Authors: Danielle E. Levitt (1Metabolic Health and Muscle Physiology Laboratory, Department of Kinesiology & Sport Management, Texas Tech University, Lubbock, TX), Kealey J. Wohlgemuth (2Neuromuscular and Occupational Performance Laboratory, Department of Kinesiology & Sport Management, Texas Tech University, Lubbock, TX), Emilie F. Burnham (3Front Line Mobile Health, Georgetown, TX), Michael J. Conner (3Front Line Mobile Health, Georgetown, TX), J. Jason Collier (4Laboratory of Islet Biology and Inflammation, Pennington Biomedical Research Center, Baton Rouge, LA), Jacob A. Mota (2Neuromuscular and Occupational Performance Laboratory, Department of Kinesiology & Sport Management, Texas Tech University, Lubbock, TX)
Categories: Article, First responder, ethanol, occupational health, mental health, tactical athlete
Source: Alcohol, clinical & experimental research
Doi: 10.1111/acer.15517
Authors: Danielle E. Levitt, Kealey J. Wohlgemuth, Emilie F. Burnham, Michael J. Conner, J. Jason Collier, Jacob A. Mota
Alcohol misuse is prevalent among firefighters, and associated adverse cardiometabolic health consequences could negatively impact readiness for duty. Mental health conditions may confer additional risk. Therefore, we aimed to determine whether alcohol misuse increases cardiometabolic risk among firefighters and whether mental health conditions modify these relationships.
Deidentified data from firefighters (N=2,405; 95.8% male, 38±9 yrs, 29.6±4.6 kg/m^2^) included demographics, Alcohol Use Disorders Identification Test (AUDIT) and AUDIT-C scores, mental health screening scores, anthropometrics, metabolic panel, and cardiorespiratory testing results. Differences in cardiometabolic parameters between firefighters with low AUDIT-C (<3 [females] or <4 [males]; no or low-risk alcohol use) or high AUDIT-C (≥3 [females] or ≥4 [males]; hazardous alcohol use) were determined and odds ratios for clinical risk factors were calculated. Post-traumatic stress disorder (PTSD), insomnia, depression, and anxiety were assessed as moderators.
Firefighters with high AUDIT-C had significantly (p<0.05) higher total cholesterol (TC), high density lipoprotein (HDL-C), and systolic blodd pressure (SBP) and diastolic blood pressure (DBP) and lower hemoglobin A1C (HbA1c) than those with low AUDIT-C. In unadjusted and/or adjusted analyses, those with high AUDIT-C had increased risk for overweight/obesity, hypercholesterolemia, and prehypertension/hypertension, and decreased risk for low HDL and elevated HbA1c. There were inverse moderation effects by post-traumatic stress disorder (PTSD), depression, and anxiety on relationships between AUDIT-C score and BP. Insomnia (directly) and anxiety (inversely) moderated relationships between AUDIT-C score and circulating lipids.
Firefighters with high AUDIT-C have differential cardiometabolic risk, with specific relationships altered by mental health status. Whether higher HDL and lower HbA1c with high AUDIT-C in firefighters is protective long-term remains to be explored. Overall, these results underscore the need for alcohol screening and intervention to maintain cardiometabolic health and long-term occupational readiness among firefighters.
Firefighters in the United States frequently perform their job duties in high-stress situations with substantial physical demands. Additionally, cardiovascular events (i.e., myocardial infarction) are the leading cause of line-of-duty deaths among firefighters (Campbell and Petrillo, 2023; Fahy et al., 2020). A common means of coping with stress that can also increase cardiometabolic risk is alcohol misuse, defined as binge or heavy drinking patterns, and firefighters are at high risk for Alcohol Use Disorder (AUD) (Zegel et al., 2022). Over half of male firefighters (Haddock et al., 2012) and nearly 40% of female firefighters (Haddock et al., 2017) report past-month binge drinking, and over half of firefighters of both sexes report that their average number of drinks on drinking days meets sex-specific criteria for heavy drinking (>1 standard drink per day for females, >2 standard drinks per day for males) in the United States (National Institute on Alcohol Abuse and Alcoholism, 2022). Such drinking patterns are normalized among firefighters (Haddock et al., 2022), underscoring the pervasiveness of problematic or hazardous alcohol use among these first responders.
Long-term alcohol misuse increases cardiometabolic risk factors including overweight and obese body mass index (BMI) categorizations (Fan et al., 2008), hypertension (Roerecke et al., 2018), dyslipidemia (Glaus et al., 2012), diabetes (Glaus et al., 2012), and overt metabolic syndrome (MetS) (Fan et al., 2008). These adverse effects could be particularly problematic in firefighters since they could negatively affect occupational readiness, career length, and overall risk of line of duty death. While firefighters appear to have similar or more favorable overall cardiometabolic health profiles compared to the general population (Moffatt et al., 2021), cardiovascular events are the leading cause of on-duty mortality among firefighters (Campbell and Petrillo, 2023; Fahy et al., 2020), urging reassessment of cardiometabolic risk and explanatory factors among this group. Furthermore, recent reports suggest a high prevalence of subclinical cardiometabolic disorders among firefighters which may influence reported cardiometabolic metrics (Smith et al., 2022). However, to date, only a small study in South Africa has directly examined the relationship between alcohol and cardiometabolic risk in firefighters, and found that alcohol volume was a significant predictor of higher systolic blood pressure (SBP) and diastolic blood pressure (DBP) (Ras and Leach, 2022). This emerging evidence underscores the need for understanding the contribution of alcohol use to mental and physical health (e.g., cardiometabolic risk) among firefighters.
Reasons for alcohol use and development of problematic alcohol use patterns vary between individuals and groups. Using alcohol to cope with stress- and trauma-related mental health sequelae, including post-traumatic stress (Smith et al., 2018), insomnia (Smith et al., 2018), depression (Karnick et al., 2022), and anxiety (Lebeaut et al., 2020), has been identified as a driver of alcohol misuse among firefighters. Importantly, each of these mental health conditions is associated with increased cardiometabolic risk (Ditmars et al., 2022; Rosenbaum et al., 2015; Sofi et al., 2014; Tang et al., 2017). Whether symptoms indicating the probable presence of these mental health conditions modifies the relationship between alcohol use and cardiometabolic risk among firefighters remains to be determined.
With widespread alcohol misuse among firefighters, established relationships between alcohol and cardiometabolic risk in the general population, and the unique nature of firefighters’ occupational demands, understanding the potential impacts of alcohol on cardiometabolic risk in firefighters is critical. This evidence will provide a foundation to support the development of tailored alcohol reduction interventions to improve cardiometabolic health. Furthermore, determining whether relationships between alcohol use and cardiometabolic risk are strengthened by post-traumatic stress, insomnia, depression, or anxiety among firefighters would improve identification of individuals who may benefit most from such interventions. Therefore, the primary purpose of the present study was to determine whether hazardous alcohol use increases cardiometabolic risk among firefighters in the United States. The secondary purpose was to determine whether post-traumatic stress, insomnia, depression, or anxiety modified relationships that occur between alcohol and cardiometabolic risk in this group.
Deidentified data from career firefighters (N=2,405; 95.8% male, 38 ± 9 yrs, 29.6 ± 4.6 kg/m^2^) across 42 departments in the United States were collected as part of routine, departmental physical examinations in 2021 and 2022. Patients provided written agreement for use of their data for research purposes, and the records were deidentified prior to analyses. Approval for the retrospective use of data was obtained from the Texas Tech University Institutional Review Board (IRB#: 2022–1009). Records for each individual firefighter were associated only with a unique code, and unique codes were examined to ensure only one record was included for each individual.
Prior to their physical examinations, firefighters’ demographic information (sex, age, race/ethnicity) was collected by an organizational administrator. Race/ethnicity was condensed into two categories (White and Black, Indigenous, and People of Color [BIPOC]) for analyses due to the small proportion of participants who self-identified as Hispanic or Latino, Black or African American, Asian, and other race or ethnicity. Participants were asked to complete surveys and visit a local CLIA-certified laboratory (e.g., Clinical Pathology Laboratories) for blood collection in the 5 weeks prior to physical examinations. Firefighters gained survey access approximately 6 weeks prior to their scheduled physical examinations. While most individuals completed the surveys prior to arriving for their examination visit, those who did not were allowed to complete the surveys on their physical examination day. Physical exams consisted of basic vital signs and anthropometric, body composition, and cardiopulmonary exercise assessments. All data were deidentified by organizational personnel and linked only by a unique code.
The AUDIT is a 10-item instrument used to assess risk for AUD based on past-year alcohol consumption (items 1–3; one standard drink = 14 g ethanol) and alcohol-related problems (items 4–10). AUDIT-C scores (first 3 questions of the AUDIT; consumption) range from 0–12, with scores <3 (females) or <4 (males) indicating low-risk alcohol use and scores ≥3 (females) or ≥4 (males) indicating hazardous alcohol use. We also report total AUDIT scores (range: 0–40, where a higher score indicates higher AUD risk). Total AUDIT scores <8 indicate low risk for AUD, and scores ≥8 indicate moderate (8–15) or high (16–40) risk for AUD.
The PCL-5 is a 20-item instrument used to assess PTSD symptoms in alignment with DSM-V diagnostic criteria. Total scores range from 0–80, where a higher score indicates more PTSD symptomology. Scores of 31–33 are proposed to detect clinically diagnosable PTSD, and a cutoff score of 32 was used for this analysis to distinguish between no diagnosable PTSD (reference category) and clinically diagnosable PTSD.
The ISI-7 is a 7-item scale used to assess presence and severity of insomnia. Total scores range from 0–28, where a higher score indicates more severe insomnia. Scores <8 indicate no clinical insomnia (reference category), and scores ≥8 indicate subthreshold (8–14), moderate (15–21), or severe (22–28) insomnia.
The PHQ-9 is a 9-item scale used to assess presence and severity of clinical depression. Total scores range from 0–27, where a higher score indicates more severe depression. Scores <5 indicate minimal depression (reference category), and scores ≥5 indicate mild (5–9), moderate (10–14), moderately severe (15–19), or severe (20–27) depression.
The GAD-7 is a 7-item scale used to assess anxiety severity. Total scores range from 0–21, where a higher score indicates more severe anxiety. Scores <5 indicate minimal anxiety (reference category), and scores ≥5 indicate mild (5–9), moderate (10–14), or severe (15–21) anxiety.
Participants visited their local laboratory between 5 weeks and 2 days prior to their physical examinations. They arrived at the laboratory after an overnight fast (no food for 9+ hours); only water and non-caloric beverages were allowed. While most participants completed this laboratory visit prior to their examination, an estimated 10–15% failed to do so and were asked to complete their laboratory visit in the days following their physical examination. Laboratory data are available for 2285 participants.
A lipid panel was performed to assess circulating total cholesterol (TC), high density lipoprotein cholesterol (HDL-C), non-HDL-C (including low and very low-density lipoproteins), and triglycerides (TG). Lower (HDL-C) or higher (TC, non-HDL-C, and TG) values were considered as indicators of increased cardiometabolic risk when assessed continuously. TC ≥200 mg/dl, HDL-C <50 mg/dl (females) or <40 mg/dl (males), non-HDL-C ≥130 mg/dl, and TG ≥150 mg/dl were cut-points to indicate increased cardiometabolic risk when assessed categorically (National Heart, Lung, and Blood Institute, 2002).
Fasting blood glucose (FBG) and glycated hemoglobin (HbA1c) were measured to assess basal glycemic control and prediabetes or diabetes risk. Higher values were considered as indicators of increased cardiometabolic risk when assessed continuously. FBG ≥100 mg/dl and HbA1c ≥5.4% were cut-points to indicate increased cardiometabolic risk when assessed categorically (Adams et al., 2009).
Participants arrived at their scheduled physical examination appointments between 00 am and 00 pm. They were asked not to have any heavy meals for 3 hours prior, to arrive well-hydrated, and to consume no caffeine on the day of the appointment. After confirming compliance with appointment requirements, height and body mass were assessed using a standard scale and stadiometer and used to calculate body mass index (BMI, kg/m^2^). BMI and body composition are indicators of cardiometabolic risk that may be easily obtained in the field (i.e., on-site at fire stations) using a standard scale and stadiometer (BMI) and multifrequency bioelectrical impedance analyses (BIA). To assess body composition, body fat percentage (BF%) was estimated using multi-frequency bioelectrical impedance analysis (InBody570, InBody USA, Los Angeles, CA). Resting blood pressure (BP) was measured by a trained provider using a manual sphygmomanometer and stethoscope. Body fat percentage, height, and body mass were used to calculate fat mass index (FMI; kg fat mass/m^2^) and fat-free mass index (FFMI; kg fat-free mass/m^2^). Higher (BP, BMI, BF%, FMI) or lower (FFMI) values were considered as indicators of increased cardiometabolic risk when assessed continuously. Systolic blood pressure (SBP) >120 mmHg or diastolic blood pressure (DBP) >80 mmHg, BMI ≥25 kg/m^2^, higher adiposity (BF%, FMI above median by sex), and lower fat-free mass (FFMI, below median by sex) (Takamura et al., 2017) were cut-points to indicate increased cardiometabolic risk when assessed categorically.
After vital signs, anthropometric measurements, and body composition assessment, a cardiopulmonary exercise test (CPET) was performed unless contraindicated (e.g., free of injuries or other medical contraindications to CPET). Aerobic capacity (VO2peak, ml/kg/min) and maximal power output relative to body mass (W/kg) were determined using an incremental intensity exercise test on a cycle ergometer. The procedures followed a ramped protocol that began with a 1-minute unloaded warm-up (70 revolutions per minute, rpm). Following the warm-up, participants were instructed to continue cycling at 70 rpm and the power output increased by 30 W every minute until the participant could no longer maintain the power output (i.e., cadence dropping below 55 rpm). Respiratory gases were monitored and continuously analyzed with open-circuit spirometry using a calibrated metabolic cart (Ultima Cardi02, MGC Diagnostics, St. Paul, Minnesota, USA). Data were analyzed by a 6-breath rolling average with the highest oxygen consumption being identified as the VO2peak (Wohlgemuth et al., 2024). Heart rate was monitored by a 12-lead electrocardiogram during the exercise test. Lower maximal aerobic capacity and maximal power output as measured during the CPET were considered as indicators of increased cardiometabolic risk (Earnest et al., 2013) when assessed continuously. Median splits by sex for each of these variables were used to determine cut-points for increased cardiometabolic risk when assessed categorically.
Data were assessed for the assumption of normality using Q-Q plots and for equality of variances using Levene’s test. Differences in descriptive characteristics between AUDIT groups were assessed used Chi Square tests (categorical variables) or t tests. We also performed t tests to determine whether cardiometabolic risk factors, assessed continuously, differed between those with low vs high AUDIT-C scores.
To determine the relative risk conferred by hazardous alcohol use (high AUDIT-C scores, ≥3 [females] or ≥4 [males]) for each of the cardiometabolic parameters indicated above, odds ratios were calculated. Unadjusted estimates and estimates adjusted for sex, age, and race/ethnicity, and BMI (except for body composition outcomes) were derived using binary logistic regression analyses.
To understand whether relationships between hazardous alcohol use and cardiometabolic risk differed by mental health status, moderation analyses were performed. AUDIT-C category (low vs high) was used as the predictor, mental health status (clinically diagnosable PTSD, and mild or greater insomnia, depression, and anxiety versus the reference category for each measure) were used as moderators, and each cardiometabolic health parameter (except FMI and FFMI) was used as a continuous outcome measure. Each model included AUDIT-C category, one mental health status indicator, and their interaction (moderation) to predict one outcome variable. To better understand significant moderation effects, subgroup analyses were conducted as t tests to determine differences in the appropriate cardiometabolic risk factor by AUDIT-C category at each level of the appropriate moderator.
For all analyses, the alpha level of significance was set to p≤0.05. Data are presented as mean ± standard deviation (tables, unless otherwise specified) or individual data points with mean ± standard error (t tests), relative risk point estimate (odds ratio, OR) with 95% confidence intervals (CI), and beta coefficients (moderation analyses). Moderation analyses were performed using R (version 4.1, R Core Team, R Foundation for Statistical Computing, Vienna, Austria) within the integrated development environment RStudio (Version 2023.12.1, Posit Software, PBC, Boston, MA). All other analyses were performed using SPSS Statistics (Version 29, IBM Corporation, Armonk, NY).
Demographic characteristics are shown in Table 1. Sex did not significantly differ between AUDIT-C groups. Those in the high AUDIT-C group were younger in age and had significantly higher AUDIT-C and AUDIT scores than those in the low AUDIT-C group. Moreover, a greater proportion identified as White and lesser proportions identified as Black/African American or Asian high the high vs low AUDIT-C group. The distribution of AUDIT and AUDIT-C scores are shown in Supplemental Figure S1.
BMI, FMI, FFMI, and BF% were similar between groups (Figure 1A–D; Supplemental Table S1). In regression analyses adjusted for sex, age, and race/ethnicity, but not in unadjusted regression analyses, high AUDIT-C increased the odds of an overweight or obese BMI (p=0.020; Table 3). High AUDIT-C did not modify risk for higher FMI, lower FFMI, or higher BF%. Odds ratios and 95% confidence intervals for each variable in the adjusted models can be found in Supplemental Table S2.
TC (p=0.003) and HDL-C (p<0.001) were significantly greater in the high AUDIT-C group (Figure 2 A–C; Supplemental Table S1). Non-HDL-C and TG were similar between groups (Figure 2D; Supplemental Table S1).
In unadjusted regression analyses, high AUDIT-C significantly increased the odds of high TC (p=0.013) and decreased the odds of low HDL-C (p<0.001; Table 2). After adjusting for sex, age, race/ethnicity, and BMI, these relationships remained (TC: p=0.002; HDL-C: p<0.001; Tables 2 and S2). High AUDIT-C did not modify risk for high non-HDL-C or TG.
HbA1c was significantly lower (p<0.001) in the high AUDIT-C group (Figure 2E–F; Supplemental Table S1). FBG was similar between groups (Figure 2E–F; Supplemental Table S1).
In unadjusted regression analyses, high AUDIT-C significantly decreased risk of an HbA1c indicative of prediabetes or diabetes (p<0.001; Table 2). After adjusting for sex, age, race/ethnicity, and BMI, this relationship remained (p<0.001; Tables 2 and S2). High AUDIT-C did not modify risk for elevated FBG in unadjusted or adjusted regression analyses (Tables 2 and S2).
SBP (p=0.024) and DBP (p=0.014) were significantly greater in the high AUDIT-C group (Figure 3; Supplemental Table S1). In unadjusted regression analyses, high AUDIT-C significantly increased the risk of elevated SBP (p=0.025) and DBP (p=0.007; Table 2). After adjusting for sex, age, race/ethnicity, and BMI, these relationships remained (SBP: p=0.005; DBP: p<0.001; Tables 2 and S2).
VO2peak and peak power output were similar between groups (Figure 4; Supplemental Table S1). High AUDIT-C did not modify risk for lower VO2peak or lower peak power output in unadjusted or adjusted regression analyses (Tables 2
and
S2).
Scores for each mental health survey (PCL-5, ISI-7, PHQ-9, and GAD-7) were significantly higher among those with high AUDIT-C, and a significantly greater proportion of those in the high AUDIT-C group had likely PTSD, subthreshold or greater insomnia, mild or greater depression, and mild or greater anxiety (Supplemental Table S3).
PTSD significantly moderated the relationship between alcohol use and DBP (β=−3.752, p=0.024). Subgroup analysis revealed that among those not meeting criteria for clinically diagnosable PTSD, DBP was significantly greater in the high AUDIT-C group (low AUDIT-C: 78.0 ± 6.7 mmHg, N=1173; high AUDIT-C: 78.8 ± 6.7 mmHg, N=1027; p=0.005; Figure 5A). Among those meeting criteria for clinically diagnosable PTSD, DBP did not differ between AUDIT-C groups (low AUDIT-C: 80.6 ± 8.5 mmHg, N=27; high AUDIT-C: 77.7 ± 7.4 mmHg, N=48; p=0.122; Figure 5B).
Insomnia significantly moderated the relationship between alcohol use and non-HDL-C (β=7.470, p=0.022). Subgroup analysis revealed that among those meeting criteria for no clinical insomnia, non-HDL-C did not differ between AUDIT-C groups (low AUDIT-C: 143.5 ± 38.0 mg/dl, N=742; high AUDIT-C: 141.0 ± 35.4 mg/dl, N=546; p=0.227; Figure 5C). Among those meeting criteria for subthreshold or greater insomnia, non-HDL-C also did not differ between AUDIT-C groups (low AUDIT-C: 142.1 ± 37.7 mg/dl, N=479; high AUDIT-C: 147.1 ± 42.4 mg/dl, N=505; p=0.052; Figure 5D), but the direction of the potential effect was opposite. Insomnia also significantly moderated the relationship between alcohol use and TG (β=14.055, p=0.032). Subgroup analysis revealed that among those meeting criteria for no clinical insomnia, TG did not differ between AUDIT-C groups (low AUDIT-C: 117.0 ± 67.9 mg/dl, N=622; high AUDIT-C: 114.0 ± 68.0 mg/dl, N=453; p=0.482; Figure 5E). Among those meeting criteria for subthreshold or greater insomnia, TG was significantly greater in the high AUDIT-C group (low AUDIT-C: 114.0 ± 66.6 mg/dl, N=407; high AUDIT-C: 125.1 ± 78.9 mg/dl, N=403; p=0.031; Figure 5F).
Depression significantly moderated the relationship between alcohol use and SBP (β=−2.980, p=0.011). Subgroup analysis revealed that among those meeting criteria for minimal depression, SBP was significantly greater in the high AUDIT-C group (low AUDIT-C: 122.0 ± 10.6 mmHg, N=1029; high AUDIT-C: 123.5 ± 10.9 mmHg, N=832; p=0.002; Figure 5G). Among those meeting criteria for mild or greater depression, SBP did not differ between AUDIT-C groups (low AUDIT-C: 123.6 ± 10.5 mmHg, N=174; high AUDIT-C: 122.1 ± 10.2 mmHg, N=242; p=0.167; Figure 5H). Depression also significantly moderated the relationship between alcohol use and DBP (β=−1.714, p=0.022). Subgroup analysis revealed that among those meeting criteria for minimal depression, DBP was significantly greater in the high AUDIT-C group (low AUDIT-C: 78.0 ± 6.8 mmHg, N=1027; high AUDIT-C: 79.0 ± 6.8 mmHg, N=833; p=0.001; Figure 5I). Among those meeting criteria for mild or greater depression, DBP did not differ between AUDIT-C groups (low AUDIT-C: 78.6 ± 6.5 mmHg, N=173; high AUDIT-C: 77.9 ± 6.8 mmHg, N=242; p=0.308; Figure 5J).
Anxiety significantly moderated the relationship between alcohol use and TG (β=−18.743, p=0.035). Subgroup analysis revealed that among those meeting criteria for minimal anxiety, TG did not differ between AUDIT-C groups (low AUDIT-C: 113.9 ± 63.8 mg/dl, N=897; high AUDIT-C: 120.0 ± 72.6 mg/dl, N=686; p=0.080; Figure 5K). Among those meeting criteria for mild or greater anxiety, TG also did not differ between AUDIT-C groups (low AUDIT-C: 128.7 ± 86.8 mg/dl, N=132; high AUDIT-C: 116.1 ± 77.0 mg/dl, N=170; p=0.183; Figure 5L), but the direction of the potential effect was opposite. Anxiety also significantly moderated between alcohol use and SBP (β=−2.968, p=0.015). Subgroup analysis revealed that among those meeting criteria for minimal anxiety, SBP was significantly greater in the high AUDIT-C group (low AUDIT-C: 122.0 ± 10.6 mmHg, N=1043; high AUDIT-C: 123.6 ± 10.9 mmHg, N=860; p=0.002; Figure 5M). Among those meeting criteria for mild or greater anxiety, SBP did not differ between AUDIT-C groups (low AUDIT-C: 123.2 ± 10.3 mmHg, N=160; high AUDIT-C: 121.8 ± 10.1 mmHg, N=214; p=0.176; Figure 5N).
While measures of physical capacity did not differ by AUDIT-C group and there were no moderation effects of mental health status on the relationships between AUDIT-C group and physical capacity, moderation analyses revealed simple main effects of mental health status on physical capacity (Supplemental Table S4). Specifically, main effects of PTSD and depression status were observed for VO2peak (PTSD: β=−2.086, p=0.005; β=−1.177, p=0.041) and peak work (PTSD: β=−0.207, p=0.002; β=−0.090, p=0.025) where VO2peak and peak work were lower among those meeting criteria for clinically diagnosable PTSD and for mild or greater depression.
The results from the present study are the first to demonstrate that key cardiometabolic risk differ by alcohol use among firefighters in the United States, and mental health status modifies relationships between alcohol use and specific cardiometabolic risk factors. Understanding the cardiometabolic impacts of hazardous alcohol use, compared with low-risk or no alcohol use, among firefighters is critical because alcohol misuse is highly prevalent in this population (Haddock et al., 2017, 2012) and cardiovascular events are the leading cause of on-duty firefighter mortality (Campbell and Petrillo, 2023; Fahy et al., 2020). Such understanding will allow for improved alcohol misuse screening among firefighters, improved monitoring of key risk factors among firefighters struggling with alcohol misuse, and development of more effective and proactive interventions to promote alcohol cessation and improved cardiometabolic health among firefighters with current or previous alcohol misuse.
Being categorized as overweight or obese and having higher body fat indicate increased cardiometabolic risk, while higher lean mass is protective against metabolic disease (Takamura et al., 2017). While overt differences in body composition between firefighters with versus without hazardous alcohol use in the present study were not observed, hazardous alcohol use conferred an increased risk for an overweight or obese BMI after adjustment for demographic variables. Inclusion of these demographic variables also revealed that risk for overweight or obese BMI was greater for those with male sex, BIPOC ethnicity, and older age. Given that approximately 3 in 4 firefighters in the United States are categorized as overweight or obese based on BMI (Beckett et al., 2023), and an even higher prevalence of overweight or obese BMI was observed among in the present study (>85%), it is imperative that fire administrators carefully consider addressing alcohol use among firefighters with overweight or obesity as part of an overall weight management program, especially among those in demographic groups at higher risk.
Components of the lipid profile are established factors in cardiometabolic health, where higher TC, non-HDL-C, and TG, and lower HDL-C increase risk (National Heart, Lung, and Blood Institute, 2002). In the present study, TC was significantly higher among firefighters with hazardous alcohol use. Additionally, hazardous alcohol use increased the odds of having high TC in unadjusted and adjusted models, indicating an increased lipid profile-based cardiometabolic risk. Moreover, while non-HDL-C did not differ by alcohol use categories, the majority of firefighters had high non-HDL-C, indicating increased lipid profile-based cardiometabolic risk in this population. Insomnia moderated the relationship between hazardous alcohol use and non-HDL-C. While subgroup analyses did not reveal significant differences, the direction of the relationship between alcohol use and non-HDL-C was opposite by insomnia status. This relationship was negative among those without reported insomnia symptoms and positive among those with subthreshold or greater insomnia, indicating that sleep duration and/or quality may contribute to a poorer lipid profile among firefighters with hazardous alcohol use. Alcohol is well known to decrease the time to sleep onset yet decrease sleep quality (Koob and Colrain, 2020). However, not all firefighters with high AUDIT-C reported symptoms of clinical insomnia in the present study. Other factors such as global mental health status, personality, and genetic factors including polymorphisms in the period circadian homolog 3 (PER3) gene, which helps regulate sleep, may contribute to the presence or absence of insomnia symptoms among those with alcohol misuse (Brower et al., 2012). While the mechanism underlying the differential relationships between alcohol use and non-HDL-C with or without insomnia is unknown, proteomic profiling of serum from patients with and without insomnia revealed differentially expressed proteins related to cholesterol metabolism, including upregulation of apolipoprotein B (apoB) (Wang et al., 2021), a protein critical for non-HDL-C particle assembly. These molecular data support a direct role for poor sleep quality or duration on cardiometabolic risk that may present in those with alcohol misuse. Although all firefighters should be encouraged to prioritize sleep, addressing insomnia may be particularly important in those with alcohol misuse.
Firefighters with hazardous alcohol use also had higher HDL-C and a decreased risk for low HDL-C, suggesting a possible protective factor alongside the aforementioned increased risk. This striking association between alcohol and HDL-C has been reported in other populations including those with advanced hepatic fibrosis (Vannier et al., 2022). The alcohol-dependent increase in HDL is likely due to increased transport via apolipoprotein A (apoA)I and II (De Oliveira E Silva et al., 2000). These and other apolipoproteins present on HDL-C may modify the protective effect of HDL-C against cardiometabolic disease (Corsetti et al., 2012) and should be further examined in firefighters with hazardous alcohol use. However, whether the protective effects of higher HDL-C are sufficient to offset the host of additional risk factors present among firefighters with hazardous alcohol use is doubtful and should not be construed as rationale to maintain high levels of alcohol intake.
High fasting glucose levels and glycated hemoglobin, indicating persistently high circulating glucose, are hallmarks of cardiometabolic disease risk and are used as screening tools for prediabetes or type II diabetes mellitus (T2DM). Development of these metabolic diseases result from impaired peripheral glucose uptake, impaired function of pancreatic beta cells, or a combination thereof (reviewed in (Galicia-Garcia et al., 2020)), and metabolic syndrome risk factors increase risk of these and other cardiometabolic diseases. While moderate alcohol intake is associated with reduced T2DM risk (Joosten et al., 2010), alcohol misuse increases risk for T2DM (Lu et al., 2023). In the present study, HbA1C values and risk for an HbA1C indicating prediabetes or T2DM were decreased in firefighters with hazardous alcohol use. Similar to these results, a recent analysis of National Health and Nutrition Examination Survey (NHANES) data revealed an inverse relationship between heavy alcohol use and HbA1C (Wiss, 2019). This finding is not at odds with increased risk for T2DM among people with alcohol misuse because alcohol appears to confer a non-canonical T2DM pathogenic trajectory, where dysfunction of the endocrine pancreas is evident prior to the development of insulin resistance (Yang et al., 2020). Moreover, without elevations in fasting glucose, people with harmful alcohol use in an older sample were more likely to meet criteria for prediabetes based on an oral glucose tolerance test (Primeaux et al., 2021). Therefore, the decrease in HbA1C among firefighters with hazardous alcohol use in this analysis likely do not translate to longer-term decreased risk for T2DM. Additional measures (e.g., glucose tolerance testing) could reveal important nuances in glycemic control among firefighters with hazardous alcohol use. Finally, in this cohort, the average FBG (96.1 mg/dl) was near the threshold for prediabetes (100 mg/dl). The average HbA1C (5.5%) was at a level associated with increased risk for T2DM (Adams et al., 2009). Together, these findings indicate the need for education and intervention targeting improvements in glycemic control.
Hypertension is a cardiovascular disease that increases risk for other cardiometabolic diseases, and such increased risk begins with even small elevations in blood pressure above 120/80 mmHg (Kshirsagar et al., 2006). Alcohol is a well-known risk factor for hypertension, and physical activity is a first-line prevention and treatment strategy for alcohol-induced hypertension (Husain et al., 2014). Given the physically demanding nature of firefighting, it is plausible that relationships between alcohol and blood pressure may have been different among firefighters compared to the general population. However, significantly higher systolic and diastolic blood pressure was observed among firefighters with hazardous alcohol use and higher odds of elevated blood pressure (120/80 mmHg or higher). It is notable that the average SBP in this cohort exceeded the threshold for prehypertension, and the average DBP was at the high end of normal. Given the potential for hypertension to contribute to cardiovascular events, and cardiovascular events being the leading cause of line of duty death in firefighters (Campbell and Petrillo, 2023; Fahy et al., 2020), blood pressure among all firefighters, and especially among those with hazardous alcohol use, should be closely monitored and interventions to decrease alcohol use and blood pressure initiated.
When stratified by PTSD, depression, or anxiety status, the alcohol-induced increases in blood pressure in firefighters without symptoms of these mental health conditions were absent in those reporting meeting criteria for PTSD or for mild or greater depression or anxiety. Alcohol use is a well-documented negative coping strategy for PTSD, depression, and anxiety, including among firefighters (Karnick et al., 2022). While alcohol increases the risk for hypertension, depression and anxiety are associated with lower blood pressure (Bhat et al., 2017). Therefore, physiological changes associated with the presence of these mental health conditions may have limited the impact of alcohol on blood pressure in the present study. In contrast, treatment of these commonly co-occurring mental health conditions with antidepressant medications may have unintended adverse effects on blood pressure (Licht et al., 2009). Significant barriers to treatment for mental health conditions have been reported (Haugen et al., 2017), suggesting that although the use of medications to treat PTSD, depression, or anxiety is unknown in the present study, it is likely low. Such information would aid in further dissecting the impact of mental health on the relationship between hazardous alcohol use and risk for prehypertension or hypertension. Notably, the median blood pressure for all groups was above the optimal 120/80 mmHg cut point, suggesting that a substantial proportion of firefighters would benefit from blood pressure reduction efforts.
Alcohol cessation is an ideal strategy to begin addressing elevated blood pressure and hypertension among firefighters with hazardous alcohol use. However, only about 1 in 6 individuals with problematic alcohol use receives treatment (Mekonen et al., 2021) and relapse is common (Durazzo and Meyerhoff, 2017). Therefore, other strategies are needed to manage blood pressure among firefighters with hazardous alcohol use. Oxidative stress and damage have been identified as key factors contributing to alcohol-associated vascular dysfunction and hypertension (Husain et al., 2014). To combat oxidative stress, antioxidant supplementation, such as with N-acetylcysteine (Martina et al., 2008), may be a feasible blood pressure management strategy in firefighters with hazardous alcohol use.
Firefighters with greater aerobic capacity have decreased risk for a host of cardiometabolic disease risk factors (McAllister et al., 2023, 2022). While results from a single study indicate a positive association between alcohol intake and physical activity among firefighters (Ras and Leach, 2022), to our knowledge, differences in firefighters’ physical capacity by alcohol use status has not previously been examined. Across the whole cohort and stratified by mental health status, hazardous alcohol use did not adversely impact cardiorespiratory capacity or peak power output. This result is promising since physical capacity predicts performance on an occupational task-related test to indicate firefighters’ readiness for duty (Wohlgemuth et al., 2024). However, the average VO2peak in this cohort was 32.8 ml O2/kg/min, which is similar to previous reports in career firefighters (Houck et al., 2020) but less than the 42 ml O2/kg/min recommended by the NFPA (National Fire Protection Association, 2022). Therefore, improving the physical capacity of firefighters should be a point of focus among fire administrators.
While physical capacity did not differ between those with and without hazardous alcohol use, the presence of PTSD or depression was independently associated with lower physical capacity. Moreover, AUDIT-C scores were significantly higher among those with these mental health concerns. These findings have implications for the occupational readiness and cardiometabolic disease risk among firefighters with PTSD or depression regardless of alcohol use. Importantly, alcohol is a common coping mechanism for these conditions (Karnick et al., 2022) and AUD, PTSD, and depression are highly comorbid (Paulus et al., 2017). Veterans with PTSD were less likely than those without PTSD to report regular exercise engagement (Adams et al., 2020), and among older adults, the presence of depression at the onset of an exercise program predicted attrition (Ha et al., 2023). While exercise engagement in the present study was not assessed, it is possible that firefighters with PTSD and depression had less exercise engagement than those without, promoting adverse changes in key physiological factors that maintain physical capacity. Since exercise training can improve PTSD and depression symptoms (McGranahan and O’Connor, 2021), exercise interventions alongside mental health support for firefighters with PTSD, depression, and hazardous alcohol use may simultaneously improve physical capacity and decrease the need to use alcohol as a coping mechanism.
While the findings herein are the first to assess hazardous alcohol use and cardiometabolic risk among firefighters in the United States, some limitations should be noted. The present study assessed hazardous alcohol use using a self-report measure, the AUDIT-C questionnaire, and there is always a risk of under-reporting. However, the AUDIT-C is considered a valid and reliable measure for identifying individuals with hazardous alcohol use (Bush et al., 1998) and is widely represented in the alcohol literature. Future studies should include additional alcohol use assessments such as biomarkers (e.g., phosphatidylethanol) and surveys measuring shorter- and longer-term alcohol use (e.g., Timeline Follow Back and Lifetime Drinking History questionnaires). Adjusted odds ratio models revealed that sex, race/ethnicity, BMI, and age contributed to cardiovascular risk. Although in-depth discussion of these factors was beyond the scope of this work, there was increased risk associated with being male (high BMI, non-HDL-C, FBG, HbA1c, SBP, DBP), BIPOC (high BMI, FMI, FBG, HbA1c), increasing age (all risk factors examined), and having an overweight or obese BMI (low HDL-C, VO2peak, peak power output; high non-HDL-C, TG, FBG, SBP, DBP). Re-examination of these risk factors in a more diverse cohort would help to identify more nuanced sex-, race-, or age-dependent differences that this study was not designed to detect. Since there are sex-differential effects in risks associated with alcohol misuse and the majority of firefighters in this cohort were male, the findings herein may have limited generalizability to female firefighters. There are also many reports of positive associations between physical activity and alcohol use, and physical activity can protect against cardiometabolic risk. Therefore, assessments of habitual physical activity should be included in firefighter health assessments and could aid in better understanding relationships between alcohol use and cardiometabolic risk in firefighters. Finally, there are many risk factors for, and factors that protect against, cardiometabolic disease. It is not possible to assess all factors in a single study. Based on the results of the present work together with existing literature, exercise engagement, dietary habits, and additional physiological and molecular measures to assess cardiometabolic risk (e.g., glucose tolerance testing, endothelial function, oxidative stress and damage, etc.) will be beneficial to include in future studies to further dissect the relationship between alcohol misuse and cardiometabolic risk in firefighters.
Firefighters with hazardous alcohol use has increased risk of overweight and obesity, dyslipidemia, and elevated blood pressure even after adjusting for demographic variables. In contrast, risks for low HDL-C and high HbA1C were decreased among firefighters with hazardous alcohol use. These relationships between alcohol intake with HDL-C and HbA1C are widely reported in the literature and do not necessarily reflect decreased cardiometabolic risk. Mental health conditions including PTSD and mild or greater insomnia, depression, and anxiety were more prevalent among firefighters with hazardous alcohol use and moderated the relationships between alcohol and cardiometabolic risk. Physical capacity, indicative of cardiometabolic risk and readiness for duty, did not differ by alcohol use but was lower among firefighters who reported symptoms indicating PTSD or depression. Since alcohol misuse is highly prevalent among firefighters, these results provide a foundation to begin addressing cardiometabolic health in firefighters most at risk. For example, addressing diet and exercise alongside mental health counseling and a comprehensive alcohol cessation program may be warranted. Additional strategies, such as antioxidant supplementation, may be useful in mitigating risk. While hazardous alcohol use increased risk for several cardiometabolic parameters, the overall cohort, representing career firefighters from 42 departments in the United States, had an overall high cardiometabolic risk profile and low physical capacity. This suggests that widespread health and fitness programs would benefit firefighters regardless of alcohol use status. To build on these results, future work should include comprehensive lifestyle assessments, examine the role of alcohol drinking patterns and beverage types, and assess additional parameters in each of the cardiometabolic health domains examined herein. Such work will help improve cardiometabolic health among firefighters, especially the substantial subset with alcohol misuse, to prevent on-duty deaths from cardiovascular events and maintain readiness for duty over the course of firefighters’ careers.