2011–2018 through a series of interviews and evaluations at an independent examination center¹⁰.

Participants <18 years of age or with a history of myocardial infarction or stroke were excluded from the analysis. Those with any missing data regarding cannabis use, ASCVD components, and any other variables analyzed were excluded. Age, gender, race, tobacco smoking status, and cannabis use were defined by self-report. Ever cannabis users answered affirmative to the question, “Have you ever, even once, used marijuana or hashish?” Users who reported using cannabis at least once per month for a period of at least 12 months answered affirmative to the question, “Have you ever smoked marijuana or hashish at least once a month for more than a year?” Current users listed an integer in response to the question, “During the last 30 days, on how many days did you use marijuana or hashish?” Light users were defined as those who reported using ≤4 days per month; heavy users were defined as those who reported using ≥5 days per month. These metrics have been used in prior studies of NHANES^11,12.

Composite cardiovascular risk was assessed using the American College of Cardiology/American Heart Association 10-year ASCVD score¹³. The ASCVD score is a validated risk assessment tool that estimates cardiovascular risk using a pooled cohort equation that uses age, race, gender, cholesterol, blood pressure, and comorbidities to provide a 10-year estimate of risk for major cardiovascular events¹³. ASCVD scores were reported as percentages. When modeled categorically, the following risk intervals were used: low (<5.0%), borderline (5.0–7.4%), intermediate (7.5–19.9%), and high (≥20.0%).

Diabetes was defined as use of an antihyperglycemic medication, fasting serum glucose ≥126 mg/dL, or hemoglobin A1c ≥6.5%. Hypertension was defined as use of an antihypertensive medication or systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥80 mmHg according to American Heart Association/American College of Cardiology guidelines¹⁴. Hyperlipidemia was defined as total cholesterol ≥200 mg/dL, serum triglycerides ≥150 mg/dL, low-density lipoprotein ≥190 mg/dL, or use of lipid-lowering medications. Obesity was defined as body mass index ≥30 kg/m².

Population characteristics were compared based on ever cannabis use. Continuous variables were reported as mean ± standard deviation. Categorical variables were reported as frequency and percentage. A chi-square test was used to compare categorical variables and a student’s t-test was used to compare continuous variables.

Multinomial logistic regression was used to estimate the association between cannabis use and ASCVD score when ASCVD score was modeled categorically; the low risk ASCVD score category was used as the reference level. Since ASCVD score is calculated using age, gender, ethnicity, diabetes status, hypertension status, tobacco smoking status, and hyperlipidemia, these variables were not adjusted for in the regression models. Associations were also measured in subgroups stratified by race, gender, hypertension, obesity, and hyperlipidemia. In the subgroup analyses, a linear regression model was used to test for interaction with the addition of the interaction term between ever cannabis use and subgroup stratification.

All statistical analyses were conducted using RStudio version 1.3.1093 (Boston, MA) and p-values were considered significant if <0.05.

Results:

In this analysis, there were 7,159 participants (age 37.8±12.4 years, 48.6% male, 61.5% Caucasian) after exclusions. About 63.9% (n=4,573) of participants had ever used cannabis. Population characteristics stratified by cannabis use status are shown in Table 1. Participants who had ever used cannabis were more likely to be male, white, and be current tobacco users. Never cannabis users were more likely to have obesity and diabetes. Ever cannabis users had higher mean ASCVD risk scores.

Table 1:

Population Characteristics

Characteristics	Cannabis User Status
	n=2586 (36.1%)	n=4573 (63.9%)
	Never	Ever
Age (years)	37.8 ± 12.5	37.8 ± 12.3
Men	1070 (41.4%)	2411 (52.7%)*
White	1493 (57.7%)	2907 (63.6%)*
Non-white	1093 (42.3%)	1666 (36.4%)*
Hypertension	941 (36.3%)	1693 (37.0%)
Obesity	1135 (43.9%)	1802 (39.4%)*
Diabetes mellitus	290 (11.2%)	359 (7.9%)*
Total Cholesterol (mg/dL)	187 ± 40.3	188 ± 42.6
LDL (mg/dL)	112 ± 34.8	111 ± 34.9
HDL (mg/dL)	53.4 ± 15.7	53.4 ± 16.0
Current Tobacco Smoker	275 (10.6%)	1585 (34.7%)*
ASCVD Risk Score (%)	3.0 ± 4.1	3.6 ± 4.6*

^*p <0.05 compared with never users

Obesity = body mass index ≥ 30 kg/m²

LDL = low-density lipoprotein

HDL = high-density lipoprotein

ASCVD = atherosclerotic cardiovascular disease

Table 2 shows the results of the multinomial logistic regression. Ever cannabis use was associated with increased odds of borderline [OR (95% CI): 1.51 (1.23–1.84, p<0.001)], intermediate [OR (95% CI): 1.49 (1.26–1.78, p<0.001)], and high [OR (95% CI): 1.60 (1.04–2.45, p=0.03)] ASCVD scores (reference group: low ASCVD score). Current cannabis use was associated with increased odds of borderline [OR (95% CI): 1.32 (1.01–1.72, p=0.04)] and intermediate [OR (95% CI): 1.37 (1.09–1.73, p=0.007)] ASCVD scores (reference group: low ASCVD score). Current cannabis use was not associated with increased odds of high [OR (95% CI): 1.41 (0.80–2.51, p=0.24)] ASCVD scores (reference group: low ASCVD score).

Table 2:

Association of Cannabis Use and ASCVD Risk Category

Cannabis Use	Reference Level	ASCVD Risk Category
		Borderline	Intermediate	High
		OR (95% CI)	OR (95% CI)	OR (95% CI)
Ever User	Never User	1.51 (1.23–1.84, p<0.001)	1.49 (1.26–1.78, p<0.001)	1.60 (1.04–2.45, p=0.03)
Current User	Never User	1.32 (1.01–1.72, p=0.04)	1.37 (1.09–1.73, p=0.007)	1.41 (0.80–2.51, p=0.24)
Every month for ≥ 1 year	Never User	1.73 (1.38–2.16, p<0.001)	1.96 (1.62–2.37, p<0.001)	1.87 (1.16–3.00, p=0.01)
≥ 2 uses per month	Never User	1.78 (1.43–2.23, p<0.001)	1.90 (1.56–2.30, p<0.001)	1.79 (1.10–2.92, p=0.02)
≥ 1 use per day	Never User	1.72 (1.38–2.16, p<0.001)	1.94 (1.60–2.35, p<0.001)	1.87 (1.16–3.01, p<0.001)

OR (95% CI) = Odds Ratio (95% Confidence Interval); ASCVD = atherosclerotic cardiovascular disease;

Borderline = ASCVD score 5.0–7.4%; Intermediate = ASCVD score 7.5%−19.9%; High = ASCVD score ≥20%

We observed a dose-response relationship between increased use of cannabis and a higher risk of ASCVD. Reported cannabis use at least once per month for ≥1 year was associated with increased odds of borderline [OR (95% CI): 1.73 (1.38–2.16, p<0.001)], intermediate [OR (95% CI): 1.96 (1.62–2.37, p<0.001)], and high [OR (95% CI): 1.87 (1.16–3.00, p=0.01)] ASCVD scores (reference group: low ASCVD score). Reported cannabis use ≥2 times per month was associated with increased odds of borderline [OR (95% CI): 1.78 (1.43–2.23, p<0.001)], intermediate [OR (95% CI): 1.90 (1.56–2.30, p<0.001)], and high [OR (95% CI): 1.79 (1.10–2.92, p=0.02)] ASCVD scores (reference group: low ASCVD score). Reported cannabis use ≥1 time per day was associated with increased odds of borderline [OR (95% CI): 1.72 (1.38–2.16, p<0.001)], intermediate [OR (95% CI): 1.94 (1.60–2.35, p<0.001)], and high [OR (95% CI): 1.87 (1.16–3.01, p<0.001)] ASCVD scores (reference group: low ASCVD score).

Table 3 summarizes the results of the multinomial regression when stratified by race, gender, hypertension, obesity, and hyperlipidemia. Results were consistent among sub-groups analyzed. Table 4 shows ASCVD scores stratified by cannabis consumption frequency. Table 5 shows cannabis consumption stratified by hyperlipidemia status.

Table 3:

Association of Ever Cannabis Use and ASCVD Risk Category Among Sub-groups

Sub-group		ASCVD Risk Category†
		Borderline	Intermediate	High
		Odds Ratio (95% CI)	Odds Ratio (95% CI)	Odds Ratio (95% CI)	Interaction p-value *
Race	Non-white	1.23 (0.91–1.66, p=0.18)	1.12 (0.89–1.42, p=0.33)	1.59 (0.91–2.78, p=0.10)	0.17
	White	1.71 (1.31–2.24, p<0.001)	2.08 (1.59–2.72, p<0.001)	1.62 (0.84–3.14, p=0.15)
Gender	Men	1.26 (0.97–1.63, p=0.08)	1.37 (1.09–1.71, p=0.006)	1.24 (0.72–2.15, p=0.44)	0.38
	Women	1.53 (1.11–2.10, p=0.009)	1.17 (0.88–1.56, p=0.29)	1.79 (0.91–3.51, p=0.09)
Hypertension	Present	1.47 (1.14–1.90, p=0.003)	1.28 (1.04–1.57, p=0.02)	1.60 (0.99–2.59, p=0.05)	0.16
	Absent	1.58 (1.13–2.21, p=0.007)	2.89 (1.87–4.47, p<0.001)	1.56 (0.61–3.98, p=0.36)
Obesity	Present	1.26 (0.96–1.66, p=0.09)	1.43 (1.13–1.82, p=0.003)	1.42 (0.85–2.37, p=0.18)	0.76
	Absent	1.82 (1.35–2.44, p<0.001)	1.55 (1.20–2.01, p<0.001)	2.02 (0.92–4.43, p=0.08)
Hyperlipidemia	Present	1.55 (1.20–1.99, p<0.00l)	1.39 (1.11–1.72, p=0.003)	1.38 (0.87–2.20, p=0.17)	0.19
	Absent	1.39 (0.99–1.94, p=0.05)	1.61 (1.19–2.18, p=0.002)	3.32 (0.97–11.33, p=0.06)

ASCVD = atherosclerotic cardiovascular disease; 95% CI = 95% Confidence Interval; Borderline = ASCVD score 5.0–7.4%;

Intermediate = ASCVD score 7.5%−19.9%; High = ASCVD score ≥20%

^†Association of ever cannabis use and ASCVD risk category

^*Interaction p-value calculated from linear regression model

Obesity = body mass index ≥30 kg/m²

Hyperlipidemia = Total cholesterol ≥200 mg/dL, serum triglycerides ≥150 mg/dL, low-density lipoprotein ≥190 mg/dL, or use of lipid-lowering medications

Table 4:

ASCVD Risk Score and Cannabis Consumption

Cannabis Use	n	ASCVD Risk Score	p-value*
Never User	2586	2.95 ± 4.09	--
Ever User	4573	3.58 ± 4.63	<0.001
Current User	1443	3.29 ± 4.45	0.02
Every month for ≥ 1 year	2362	3.92 ± 4.89	<0.001
≥ 2 uses per month	2219	3.87 ± 4.79	<0.001
≥ 1 use per day	2351	3.91 ± 4.88	<0.001

ASCVD = atherosclerotic cardiovascular disease

^*Compared to never users

Table 5:

Cannabis Consumption and Hyperlipidemia

Cannabis Use	n	Hyperlipidemia	p-value*
Never User	2586	1097 (42.4%)	--
Ever User	4573	1966 (43.0%)	0.65
Current User	1443	498 (34.5%)	<0.001
Every month for ≥ 1 year	2362	978 (41.4%)	0.47
≥ 2 uses per month	2219	918 (41.4%)	0.46
≥ 1 use per day	2351	974 (41.4%)	0.48

Hyperlipidemia = Total cholesterol ≥200 mg/dL, serum triglycerides ≥150 mg/dL, low-density lipoprotein ≥190 mg/dL, or use of lipid-lowering medications

^*Compared to never users

Discussion:

In this cross-sectional analysis, cannabis use was significantly associated with a composite of CVD risk factors estimated using the 10-year ASCVD risk score. In addition, we also observed a dose-response relationship between increased use of cannabis and a higher risk of ASCVD. These results were consistent in subgroups stratified by demographics and comorbidities.

Cannabis is the most commonly used drug of abuse in the United States³. The use of cannabis and its derivatives continues to increase as states approve the legalization of these products for both medicinal and recreational use¹⁵. ASCVD, which comprises ischemic heart disease, ischemic cerebrovascular disease, and peripheral arterial disease, is a disease no longer restricted to older age. Recent epidemiological data have shown an upsurge in the incidence of all major types of ASCVD in young and middle-aged adults¹⁶. Understanding the potential cardiovascular risk linked to cannabis use is critical for planning preventive strategies¹⁷.

The effects of cannabis use on the various risk factors of ASCVD are inconsistent. Early reports showed that marijuana use causes tachycardia, peripheral vasodilation, and elevation in both systolic and diastolic blood pressures (BP) when supine¹⁸. Tachycardia is believed to be a result of increased sympathetic nervous system activity after marijuana use¹⁸. It is postulated that marijuana-induced sympathetic stimulation increases myocardial oxygen demand. This, along with a decreased oxygen supply due to carboxyhemoglobin formation from inhalation of products of combustion in marijuana cigarettes decreases exercise capacity¹⁹. In a recent report from the NHANES, recent cannabis use was associated with an increased systolic BP. However, no association was detected between lifetime history of cannabis use and BP levels²⁰. Among elderly adults, one prospective cohort study demonstrated a reduction in both diastolic and systolic BP after three months of medical cannabis treatment²¹. Similar inconsistent relationships appear in research at the intersection of cannabis, obesity, and diabetes. Physiologically, preclinical evidence suggests central activation of cannabinoid-1 (CB1) receptors promotes hyperphagia which would plausibly suggest an association with obesity²². Furthermore, CB1 receptors activation in liver increases de-novo fatty acid synthesis, decreases lipolysis, and induces insulin resistance²³. Despite this backdrop of physiologic evidence, epidemiological studies suggest lower obesity prevalence, lower biomarker levels of impaired glucose metabolism, and lower diabetes prevalence^11,24. In a meta-analysis, these inverse relationships remain replicable²⁵ and this “protective paradox” persists. However, a major limitation of these findings is that they were all conducted among various cross-sectional field surveys. Lastly, high-quality evidence regarding the effect of cannabis on lipoproteins remains sparse and inconsistent. In numerous translational studies, cannabinoids have been shown to modulate the immune system, alter lipid metabolism, and affect endothelial cells²⁶. This is important because inflammatory cytokines, oxidized low-density lipoprotein and macrophages play key roles in pathogenesis of atherosclerosis²⁶. Despite these findings, smoking cannabis appears to have marginal effects in a favorable direction with respect to triglycerides and HDL-C in epidemiological studies²⁷. Interestingly, Steffens et al. showed a decrease in progression of atherosclerotic lesions in murine models after oral administration of low-dose THC²⁸.

Our results corroborate recent findings linking cannabis use to premature ASCVD independent of traditional atherosclerotic risk factors or concomitant use of other drugs². Despite the known synergism between tobacco use and other illicit drugs such as cocaine and cannabis²⁹, our study adds to a growing body of evidence suggesting the importance of facilitating a discussion among young cannabis users regarding their risk for development of ASCVD. Since we excluded those with a history of CVD, our findings also have implications from a preventive cardiology standpoint.

Our study suffers from numerous limitations. The cross-sectional design is subject to temporality and residual confounding biases. Also, cannabis use was defined by self-report and is vulnerable to both reporting and recall bias, so the prevalence may have been underestimated. We have various categories of cannabis use history, but we do not have data regarding the dosage, route of administration, or periods of abstinence. Further, we do not have data regarding the type of cannabis used. Cannabis is a diverse genus that comes in a wide variety of potencies³⁰. Therefore, the differential composition of cannabis may also influence its mechanism of ASCVD. Despite these weaknesses, our study demonstrates a novel link between cannabis use and ASCVD. Other strengths include a racially diverse population and large sample size.

Cannabis use is associated with elevated ASCVD risk score after adjusting for traditional atherosclerotic risk factors.