Greater perceived stress and lower self-regulation associated with substance use behaviors in a national sample of United States emergency medical service clinicians.

Greater perceived stress and lower self-regulation associated with substance use behaviors in a national sample of United States emergency medical service clinicians

Abstract

Background:

Emergency medical service (EMS) clinicians are first responders who can experience work-related stressors. Substances can be used as a coping strategy to manage stress. The ability to self-regulate motivation, cognition, and emotions may protect against substance use. Our objective was to examine the associations among stress, self-regulation, and substance use in EMS clinicians to identify potential psychological targets for substance use prevention.

Methods:

We invited a random sample of 19,000 full-time United States EMS clinicians from the largest national EMS database, of whom 1510 completed the survey. The online survey distributed by email included validated scales for perceived stress and self-regulation, as well as items assessing the number of days with binge drinking, cannabis use, and nicotine use in the prior month. We examined substance use outcomes using multivariable zero-inflated negative binomial models weighted to reflect population sociodemographic distributions.

Results:

Each 1-point increase in perceived stress was associated with 21% more binge drinking days (aIRR=1.21 [1.05, 1.39]), 79% more cannabis use days (aIRR=1.79 [1.22, 2.61]), and 5% more nicotine use days (aIRR=1.05 [1.01, 1.10]). In contrast, each 1-point increase in self-regulation was associated with 97% lower odds of using cannabis at all (aOR=0.03 [0.01, 0.57]) and 33% fewer binge drinking days (aIRR=0.67 [0.51, 0.88]).

Conclusions:

In this national cross-sectional study of full-time United States EMS clinicians, we identified stress and self-regulation as factors significantly associated with substance use. Future studies should determine whether workplace and individualized interventions targeting perceived stress and self-regulation help reduce substance use among EMS clinicians.

Keywords: Emergency medical service, Perceived stress, Self-regulation, Alcohol binge drinking, Cannabis, Nicotine

1.Introduction

Emergency medical service (EMS) clinicians are first responders who deliver life-saving care in the prehospital setting (Ehlers et al., 2023; Panchal et al., 2022). Their essential job-related tasks include evaluating patients, identifying and treating life-threatening conditions, administering medications, performing procedural interventions, and transporting patients to the hospital (Ehlers et al., 2023; Panchal et al., 2022). EMS clinicians perform these essential job tasks while only receiving an average wage of $17 per hour (U.S. Bureau of Labor Statistics, 2025). In the process of delivering this high-intensity prehospital care, EMS clinicians are frequently placed in isolated and unpredictable situations (e.g., extreme weather, unsafe buildings, remote forests) often with limited access to life-saving equipment, sufficient personnel, and medical oversight (Alghamdi, 2022; Bonumwezi et al., 2022; Huang et al., 2022; Mountfort and Wilson, 2022).

EMS clinicians report higher rates of substance use during their EMS careers compared to before entering the field (Basting et al., 2023). Stress is related to the use of alcohol (Keyes et al., 2011; Sinha, 2022), cigarettes (McKee et al., 2003; Slopen et al., 2013), and cannabis (Glodosky and Cuttler, 2020) among other populations of adults. Repeated exposure to difficult situations can increase stress, which could lead some first responders to use substances as a maladaptive coping strategy (Gill and Lauren, 2025; Huang et al., 2022). Specifically, the Self-Medication Hypothesis suggests that individuals may use substances to alleviate distress and tension (Khantzian, 1985). While this can provide short-term relief, reliance on this strategy over time can contribute to substance use disorders and dependence, regardless of stressors (Khantzian, 1985).

In one study, 33% of the 3000 first responders surveyed reported using substances to manage their emotional distress (Witkowski et al., 2023). The most frequently endorsed substances among first responders are alcohol, cannabis, and nicotine, with substance co-use commonly reported (Gabay et al., 2025). Specifically, firefighters who smoke cigarettes are five times more likely to endorse binge drinking (Jitnarin et al., 2015, 2013) and significantly more likely to endorse cannabis use (Testoff et al., 2025) compared to those who do not smoke. While known associations exist between stress and higher likelihood of substance use among convenience samples of first responders (Arble and Arnetz, 2017; Gabay et al., 2025; Witkowski et al., 2023), studies have primarily included composite measures of substance use (e.g., the nonspecific use of any substances), which limits the identification of stress as a potential risk factor across different substance use behaviors (Arble and Arnetz, 2017; Gabay et al., 2025; Witkowski et al., 2023).

In contrast to stress, studies in general adult populations have found that greater self-regulation is associated with less substance use (González-Roz et al., 2024; Hand et al., 2024; Roos et al., 2020). Self-regulation is frequently defined as the ability to control motivations, cognition, and emotions to avoid immediate gratification and better achieve long-term goals (Eisenberg et al., 2018). Across first responders and other adult populations with health-risk behaviors, lower self-regulation has been associated with greater substance use risks (Arble and Arnetz, 2017; Chopko et al., 2013; de Ridder and de Wit, 2008; Díaz-Tamayo et al., 2022; Roos and Witkiewitz, 2017). Prior studies have only examined composite measures of substance use or only one substance in isolation. No studies of first responders have examined self-regulation and the use of individual substances, despite these substances all having different health risks, use patterns, and treatment options (Lee et al., 2020; Sokolovsky et al., 2020; Vergara et al., 2018).

Among all first responder groups, including career police officers and professional firefighters, EMS clinicians might be among those at the highest risk for experiencing mental health and substance use (Gill and Lauren, 2025; Huang et al., 2022). Therefore, in line with the Self-Medication Hypothesis (Khantzian, 1985), this cross-sectional study examined the associations of perceived stress and self-regulation with the frequency of prior-month binge drinking, cannabis use, and nicotine use in a national random sample of United States EMS clinicians. We hypothesized that greater perceived stress would be associated with more frequent substance use, whereas greater self-regulation would be associated with less frequent use. Since the most commonly endorsed substances among first responders include alcohol, cannabis, and nicotine, we specifically examined associations across these three substances (Gabay et al., 2025). These findings can help determine whether future interventions aimed at decreasing stress or increasing self-regulation might help reduce substance use behaviors among first responders.

2.Methods

2.1.Participants

Inclusion was limited to United States EMS clinicians aged ≥ 18 years who worked full time responding to 9–1–1 emergencies. We included EMS clinicians across the United States who provided independent patient care and held Emergency Medical Technician (EMT), Advanced EMT, or paramedic certifications. We excluded emergency medical responders since this certification level has a very limited scope of practice and only represents 2–3% of the national EMS database (Cash et al., 2021). We limited our inclusion criteria to EMS clinicians working in emergency care settings (e.g., immediate response to injuries and illness), because these settings involve greater exposure to occupational stressors (Cannuscio et al., 2016). We excluded individuals who only worked in community settings (e.g., vaccine clinics or physician offices).

2.2.Procedures

2.2.1.Study approval and ethics

This cross-sectional study was conducted in collaboration with the National Registry of EMTs. All participants provided informed consent. No compensation was provided. This study was approved by the American Institutes of Research Institutional Review Board (#EX00678) and Dartmouth College Committee for Protection of Human Subjects (#00032966).

2.2.2.Participant recruitment

The National Registry maintains the largest database of EMS clinicians in the United States, containing records for up to 95% of certified clinicians (Cash et al., 2021). We invited a simple random sample of 19, 000 EMS clinicians from the national EMS database who met the above eligibility criteria to complete an anonymous online survey distributed by email between April and June 2024. The sampling frame of 19,000 individuals was large due to low expected response rates in online surveys of EMS clinicians (5–8%) (Cash et al., 2019; Gregory et al., 2022; Melgoza et al., 2025; Panchal et al., 2022; Powell et al., 2025) and was designed to achieve our desired sample of 1205 based on statistical power calculations. The National Registry sent reminders two and four weeks after the initial email using the modified Dillman method (Hoddinott and Bass, 1986). The overall survey completion rate was 7.9% (n = 1510).

2.3.Measures

2.3.1.Demographics

Sociodemographic variables were collected for covariate adjustments and included age, sex, race, ethnicity, education, and EMS certification (Rivard et al., 2021). A combined minority-status variable was created since individual racial and ethnic subgroups were small. We selected these covariates for adjustment since younger age, male sex, minority status, lower education, and occupational status are associated with different substance use risks across other populations (Ellis et al., 2024; Li et al., 2023). The National Registry provided de-identified demographic and occupational characteristics from their database for all participants (Rivard et al., 2021). EMS clinicians update their data every two years during their recertification cycle.

2.3.2.Perceived stress

The Perceived Stress Scale is a 10-item validated measure of thoughts and feelings related to stress over the prior month (Cohen et al., 1983). Participants rated how often they experienced each item on a 5-point Likert scale from (0) “Never” to (4) “Very often.” The item scores were averaged to create a mean continuous score (range=0–4), where higher scores indicated greater stress (Cherry and Wilcox, 2020; Sischka et al., 2024). Internal consistency was excellent for perceived stress (α=0.91) in our sample.

2.3.3.Self-regulation

A 12-item self-regulation scale measured the ability to control motivations, cognition, and emotions to avoid seeking immediate gratification and better achieve long-term goals during the prior month (Scherer et al., 2022). The scale assessed the ability to manage emotional experiences, direct attention and awareness, set and pursue goals, and regulate risky behaviors on a 5-point Likert scale from (1) “Not at all like you” to (5) “Extremely like you.” All items were scored as described in the measure development paper (Scherer et al., 2022). Item scores were then averaged to create an overall continuous mean score (1–5) for each participant, where higher scores indicated greater self-regulation. Internal consistency was good for self-regulation (α=0.85) in our sample.

2.3.4.Substance use frequency

Our primary outcome of interest in the models was substance use frequency. Three separate items from the National Survey on Drug Use and Health were used to measure the number of days in the prior month that participants engaged in (1) alcohol binge drinking (defined in the survey as 4 + standardized drinks in one sitting for women and 5 + in one sitting for men), (2) cannabis/marijuana use, and (3) nicotine use (Substance Abuse and Mental Health Services Administration [SAMHSA], 2024). For example, the survey provided a definition of a standardized alcohol drink and then asked participants, “During the past month, on how many days did you have [branched: men = ‘5’; women = ‘4’] or more drinks on the same occasion? By ‘occasion,’ we mean at the same time or within a couple of hours of each other.” For each separate substance use item, participants entered the number of days that they used that substance as a count ranging from 0 (none) to 30 days (every day) within the prior month.

2.3.5.Substance use motives

To better understand the motivation for substance use in EMS clinicians, we descriptively examined two substance use items from the Brief COPE. The Brief COPE is a validated measure of coping strategies used in response to stressful events, and was slightly adapted to ask how often participants engaged in substance use to cope with “stressful EMS work-related events” in the prior month (Carver, 1997). Participants rated how often they used alcohol or other drugs to feel better or to help them get through it on a scale from (1) “I haven’t been doing this at all” to (4) “I’ve been doing this a lot” (Carver, 1997). We first descriptively examined the extent to which EMS clinicians endorsed substance use to cope with work-related events (continuous score) and then dichotomized the items into never (score: 1) vs. at least some (scores: 2–4) to report on the proportion of participants who used substances to help manage their work stress.

2.4.Statistical analyses

2.4.1.Software and missing data

Statistical analyses were conducted using Stata v.18.0 (StataCorp LLC., College Station, TX). To address missingness in covariates merged from the national EMS database, we used a missing-category (missing-indicator) approach and created a “missing/unknown” level to retain records with incomplete covariate data for participants. This approach helped preserve the sample size and minimize potential selection bias from more restrictive complete-case analysis.

2.4.2.Survey weighting

In recent years, online survey response rates have been declining substantially, especially among healthcare professionals (Timmins et al., 2023). It is common to have response rates of 10–15% among healthcare professionals (Sammut et al., 2021), and 5–8% in EMS clinicians specifically (Cash et al., 2019; Gregory et al., 2022; Melgoza et al., 2025; Panchal et al., 2022; Powell et al., 2025). Therefore, these surveys have high risk for nonresponse bias (Lynn, 1996). Survey weighting to help correct for nonresponse bias is a common methodology to reproduce the known sociodemographic distributions from the random sampling frame and adjust for the under- or over-representation of certain respondents (Jensen et al., 2022). Therefore, we applied survey weighting to help account for nonresponse bias and increase the generalizability of our results to the broader workforce of United States EMS clinicians (Jensen et al., 2022). We calculated weights by comparing distributions of age, sex, minority status, education, and EMS certification among the 19,000 randomly sampled clinicians with those among the 1510 respondents. Composite base weights were assigned to individuals to account for numerous variables in the dataset. Our sampling frame was a simple random sample of the national EMS database, and thus intended to be representative of the larger EMS population.

2.4.3.Primary models

We fit three survey-weighted multivariable zero-inflated negative binomial (ZINB) regression models. These included one for each substance-use outcome (e.g., alcohol binge drinking, cannabis use, and nicotine use). ZINB models were selected because the count outcomes were non-negative, right-skewed (Prost et al., 2021), overdispersed (e. g., variance ≥ mean) (Ver Hoef and Boveng, 2007), and contained excess zeros (Yang et al., 2010). The substance use frequency data (see Section 2.3.4 Substance Use Frequency) were entered into the ZINB model as count outcomes. The zero-inflation component modeled the odds of belonging to the excess-zero group. We inverted estimates from this component so that adjusted odds ratios represented the odds of any use of a certain substance (≥ 1 day/month) versus no use (0 days/month). Results from this component are presented as adjusted odds ratios (aORs) with 95% confidence intervals (CI) per 1-unit increase in the exposure. The negative-binomial count component estimated relative differences in the expected number of substance-use days after accounting for the excess-zero process (Ver Hoef and Boveng, 2007). Results from the count component are presented as adjusted incidence rate ratios (aIRRs) with 95% CIs and interpreted as the relative change in the expected number of days with substance use per 1-unit increase in the exposure.

2.4.4.Adjusted covariates

All ZINB models adjusted for age, sex, minority status, education, and level of EMS certification because prior research identified these characteristics as potential confounders (Ellis et al., 2024; Li et al., 2023). Prior studies suggest that the use of one substance can confound associations with another substance-use outcome (Schweinsburg et al., 2011). Therefore, given high rates of alcohol, cannabis, and nicotine co-use in first responders (Gabay et al., 2025), we controlled for the use of other substances within each model. For example, when alcohol binge drinking was our primary outcome of interest, we also adjusted for use of cannabis and nicotine. Thus, the models estimated associations with each substance use outcome after accounting for use of other substances (Schweinsburg et al., 2011). Multicollinearity among exposures was assessed in the three substance use models by examining variance inflation factors (VIF). VIF scores ranged from 1.14 to 1.16 for all three models, indicating no problematic multicollinearity (Kim, 2019).

3.Results

3.1.Descriptive characteristics

The mean age of participants was 42.6 years (SD = 11.4, range 18–79 years) (Table 1). Most participants were male (75.4%), White (91.0%), and non-Hispanic (92.7%). In total, 11.3% were classified as members of a racial or ethnic minority group. Most participants were paramedics (66.0%) and 53.7% held an associate degree or higher. Mean perceived stress scores over the prior month were moderate (Mean=1.8, SD=0.8, range=1.0–4.0) (Cohen et al., 1983). Mean self-regulation over the prior month was moderate (Mean=3.2, SD=0.4, range=1.5–4.2) (Scherer et al., 2022).

Table 1Participant descriptive characteristics.

Characteristic

Age in years,mean (SD) 42.6 (11.4)

Sex, n (%)

Female 357 (23.6)

Male 1138 (75.4)

Other/prefer not to say 15 (1.0)

Race, n (%)

Indigenous American or Alaska Native 19 (1.3)

Asian 7 (0.5)

Black or African American 32 (2.1)

Native Hawaiian or Pacific Islander 7 (0.5)

White 1374 (91.0)

Prefer not to say 71 (4.7)

Ethnicity, n (%)

Hispanic or Latino 110 (7.3)

Not Hispanic or Latino 1400 (92.7)

Minority statusa, n (%)

Minority 170 (11.3)

Not a minority 1340 (88.7)

Education, n (%)

High school graduate/GED 181 (12.0)

Some college 502 (33.2)

Associate degree 344 (22.8)

Bachelor’s degree or higher 467 (30.9)

Unknown 16 (1.1)

EMS certification, n (%)

EMT or Advanced EMT 514 (34.0)

Paramedic 996 (66.0)

Note: Categorical or dichotomous data are reported as n (%). Continuous data are reported as mean (SD);

aMinority status variable was defined as Non-White race and/or Hispanic or Latino ethnicity. EMT = Emergency Medical Technician; EMS = Emergency Medical Service; GED = General Educational certification.

3.2.Substance use frequency

In the prior month, 36.5% of participants reported alcohol binge drinking, 8.3% reported cannabis use, and 36.4% reported nicotine use (Table 2 and Fig. 1). Overall, 56.8% of participants reported at least one of the three substance-use outcomes. Regarding co-use, 7.2% endorsed binge drinking and cannabis, 28.7% endorsed binge drinking and nicotine, and 6.2% endorsed cannabis and nicotine. Among participants reporting each substance-use outcome, the mean number of use days was 5.4 for binge drinking, 13.5 for cannabis, and 24.2 for nicotine.

Table 2Mental health and substance use characteristics.

Characteristic

Perceived stress,mean (SD) 1.8 (0.8)

Self-regulation,mean (SD) 3.2 (0.4)

Any prior-month substance use,an (%)

Any binge drinking, cannabis use, or nicotine use 858 (56.8)

Alcohol binge drinking 551 (36.5)

Cannabis use 125 (8.3)

Nicotine use 550 (36.4)

Days in prior month with substance use,bmean (SD)

Alcohol binge drinking 5.4 (6.0)

Cannabis use 13.5 (11.0)

Nicotine use 24.2 (10.2)

Any substance use to cope,an (%)

To feel better after work-related stress 504 (33.4)

To help cope with work-related stress 447 (29.7)

Frequency of substance use motives,cmean (SD)

To feel better after work-related stress 2.61 (0.80)

To help cope with work-related stress 2.61 (0.78)

Note: Categorical or dichotomous data are reported as n (%). Continuous data are reported as mean (SD).

aIndividuals could endorse more than 1 response option for substance use, so the total number of participants and percentages in the categories can exceed 100%.

bAmong participants reporting at least one day of use of the corresponding substance.

cAmong participants endorsing the corresponding motive.

3.3.Substance use motives

When participants were asked if they used substances to help cope with stressful work events, 33.4% of participants reported using alcohol or other drugs to make themselves feel better, while 29.7% reported using alcohol or other drugs to help get through it. Among participants who endorsed each motive, the mean frequency was a “little-to-medium” amount at 2.61 (SD = 0.80) for using substances to feel better and 2.61 (SD = 0.78) for using substances to get through the event (Carver, 1997). In the full sample, the corresponding frequencies were classified as “little-to-no” amounts at 1.53 (SD = 0.88) and 1.47 (SD = 0.84), respectively (Carver, 1997).

3.4.Associations with alcohol binge drinking

Among the entire sample, neither perceived stress (aOR=1.25 [0.92, 1.69]) nor self-regulation (aOR=0.80 [0.44, 1.46]) was associated with the odds of reporting any prior-month binge drinking (Table 3). In the count component for prior-month binge drinking, each 1-point increase in perceived stress was associated with a 21% higher expected number of prior-month binge drinking days (aIRR=1.21 [1.05, 1.39]), while each 1-point increase in self-regulation was associated with a 33% lower expected number of binge-drinking days (aIRR=0.67 [0.51, 0.88]).

Table 3Weighted associations of perceived stress and self-regulation with substance use.

Characteristic Binge Drinking Prior Month Cannabis Use Prior Month Nicotine Use Prior Month

Any use vs. none entire sample Expected number of use days, count Any use vs. none entire sample Expected number of use days, count Any use vs. none entire sample Expected number of use days, count

aOR (95% CI) aIRR (95% CI) aOR (95% CI) aIRR (95% CI) aOR (95% CI) aIRR (95% CI)

Perceived stress 1.25 (0.92, 1.69) 1.21 (1.05, 1.39)* 0.76 (0.31, 1.85) 1.79 (1.22, 2.61)* 0.93 (0.76, 1.14) 1.05 (1.01, 1.10)*

Self-regulation 0.80 (0.44, 1.46) 0.67 (0.51, 0.88)* 0.03 (0.01, 0.57)* 1.79 (0.87, 3.67) 0.88 (0.61, 1.28) 1.11 (0.99, 1.22)

Age, in years 0.96 (0.94, 0.98)* 1.01 (1.01, 1.02)* 0.95 (0.88, 1.02) 1.01 (0.98, 1.03) 0.97 (0.96, 0.98)* 1.00 (1.00, 1.01)

Male sex 1.43 (0.86, 2.35) 1.01 (0.79, 1.27) 1.41 (0.37, 5.57) 0.58 (0.33, 1.02) 1.37 (1.02, 1.85)* 0.98 (0.90, 1.06)

Minority status 0.79 (0.34, 1.84) 1.09 (0.79, 1.52) 1.52 (0.19, 12.87) 1.16 (0.57, 2.39) 1.05 (0.72, 1.54) 1.14 (1.01, 1.30)*

Education 0.85 (0.68, 1.07) 1.09 (0.99, 1.21) 1.28 (0.59, 2.78) 1.23 (0.86, 1.75) 0.69 (0.62, 0.79)* 0.94 (0.91, 0.97)*

Paramedic 1.39 (0.87, 2.24) 0.94 (0.76, 1.17) 1.23 (0.43, 3.57) 1.72 (1.12, 2.64)* 0.91 (0.70, 1.18) 0.96 (0.90, 1.02)

Binge drinking - - 6.16 (1.45, 34.18)* 0.80 (0.66, 0.98)* 1.49 (1.33, 1.67)* 1.00 (0.98, 1.02)

Cannabis use Non-estimablea 0.92 (0.86, 0.98)* - - 1.14 (1.03, 1.23)* 1.00 (0.97, 1.02)

Nicotine use 1.07 (0.99, 1.15) 1.08 (1.04, 1.12)* 0.93 (0.72, 1.20) 1.10 (1.01, 1.21)* - -

Note. Survey weights were derived from the sampling frame (N = 19,000) and applied to the respondent sample (n = 1510). All models adjusted for the covariates shown. An asterisk indicates that the 95% CI does not include 1.00.

aOf note, cannabis use perfectly predicted binge drinking in the zero-inflation component of the binge-drinking model, resulting in a nonestimable coefficient.

3.5.Associations with cannabis use

Among the entire sample, perceived stress was not associated with the odds of reporting any prior-month cannabis use (aOR=0.76 [0.31, 1.85]) (Table 3). However, each 1-point increase in self-regulation was associated with 97% lower odds of reporting any prior-month cannabis use (aOR=0.03 [0.01, 0.57]). In the count component for prior-month cannabis use, each 1-point increase in perceived stress was associated with a 79% higher expected number of prior-month cannabis use days (aIRR=1.79 [1.22, 2.61]), while self-regulation (aIRR=1.79 [0.87, 3.67]) was not significantly associated with the number of prior-month cannabis use days.

3.6.Associations with nicotine use

Among the entire sample, neither perceived stress (aOR=0.93 [0.76, 1.14]) nor self-regulation (aOR=0.88 [0.61, 1.28]) was associated with reporting any prior-month nicotine use. In the count component for prior-month nicotine use, each 1-point increase in perceived stress was associated with a 5% higher expected number of nicotine use days (aIRR=1.05 [1.01, 1.10]) (Table 3). However, in the count component for prior-month nicotine use, self-regulation was not associated with the number of prior-month nicotine use days (aIRR=1.11 [0.99, 1.22]).

4.Discussion

4.1.Summary of results

Greater perceived stress and lower self-regulation were modestly associated with more binge drinking days. Greater perceived stress was also associated with more cannabis and nicotine use days, although the associations with nicotine use were small in magnitude. Greater self-regulation was associated with lower odds of any cannabis use. However, the cannabis estimates were imprecise and should be considered preliminary as they warrant replication in a larger sample. Overall, these findings suggest that future research should further examine associations of these psychological factors with cannabis use and alcohol binge drinking among EMS clinicians.

4.2.Demographic characteristics

We recruited a random sample of EMS clinicians through the National EMS Database. Overall, respondents were similar to the national EMS population, but differed on several characteristics. Compared with the national EMS population, respondents included higher proportions who identified as White or Hispanic/Latino, held a bachelor’s degree or higher, and were certified as paramedics (Cash et al., 2021). The proportion of female participants was similar in the respondent sample and national population (23.6% vs. 24.2%) (Cash et al., 2021). Some studies suggest that rates of substance use might be higher among individuals identifying as non-White race and those with less education (Ellis et al., 2024; Li et al., 2023). Therefore, it is possible that rates of substance use and the resulting associations may be underestimated in our current sample compared to the overall population of EMS clinicians. However, data from the Substance Abuse and Mental Health Services Administration suggest that rates of alcohol use may actually be higher among individuals who identify as White race when compared to all other racial/ethnic groups (Center for Behavioral Health Statistics and Quality, 2021). Therefore, we cannot confidently state that rates of substance use within our sample differ from the EMS population. Additionally, full-time paramedics often have higher socio-economic status than EMTs (Cash et al., 2022), which may have contributed to underestimation of prevalence and smaller associations (Ellis et al., 2024; Li et al., 2023). EMS clinicians also report high rates of housing insecurity, food insecurity, and health care or insurance costs (Basting et al., 2023). Therefore, future studies should examine potential associations between social determinants of health and substance use among large, national samples of EMS clinicians.

4.3.Results in context

We identified a high prevalence of alcohol binge drinking and nicotine use in this national sample of EMS clinicians, where almost one-third endorsed co-use of both substances. In comparison, these substance use rates were higher than those reported among United States adults which used methods and measures similar to those used in our survey (Fig. 1) (SAMHSA, 2024). The Self-Medication Hypothesis suggests that one potential reason why individuals might engage in substance use behaviors is as a maladaptive coping strategy to alleviate distress, tension, or withdrawal symptoms from substances (Khantzian, 1985). Given that EMS clinicians work in difficult environments (Bentley et al., 2013) and can experience greater occupational stress than workers in other occupations (Gill and Lauren, 2025), they might be using legal and accessible substances (e.g., alcohol and nicotine) to help manage their stress.

Despite higher rates of binge drinking and nicotine use, only 8.3% of respondents reported prior-month cannabis use, which is lower than the 16.3% prevalence in United States adults (Fig. 1) (SAMHSA, 2024). EMS clinicians may be more likely than other adults to abstain from cannabis use given greater stigma among healthcare professionals (Coughenour et al., 2021) and continued illegality or workplace restrictions in some states. While laws vary, first responders are often subject to routine job-related drug testing, which might also contribute to lower cannabis use (SAMHSA, 2024). In a large national study of working adults, individuals from workplaces that performed routine drug testing reported lower rates of cannabis use (Carpenter, 2007). However, stigmatized behaviors are also associated with underreporting (Binkowska et al., 2025). Some respondents may have underreported cannabis use because of social-desirability concerns (Binkowska et al., 2025). Additionally, some respondents may have chosen not to disclose potentially illicit behaviors (e.g., depending on state and workplace policies) in a survey associated with their EMS certification. Therefore, rates of substance use within this survey could be influenced by both lower prevalence and underreporting of use.

In adjusted analyses, increased perceived stress was not associated with higher odds of any prior-month use (versus none) among EMS clinicians. However, the count components indicated that greater perceived stress was associated with more alcohol binge drinking, cannabis use, and nicotine use days. Therefore, perceived stress may be significantly associated with more days of use among individuals who already use these substances. One possible explanation is that stress reinforces substance use craving (Ruisoto and Contador, 2019). Acute stress responses can disrupt the brain regions responsible for inhibitory control, which can contribute to substance use craving and increased use (Sinha, 2024). Together, our results suggest that intervention studies should determine whether stress-management interventions can reduce substance-use frequency.

Associations were stronger for cannabis use and binge drinking than for nicotine use. These associations can be contextualized through frequency of use. EMS clinicians reported more episodic binge drinking (5 days/month) and cannabis use (14 days/month). However, EMS clinicians reported frequent nicotine use (24 days/month), consistent with a more habitual pattern of nicotine use (Benowitz, 2010). Therefore, EMS clinicians may engage in episodic binge drinking or cannabis use in response to increased perceived stress. This is consistent with prior epidemiological studies, which found that up to 50% of people (Han et al., 2023) who use nicotine develop dependence, while only approximately 10% of individuals develop alcohol dependence and 7–8% develop cannabis dependence (Esser et al., 2014; Kritikos et al., 2025). In other populations, participants also reported that they engaged in more cannabis use (Cuttler et al., 2018) and binge drinking (Révész et al., 2022) when their stress or anxiety levels were high.

Lastly, in adjusted analyses, greater self-regulation was significantly associated with lower odds of using any cannabis and fewer days with binge drinking in the prior month. Given the limited number of individuals who used cannabis in this sample and the concerns for model instability (e.g., wide confidence intervals), our results should be considered preliminary until replicated in other samples. However, in other populations, individuals do report cannabis use more often than any other substance as a coping tool to deal with negative emotions when self-regulation was lower (Révész et al., 2022). This might explain why greater self-regulation was associated with lower odds of using any cannabis in the prior month among all respondents. Together, our results also suggest that future intervention studies should determine whether increasing self-regulation can reduce cannabis use and binge drinking among EMS clinicians.

4.4.Implications and directions for future research

Our findings, together with prior studies (Gabay et al., 2025; Huang et al., 2022), suggest that EMS clinicians might be at greater risk for substance use than the general public (SAMHSA, 2024). Therefore, EMS clinicians may benefit from interventions tailored to their unique work environment. For example, restrictions on workplace cigarette smoking can be effective for the general public (Centers for Disease Control and Prevention, 2014; Cummings, 2016; Meza et al., 2023), but restrictions (International Association of Fire Fighters, 2019) are often less effective for first responders (Phan et al., 2022). Targeting psychological factors, such as perceived stress and self-regulation, through tailored interventions might reduce substance use in this sample.

Workplace resilience programs have been shown to improve self-regulation (e.g., goal setting, awareness) and adaptive coping among first responders (Grill et al., 2019). Additionally, general workplace support strategies, including providing dedicated spaces to decompress while at work, are negatively associated with problematic substance use in first responders, possibly by reducing their perceived stress levels (Witkowski et al., 2023). Given that alcohol and nicotine can be used to promote social bonding among first responders (e.g., sharing cigarettes while at work), interventions tailored to these unique social-occupational risks among first responders may be critically important (Haddock et al., 2022; Jitnarin et al., 2021). Finally, access to licensed mental health clinicians can reduce frontline worker stress (Rosen et al., 2022). Numerous therapies, such as dialectical behavior therapy, cognitive behavioral therapy, and mindfulness-based cognitive therapy, can help improve self-regulation (Jadhakhan et al., 2022).

However, 33% of first responders reported that stigma prevents them from seeking mental health care (Haugen et al., 2017). First responders are concerned about confidentiality, potential career repercussions, scheduling appointments with rotating schedules, and a lack of available resources (Haugen et al., 2017). Given these barriers, digital health apps and web-based interventions can provide convenient, confidential, and immediate access to stress management and self-regulation interventions (Knaak et al., 2017). Digital health interventions have been deployed to reduce perceived stress (Linardon et al., 2024, 2019), improve self-regulation (Jadhakhan et al., 2022) and decrease substance use in other populations (Staiger et al., 2020). Although stress may be unavoidable, future studies should investigate the acceptability and feasibility of digital health interventions among EMS clinicians. EMS agencies and healthcare systems should consider testing digital tools for EMS clinicians to lower substance use risks.

5.Limitations

Despite the strengths of this study, we must consider limitations. First, while the Self-Medication Hypothesis is one explanation for substance use in EMS clinicians, other potential mechanisms include (1) genetic predisposition (e.g., Quantitative Genetic Theory) (Deak and Johnson, 2021), (2) social reinforcement, such as drinking together after difficult shifts (e.g., Social Learning Theory) (Smith, 2021), and (3) interacting genetic, psychological, and social factors (e.g., Biopsychosocial Model) (Feingold and Tzur Bitan, 2022). Therefore, future research should examine genetic and social factors among EMS clinicians. Similarly, future studies should adjust for potential psychological (e.g., post-traumatic stress disorder) and occupational (e.g., hours worked) confounders (Carey et al., 2011; Witkowski et al., 2023) in the analyses, as these factors have been associated with greater mental health and substance use risks among first responders.

Second, this cross-sectional design does not permit causal inference since all measures were anchored to the prior month. We could also not perform cross-sectional mediation analyses since those methods can produce biased estimates because the temporal ordering of the variables cannot be established (Pek and Hoyle, 2016). Given the temporal ambiguity from the cross-sectional survey design, it is also possible that more substance use contributes to higher perceived stress or lower self-regulation. There are likely reverse or bidirectional associations in which substance use can also impact stress and self-regulation (Juliano et al., 2025). Therefore, future studies should also use longitudinal designs to examine temporality and directionality.

Third, although our total sample size was large and our response rate was in line with other online studies of EMS clinicians (Cash et al., 2019; Gregory et al., 2022; Melgoza et al., 2025; Panchal et al., 2022; Powell et al., 2025), nonresponse bias is likely. Respondents were more likely to identify as White race, have higher education, and be certified as paramedics compared to the National EMS Database (Cash et al., 2021). These factors may be associated with higher socioeconomic status (Cash et al., 2022) and lower rates of substance use (Ellis et al., 2024; Li et al., 2023). Therefore, it is possible that this nonresponse bias contributed to underestimated substance use prevalence and smaller effect sizes in this particular sample of EMS clinicians. It is widely accepted that national, online surveys have nonresponse bias (Lynn, 1996), so we recruited a large, random sample and performed survey weighting, which may help reduce the potential for nonresponse bias and improve sample representativeness (Jensen et al., 2022).

Fourth, substance use is subject to recall and social desirability bias (Binkowska et al., 2025). Some participants may have underreported their substance use, and specifically cannabis use, which would underestimate both the prevalence and strength of our associations. While the National EMS database allowed us to recruit a national, random sample and perform robust survey weighting, we suggest that future studies determine whether recruiting EMS clinicians through independent sources (e.g., social media advertisements, online blog posts) results in significantly different prevalence rates of cannabis use compared to our current random sample. Additionally, the small number of participants reporting cannabis use may have contributed to the large observed confidence intervals and could indicate model instability. Therefore, we should interpret the effect sizes for cannabis use with caution until subsequent studies are conducted. We consider cannabis estimates to be preliminary until replicated in future studies.

Fifth, while we invited a random, nationally representative sampling frame of EMS clinicians, our survey respondents were predominantly White and non-Hispanic. The survey weights likely reduced some nonresponse bias but cannot entirely eliminate this threat to external validity (Jensen et al., 2022). Given that rates of substance use can be higher for individuals identifying as non-White race (Li et al., 2023), future studies should also consider oversampling individuals from racial and ethnic minority groups to increase generalizability to the entire EMS population.

6.Conclusions

This cross-sectional study identified psychological factors associated with substance use in a national sample of United States EMS clinicians. Greater perceived stress was significantly associated with more days of binge drinking, cannabis use, and nicotine use. In contrast, greater self-regulation was significantly associated with lower odds of endorsing any cannabis use at all and fewer days of binge drinking. Future longitudinal studies should examine temporal patterns, while future intervention studies should determine whether workplace and tailored interventions targeting both perceived stress and self-regulation reduce substance use among EMS clinicians.

Acknowledgements

We would like to thank Shea van den Bergh at the National Registry of EMTs and Nancy Hoffmann at the National Association of EMTs for administrative assistance.

Funding

This work was supported by a National Institute on Drug Abuse Center grant (P30DA029926), National Institute on Drug Abuse Training grant (5T32DA037202), and National Institute on Drug Abuse Individual Predoctoral Fellowship grant (1F31DA062393).

Declaration of Competing Interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Enzo Plaitano reports financial support was provided by National Institute on Drug Abuse. Lisa Marsch reports financial support was provided by National Institute on Drug Abuse. Lisa Marsch reports a relationship with Square2 Systems that includes: consulting or advisory. Lisa Marsch reports a relationship with Boehringer Ingelheim that includes: consulting or advisory. Lisa Marsch reports a relationship with Click Therapeutics that includes: consulting or advisory. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Glossary

(EMS)Emergency Medical Service

(EMT)Emergency Medical Technician

(ZINB)zero-inflated negative binomial regression

Footnotes

CRediT authorship contribution statement

Plaitano Enzo G: Writing – review & editing, Writing – original draft, Visualization, Software, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Ashish R. Panchal: Writing – review & editing, Supervision, Resources, Methodology, Investigation. Madelyn R. Frumkin: Writing – review & editing, Supervision, Investigation. Jonathan R. Powell: Writing – review & editing, Methodology, Investigation, Data curation. Christopher B. Gage: Writing – review & editing, Methodology, Data curation. Lisa A. Marsch: Writing – review & editing, Supervision, Resources, Investigation, Funding acquisition. Christine M. Gunn: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization.

Ethics declaration

Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed.

This study was performed in accordance with relevant laws, regulatory frameworks and guidelines where the research took place. This study was approved by the Dartmouth College Committee for Protection of Human Subjects. (Approval No. 00032966)

Ethical approvals

This study was approved by the American Institutes of Research Institutional Review Board (#EX00678) and Dartmouth College Committee for Protection of Human Subjects (#00032966).

Declaration of Generative AI and AI-assisted technologies in the writing process

The authors did not use any generative AI or AI-assisted technologies.

Data availability

The raw data supporting the conclusions of this article are not yet publicly available as they are an essential component of a doctoral thesis that is still in progress. In the interim, the raw data supporting the conclusions of this article will be made available to individuals by the corresponding author upon reasonable request and will be public in the future.

References

Alghamdi AA, 2022. The psychological challenges of emergency medical service providers during disasters: a mini-review february 2022. Front. Psychiatry
13, 773100. 10.3389/fpsyt.2022.773100.

Arble E, Arnetz BB, 2017. A model of first-responder coping: an Approach/Avoidance Bifurcation. Stress. Health
33, 223–232. 10.1002/smi.2692.

Basting J, Wong J, Berger D, Caldwell C, Saggi S, Mann J, 2023. Prevalence of Social Needs & social risks among EMS providers. JEMS Exclus. 2023.

Benowitz NL, 2010. Nicotine addiction. N. Engl. J. Med. 362, 2295–2303. 10.1056/NEJMra0809890.

Bentley MA, Crawford JM, Wilkins JR, Fernandez AR, Studnek JR, 2013. An assessment of depression, anxiety, and stress among nationally certified EMS professionals. Prehosp. Emerg. Care
17, 330–338. 10.3109/10903127.2012.761307.

Best practices for comprehensive tobacco control programs - 2014. Atlanta: U.S Department of Health and Human Services, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014. Centers for Disease Control and Prevention.

Binkowska AA, Pałczyński P, Jakubowska N, Czarny J, Raczkowski M, Brzezicka A, 2025. Substance use assessment: comparing self-reports with objective data in a research setting. Front. Public. Health
13, 1628519. 10.3389/fpubh.2025.1628519.

Bonumwezi JL, Tramutola D, Lawrence J, Kobezak HM, Lowe SR, 2022. Posttraumatic stress disorder symptoms, work-related trauma exposure, and substance use in first responders. Drug. Alcohol. Depend. 237, 109439. 10.1016/j.drugalcdep.2022.109439.

Cannuscio CC, Davis AL, Kermis AD, Khan Y, Dupuis R, Taylor JA, 2016. A Strained 9-1-1 System and Threats to Public Health. J. Community Health
41, 658–666. 10.1007/s10900-015-0142-x.

Carey MG, Al-Zaiti SS, Dean GE, Sessanna L, Finnell DS, 2011. Sleep problems, depression, substance use, social bonding, and quality of life in professional firefighters. J. Occup. Environ. Med. 53, 928–933. 10.1097/JOM.0b013e318225898f.

Carpenter CS, 2007. Workplace drug testing and worker drug use. Health Serv. Res. 42, 795–810. 10.1111/j.1475-6773.2006.00632.x.

Carver CS, 1997. You want to measure coping but your protocol’s too long: consider the brief COPE. Int. J. Behav. Med. 4, 92–100. 10.1207/s15327558ijbm0401_6.

Cash RE, Powell JR, Peters GA, Goldberg SA, Panchal AR, Camargo CA Jr, 2022. Trends in demographic and employment characteristics of US emergency medical technicians and paramedics, 2011-2019. J. Am. Coll. Emerg. Physicians Open. 3, e12776. 10.1002/emp2.12776.

Cash RE, Rivard MK, Chrzan K, Mercer CB, Camargo CA Jr., Panchal AR, 2021. Comparison of volunteer and paid EMS professionals in the United States. Prehosp. Emerg. Care
25, 205–212. 10.1080/10903127.2020.1752867.

Cash RE, White-Mills K, Crowe RP, Rivard MK, Panchal AR, 2019. Workplace incivility among nationally certified EMS professionals and associations with workforce-reducing factors and organizational culture. Prehosp. Emerg. Care
23, 346–355. 10.1080/10903127.2018.1502383.

Center for Behavioral Health Statistics and Quality. (2021). Racial/ethnic differences in substance use, substance use disorders, and substance use treatment utilization among people aged 12 or older (2015–2019) (Publication No. PEP21-07-01-001). Substance Abuse and Mental Health Services Administration. https://www.samhsa.gov/data/

Cherry ML, Wilcox MM, 2020. Decreasing perceived and academic stress through emotion regulation and nonjudging with trauma-exposed college students. Int. J. Stress. Manag. 27, 101–110. 10.1037/str0000138.

Chopko BA, Palmieri PA, Adams RE, 2013. Associations between police stress and alcohol use: implications for practice. J. Loss Trauma. 18, 482–497. 10.1080/15325024.2012.719340.

Cohen S, Kamarck T, Mermelstein R, 1983. A global measure of perceived stress. J. Health Soc. Behav. 24, 385–396.

Coughenour P, Sadicario JS, Karjane N, Parlier-Ahmad AB, Phipps L, Svikis DS, 2021. Prevalence and social acceptability of cannabis, tobacco, and alcohol use in adult women. Women’s. Health Rep. (New. Rochelle)
2, 452–458. 10.1089/whr.2021.0042.

Cummings KM, 2016. Smoking isn’t cool anymore: The success and continuing challenge of public health efforts to reduce smoking. J. Public. Health Manag. Pract. 22, 5–8. 10.1097/PHH.0000000000000360.

Cuttler C, Spradlin A, McLaughlin RJ, 2018. A naturalistic examination of the perceived effects of cannabis on negative affect. J. Affect. Disord. 235, 198–205. 10.1016/j.jad.2018.04.054.

Deak JD, Johnson EC, 2021. Genetics of substance use disorders: a review. Psychol. Med. 51, 2189–2200. 10.1017/S0033291721000969.

Díaz-Tamayo AM, Escobar-Morantes JR, García-Perdomo HA, 2022. Coping Strategies for Exposure to Trauma Situations in First Responders: A Systematic Review. Prehosp. Disaster Med. 37, 810–818. 10.1017/S1049023X22001479.

Ehlers J, Fisher B, Peterson S, Dai M, Larkin A, Bradt L, Mann NC, 2023. Description of the 2020 NEMSIS public-Release Research Dataset. Prehosp. Emerg. Care
27, 473–481. 10.1080/10903127.2022.2079779.

Eisenberg IW, Bissett PG, Canning JR, Dallery J, Enkavi AZ, Whitfield-Gabrieli S, Gonzalez O, Green AI, Greene MA, Kiernan M, Kim SJ, Li J, Lowe MR, Mazza GL, Metcalf SA, Onken L, Parikh SS, Peters E, Prochaska JJ, Scherer EA, Stoeckel LE, Valente MJ, Wu J, Xie H, MacKinnon DP, Marsch LA, Poldrack RA, 2018. Applying novel technologies and methods to inform the ontology of self-regulation. Behav. Res. Ther. 101, 46–57. 10.1016/j.brat.2017.09.014.

Ellis RA, Bailey AJ, Jordan C, Shapiro H, Greenfield SF, McHugh RK, 2024. Gender differences in illicit drug access, use and use disorder: analysis of National Survey on Drug Use and Health data. J. Psychiatr. Res. 175, 118–122. 10.1016/j.jpsychires.2024.05.017.

Esser MB, Hedden SL, Kanny D, Brewer RD, Gfroerer JC, Naimi TS, 2014. Prevalence of alcohol dependence among US adult drinkers, 2009–2011. Prev. Chronic Dis. 11. 10.5888/pcd11.140329.

Feingold D, Tzur Bitan D, 2022. Addiction psychotherapy: going beyond self-medication. Front. Psychiatry
13, 820660. 10.3389/fpsyt.2022.820660.

Gabay O, Frenkel A, Biederko R, Bismuth Y, Wacht O, Grinstein-Cohen O, 2025. The association of exposure to traumatic events with binge drinking in paramedics and EMTs. Workplace Health Saf. 73, 496–502. 10.1177/21650799251339584.

Gill SK, Lauren R, K.-M., 2025. A systematic review of occupational stress and burnout in emergency medical technician and paramedic populations and associated consequences. Int. J. Qual. Health Care
37, mzaf033. 10.1093/intqhc/mzaf033.

Glodosky NC, Cuttler C, 2020. Motives matter: Cannabis use motives moderate the associations between stress and negative affect. Addict. Behav. 102, 106188. 10.1016/j.addbeh.2019.106188.

González-Roz A, Castaño Y, Krotter A, Salazar-Cedillo A, Gervilla E, 2024. Emotional dysregulation in relation to substance use and behavioral addictions: Findings from five separate meta-analyses. Int. J. Clin. Health Psychol. 24, 100502. 10.1016/j.ijchp.2024.100502.

Gregory ME, Powell JR, MacEwan SR, Kurth JD, Kenah E, Panchal AR, McAlearney AS, 2022. COVID-19 vaccinations in EMS professionals: Prevalence and predictors. Prehosp. Emerg. Care
26, 632–640. 10.1080/10903127.2021.1993391.

Grill G, Marks M, Grill J, Gunderson P, Callahan M, 2019. Marks An evidence-based program for improving and sustaining first responder behavioral health. J. Emerg. Med. Serv. 3, 1304–1309.

Haddock CK, Jitnarin N, Caetano R, Jahnke SA, Hollerbach BS, Kaipust CM, Poston WSC, 2022. Norms about alcohol use among US firefighters. Saf. Health Work. 13, 387–393. 10.1016/j.shaw.2022.08.008.

Han B, Einstein EB, Compton WM, 2023. Patterns and characteristics of nicotine dependence among adults with cigarette use in the US, 2006-2019. JAMA Netw. Open. 6, e2319602. 10.1001/jamanetworkopen.2023.19602.

Hand LJ, Paterson LM, Lingford-Hughes AR, 2024. Re-evaluating our focus in addiction: emotional dysregulation is a critical driver of relapse to drug use. Transl. Psychiatry
14, 467. 10.1038/s41398-024-03159-5.

Haugen PT, McCrillis AM, Smid GE, Nijdam MJ, 2017. Mental health stigma and barriers to mental health care for first responders: a systematic review and meta-analysis. J. Psychiatr. Res. 94, 218–229. 10.1016/j.jpsychires.2017.08.001.

Hoddinott SN, Bass MJ, 1986. The dillman total design survey method. Can. Fam. Physician
32, 2366–2368.

Huang G, Chu H, Chen R, Liu D, Banda KJ, O’Brien AP, Jen H-J, Chiang K-J, Chiou J-F, Chou K-R, 2022. Prevalence of depression, anxiety, and stress among first responders for medical emergencies during COVID-19 pandemic: a meta-analysis. J. Glob. Health
12, 05028. 10.7189/jogh.12.05028.

International Association of Fire Fighters. (2019). IAFF-IAFC fire service joint labor management wellness-fitness initiative. http://www.iaff.org/hs/wfiresource/default.html.

Jadhakhan F, Blake H, Hett D, Marwaha S, 2022. Efficacy of digital technologies aimed at enhancing emotion regulation skills: literature review. Front. Psychiatry
13, 809332. 10.3389/fpsyt.2022.809332.

Jensen HAR, Lau CJ, Davidsen M, Feveile HB, Christensen AI, Ekholm O, 2022. The impact of non-response weighting in health surveys for estimates on primary health care utilization. Eur. J. Public. Health
32, 450–455. 10.1093/eurpub/ckac032.

Jitnarin N, Haddock CK, Poston WSC, Jahnke S, 2013. Smokeless tobacco and dual use among firefighters in the central United States. J. Environ. Public. Health
2013, 675426. 10.1155/2013/675426.

Jitnarin N, Poston WSC, Haddock CK, Jahnke SA, Day RS, 2015. Tobacco use pattern among a national firefighter cohort. Nicotine Tob. Res. 17, 66–73. 10.1093/ntr/ntu131.

Jitnarin N, Poston WSC, Jahnke SA, Haddock CK, Kelley HN, Severson HH, 2021. A qualitative study to assess perceptions, barriers, and motivators supporting smokeless tobacco cessation in the US fire service. PLoS. One
16, e0251128. 10.1371/journal.pone.0251128.

Juliano VAL, Albernaz-Mariano KA, Covre LHH, Jucá PM, Pereira RM, Shigeo-de-Almeida A, Sampaio LL, Duque E, de A, Munhoz CD, 2025. Neurobiological intersections of stress and substance use disorders. Front. Neurosci. 19, 1548372. 10.3389/fnins.2025.1548372.

Keyes KM, Hatzenbuehler ML, Hasin DS, 2011. Stressful life experiences, alcohol consumption, and alcohol use disorders: the epidemiologic evidence for four main types of stressors. Psychopharmacol. (Berl. )
218, 1–17. 10.1007/s00213-011-2236-1.

Khantzian EJ, 1985. The self-medication hypothesis of addictive disorders: focus on heroin and cocaine dependence. Am. J. Psychiatry
142, 1259–1264. 10.1176/ajp.142.11.1259.

Kim JH, 2019. Multicollinearity and misleading statistical results. Korean J. Anesthesiol. 72, 558–569. 10.4097/kja.19087.

Knaak S, Mantler E, Szeto A, 2017. Mental illness-related stigma in healthcare: Barriers to access and care and evidence-based solutions. Healthc. Manag. Forum
30, 111–116. 10.1177/0840470416679413.

Kritikos AF, Taylor B, Lamuda P, Pollack H, Schneider JA, 2025. Patterns of past month cannabis consumption and cannabis use disorder - Insights from a nationally representative survey. Drug. Alcohol. Depend. 272, 112680. 10.1016/j.drugalcdep.2025.112680.

Lee CM, Patrick ME, Fleming CB, Cadigan JM, Abdallah DA, Fairlie AM, Larimer ME, 2020. A daily study comparing alcohol-related positive and negative consequences for days with only alcohol use versus days with simultaneous alcohol and marijuana use in a community sample of young adults. Alcohol. Clin. Exp. Res. 44, 689–696. 10.1111/acer.14279.

Li B, Allebeck P, Burstöm B, Danielsson A-K, Degenhardt L, Eikemo TA, Ferrari A, Knudsen AK, Lundin A, Manhica H, Newton J, Whiteford H, Flodin P, Sjöqvist H, Agardh EE, 2023. Educational level and the risk of mental disorders, substance use disorders and self-harm in different age-groups: A cohort study covering 1,6 million subjects in the Stockholm region. Int. J. Methods Psychiatr. Res. 32, e1964. 10.1002/mpr.1964.

Linardon J, Cuijpers P, Carlbring P, Messer M, Fuller-Tyszkiewicz M, 2019. The efficacy of app-supported smartphone interventions for mental health problems: a meta-analysis of randomized controlled trials. World Psychiatry
18, 325–336. 10.1002/wps.20673.

Linardon J, Firth J, Torous J, Messer M, Fuller-Tyszkiewicz M, 2024. Efficacy of mental health smartphone apps on stress levels: a meta-analysis of randomised controlled trials. Health Psychol. Rev. 18, 839–852. 10.1080/17437199.2024.2379784.

Lynn P, 1996. Weighting for non-response. In: Banks R (Ed.), Survey and Statistical Computing. Association for Survey Computing, pp. 205–214.

McKee SA, Maciejewski PK, Falba T, Mazure CM, 2003. Sex differences in the effects of stressful life events on changes in smoking status. Addiction
98, 847–855. 10.1046/j.1360-0443.2003.00408.x.

Melgoza E, Farcas A, Powell JR, Gage CB, Crowe R, Bosson N, Joiner A, Haamid A, Schlesinger S, Panchal AR, 2025. Factors associated with EMS clinician preparedness to provide care for patients with limited English proficiency. Prehosp. Emerg. Care
16. 10.1080/10903127.2025.2524744.

Meza R, Cao P, Jeon J, Warner KE, Levy DT, 2023. Trends in US adult smoking prevalence, 2011–2022. JAMA Health Forum
4, e234213. 10.1001/jamahealthforum.2023.4213.

Mountfort S, Wilson J, 2022. EMS Provider Health And Wellness. StatPearls Publishing.

Panchal AR, Rivard MK, Cash RE, Corley JP Jr, Jean-Baptiste M, Chrzan K, Gugiu MR, 2022. Methods and implementation of the 2019 EMS Practice Analysis. Prehosp. Emerg. Care
26, 212–222. 10.1080/10903127.2020.1856985.

Pek J, Hoyle RH, 2016. On the (in)validity of tests of simple mediation: Threats and solutions. Soc. Personal. Psychol. Compass
10, 150–163. 10.1111/spc3.12237.

Phan L, McNeel TS, Jewett B, Moose K, Choi K, 2022. Trends of cigarette smoking and smokeless tobacco use among US firefighters and law enforcement personnel, 1992-2019. Am. J. Ind. Med. 65, 72–77. 10.1002/ajim.23311.

Powell JR, Gage CB, Crowe RP, Rush LJ, MacEwan SR, Dixon G, McAlearney AS, Panchal AR, 2025. National evaluation of emergency medical services clinician burnout and workforce-reducing factors. J. Am. Coll. Emerg. Physicians Open. 6, 100024. 10.1016/j.acepjo.2024.100024.

Prost V, Gazut S, Brüls T, 2021. A zero inflated log-normal model for inference of sparse microbial association networks. PLoS. Comput. Biol. 17, e1009089. 10.1371/journal.pcbi.1009089.

Révész D, Bours MJL, Weijenberg MP, Mols F, 2022. Longitudinal associations of former and current alcohol consumption with psychosocial outcomes among colorectal cancer survivors 1-15 years after diagnosis. Nutr. Cancer
74, 3109–3117. 10.1080/01635581.2022.2044063.

de Ridder DTD, de Wit JBF, 2008. Self-regulation in health behavior: Concepts, theories, and central issues, in: Self-Regulation. Health Behavior. John Wiley & Sons, Ltd, West Sussex, England, pp. 1–23. 10.1002/9780470713150.ch1.

Rivard MK, Cash RE, Mercer CB, Chrzan K, Panchal AR, 2021. Demography of the National Emergency Medical Services Workforce: A Description of Those Providing Patient Care in the Prehospital Setting. Prehosp. Emerg. Care
25, 213–220. 10.1080/10903127.2020.1737282.

Roos CR, Kober H, Trull TJ, MacLean RR, Mun CJ, 2020. Intensive longitudinal methods for studying the role of self-regulation strategies in substance use behavior change. Curr. Addict. Rep. 7, 301–316. 10.1007/s40429-020-00329-5.

Roos CR, Witkiewitz K, 2017. A contextual model of self-regulation change mechanisms among individuals with addictive disorders. Clin. Psychol. Rev. 57, 117–128. 10.1016/j.cpr.2017.08.008.

Rosen B, Preisman M, Read H, Chaukos D, Greenberg RA, Jeffs L, Maunder R, Wiesenfeld L, 2022. Providers’ perspectives on implementing resilience coaching for healthcare workers during the COVID-19 pandemic. BMC Health Serv. Res. 22, 780. 10.1186/s12913-022-08131-x.

Ruisoto P, Contador I, 2019. The role of stress in drug addiction. An integrative review. Physiol. Behav. 202, 62–68. 10.1016/j.physbeh.2019.01.022.

Sammut R, Griscti O, Norman IJ, 2021. Strategies to improve response rates to web surveys: A literature review. Int. J. Nurs. Stud. 123, 104058. 10.1016/j.ijnurstu.2021.104058.

Scherer EA, Kim SJ, Metcalf SA, Sweeney MA, Wu J, Xie H, Mazza GL, Valente MJ, MacKinnon DP, Marsch LA, 2022. Momentary Self-regulation: Scale Development and Preliminary Validation. JMIR Ment. Health
9, e35273. 10.2196/35273.

Schweinsburg AD, Schweinsburg BC, Nagel BJ, Eyler LT, Tapert SF, 2011. Neural correlates of verbal learning in adolescent alcohol and marijuana users: fMRI in adolescent users. Addiction
106, 564–573. 10.1111/j.1360-0443.2010.03197.x.

Sinha R, 2022. Alcohol’s negative emotional side: The role of stress neurobiology in alcohol use disorder. Alcohol. Res. 42, 12. 10.35946/arcr.v42.1.12.

Sinha R, 2024. Stress and substance use disorders: risk, relapse, and treatment outcomes. J. Clin. Invest. 134. 10.1172/jci172883.

Sischka PE, Grübbel L, Reisinger CV, Neufang KM, Schmidt AF, 2024. On the dimensionality, suitability of sum/mean scores, and cross-country measurement invariance of the Perceived Stress Scale 10 (PSS-10)—Evidence from 41 countries. Int. J. Stress. Manag. 10.1037/str0000330.

Slopen N, Kontos EZ, Ryff CD, Ayanian JZ, Albert MA, Williams DR, 2013. Psychosocial stress and cigarette smoking persistence, cessation, and relapse over 9-10 years: a prospective study of middle-aged adults in the United States. Cancer Causes Control. 24, 1849–1863. 10.1007/s10552-013-0262-5.

Smith MA, 2021. Social learning and addiction. Behav. Brain Res. 398, 112954. 10.1016/j.bbr.2020.112954.

Sokolovsky AW, Gunn RL, Micalizzi L, White HR, Jackson KM, 2020. Alcohol and marijuana co-use: Consequences, subjective intoxication, and the operationalization of simultaneous use. Drug. Alcohol. Depend. 212, 107986. 10.1016/j.drugalcdep.2020.107986.

Staiger PK, O’Donnell R, Liknaitzky P, Bush R, Milward J, 2020. Mobile apps to reduce tobacco, alcohol, and illicit drug use: Systematic review of the first decade. J. Med. Internet Res. 22, e17156. 10.2196/17156.

Substance Abuse and Mental Health Services Administration. (2024). Drug testing for safety- and security-sensitive industries. https://www.samhsa.gov/substance-use/drug-free-workplace/employer-resources/safety-security-sensitive#:~:text=Most%20stat%20and%20local%20law,stat%20and%20agency%20to%20agency.

Substance Abuse and Mental Health Services Administration. (2024). 2023 NSDUH detailed tables. https://www.samhsa.gov/data/report/2023-nsduh-detailed-tables

Testoff AC, Koru-Sengul T, Solle NS, Vidot DC, Lee DJ, Kobetz EN, Caban-Martinez AJ, 2025. Predictors of ever using marijuana among US first responders. J. Occup. Environ. Med. 67, e207–e211. 10.1097/JOM.0000000000003310.

Timmins F, Ottonello G, Napolitano F, Musio ME, Calzolari M, Gammone M, Catania G, Zanini M, Aleo G, Sasso L, Bagnasco A, 2023. The state of the science-the impact of declining response rates by nurses in nursing research projects. J. Clin. Nurs. 32, e9–e11. 10.1111/jocn.16597.

U.S. Bureau of Labor Statistics. (2025). Emergency medical technicians and paramedics. https://www.bls.gov/oes/2019/may/oes292040.htm

Ver Hoef JM, Boveng PL, 2007. Quasi-Poisson vs. negative binomial regression: how should we model overdispersed count data?
Ecology
88, 2766–2772. 10.1890/07-0043.1.

Vergara VM, Weiland BJ, Hutchison KE, Calhoun VD, 2018. The impact of combinations of alcohol, nicotine, and cannabis on dynamic brain connectivity. Neuropsychopharmacology
43, 877–890. 10.1038/npp.2017.280.

Witkowski K, Lofaro RJ, Headley AM, Contreras S, Remington CL, Ganapati NE, 2023. Understanding problematic substance use among first responders during the COVID-19 pandemic: a survey of law enforcement, fire, and EMS workers in the United States. Int. J. Drug. Policy
123, 104261. 10.1016/j.drugpo.2023.104261.

Yang Z, Hardin JW, Addy CL, 2010. Score tests for zero-inflation in overdispersed count data. Commun. Stat. Theory Methods
39, 2008–2030. 10.1080/03610920902948228.

Associated Data

Data Availability Statement

The raw data supporting the conclusions of this article are not yet publicly available as they are an essential component of a doctoral thesis that is still in progress. In the interim, the raw data supporting the conclusions of this article will be made available to individuals by the corresponding author upon reasonable request and will be public in the future.