Physical demands and physiological strain of American football referees while officiating

ABSTRACT Objectives Certain occupations are associated with greater risk of triggering a sudden cardiac event because of high levels of physical exertion and extreme thermal environments in which they occur. The extent to which sports officials––particularly high school (HS) American football referees––experience these conditions is unknown. Forty-six male HS officials (72% White/Caucasian; age = 48 ± 12 years, body mass index = 31.7 ± 6.6 kg·m−2) were studied to quantify the physiological strain and physical demands of officiating. Methods Referee demographics (e.g., experience, habitual exercise), pre-game urine specific gravity (USG), thermal (peak core temperature [Tcore]) and cardiovascular (average heart rate [HR]) strain, kinematic activity (e.g., total distance, speed, mechanical intensity), and environmental conditions were measured during 10 regular season varsity HS football games (≈2.5 h each) in the Southeastern United States (average wet bulb globe temperature and relative humidity: 18.9 ± 6.0 °C and 78.2% ± 12.1%). Analyses included descriptive statistics, bivariate correlations, and linear regression. Results Referees covered 5.2 ± 1.2 km per game, eliciting average HR and peak Tcore of 71.5% ± 8.0% HRmax and 38.3 ± 0.5 °C, respectively; 38% began games dehydrated (USG = 1.026 ± 0.004). Multiple regression analyses revealed that obesity (β = 0.34), not participating in regular exercise (β = −0.36), and officiating at lower mechanical intensity (β = −0.33) predicted greater cardiovascular strain (all p ≤ 0.03). White/Caucasian race/ethnicity (β = 0.59), younger age (β = −0.46), and obesity (β = 0.28) predicted greater thermal strain (all p ≤ 0.01). Conclusion HS football referees experienced elevated levels of physiological strain while officiating, with individual factors modulating the magnitude of strain. Strategies aimed at reducing obesity, increasing exercise participation, and improving cardiovascular health should be emphasized to mitigate strain and prevent cardiac events.


Introduction
Sudden cardiac death is a leading cause of mortality in the United States, accounting for 50% of cardiovascular deaths annually (≈300,000 cases per year) [1]. Many occupational cardiac events are exertion-related, occurring most frequently among occupations characterized by high levels of physical exertion, physiological strain, and extreme thermal environments [2][3][4][5]. A high proportion of exertion-related cardiac events can also be attributed to underlying cardiovascular disease (CVD) (≈80-90%) [6]. Despite employment standards for many physically demanding occupations (e.g., firefighters, first responders, law enforcement, and military) [7], workers often have suboptimal health and fitness, creating a mismatch between the physical demands of the job and their capacity to safely perform the work [3,4,8]. Furthermore, little is known about worker health and performance in other physically demanding occupations, such as sports officiating, where workers are often hired as independent contractors, and there is little (if any) assessment of health and work capacity. This lack of screening is problematic because sports officials are typically of middle and older age, overweight to obese, have low-to-fair cardiorespiratory fitness (CRF), engage in lower amounts of habitual physical activity, and have at least one CVD risk factor [9]. This health and fitness profile increases their chance of suffering an acute cardiac event [3,6,10,11].
In addition to the health and fitness profile, the nature of the job can elevate cardiac risk. Most referees officiate parttime as an avocation; many work outdoors (in all types of weather conditions), at irregular hours, and in some cases, after having traveled long distances. Officials are a critical aspect of any athletic or sporting event as they are tasked with maintaining the standard of play. Depending on the sport, officiating may require standing, squatting, sprinting, or jogging for an extended period. American football officiating, in particular, can involve intense physical activity. For instance, the games are intermittent in nature and involve short-duration, repeated bouts of high-intensity exercise interspersed with periods of low-intensity activity between plays. American football officiating can be also psychologically stressful because split-second decisions and rulings in noisy--and sometimes hostile--environments must be made, while managing strong emotional responses from players, coaches, and spectators. These transient exposures to high physiological strain and psychological stress increase the risk for triggering a cardiac event five-fold when compared to periods of rest or lower strain [10,11].
Despite the aforementioned concerns, sports officials represent an understudied occupational group, and American football referees, especially those at the amateur (high school) level, have been largely excluded from sports officiating research [12]. Therefore, we sought to comprehensively summarize key demographic and health information in American high school football referees and to characterize the physical demands, kinematic activity, environmental stress, and physiological strain they experienced while officiating. We also sought to explore relationships between study variables of interest and indices of physiological strain. Although exploratory in nature, we hypothesized that American football referees would have suboptimal CVD risk and fitness profiles, and poorer CVD risk and fitness profiles would be associated with greater physiological strain while officiating.

Participants and study design
This cross-sectional study conformed to the standard set by the Declaration of Helsinki and was approved by the local university's institutional review board. High school football referees who were registered with their local high school athletic association were recruited to participate. All data were collected on-site in the field. Officials reported to an offsite meeting location approximately 2-3 h prior to the scheduled game start where they were briefed by research personnel about the study. Interested referees were consented before completing a self-administered questionnaire and baseline measures of standing height (model 213, Seca Ltd., Hamburg, Germany), body weight (model BWB-800, Tanita © , Arlington Heights, IL, USA), and pre-game hydration status. We used body mass index (BMI; calculated from height and weight) as our index of obesity: 18.5-24.9, 25-29.9, and ≥ 30 kg·m −2 , for normal, overweight, and obese, respectively [13].

Environmental conditions
Dry bulb (ambient temperature), wet bulb globe temperature (WBGT), relative humidity, and heat index were measured during warm-up and over the course of the game using a portable weather meter (Kestrel 5400 Heat Stress Tracker Pro with Compass and LiNK, Nielsen-Kellerman, Boothwyn, PA, USA). Kestrel weather meters are factory tested and individually calibrated and/or validated against standards traceable to the National Institute of Standards and Technology (temperature, WBGT, relative humidity, and heat index accuracy of ± 0.5 °C, ± 0.70 °C, ± 2.0%, and ± 4.0 °C, respectively) [14].

Hydration
Urine-specific gravity (U SG ) was measured using a digital handheld refractometer (Atago, PAL-10S, Tokyo, Japan) as a marker of hydration. U SG is a valid and reliable estimate of hydration, especially in field settings (r = 0.98, p < 0.001 with urine osmolality, a more accurate measure of urine concentration) [17][18][19]. Although urinary indices can be confounded by several factors [20], U SG cutoffs are routinely included with other biomarkers to monitor and assess hydration status in a variety of settings. U SG values ≤ 1.020 were considered an euhydrated state; values > 1.020 reflected a dehydrated state [21].
Peak T core , the highest value achieved during the game, served as our primary marker of thermal strain. All T core data were manually inspected prior to analyses; data recorded during half-time (when food and beverage intake were not controlled) and any values that appeared to be contaminated by fluid intake (< 36.0 °C) were removed [27].
Average HR response for the entire officiating period, expressed as a percentage of age-predicted maximal HR (HR max ), served as our primary marker of cardiovascular strain. Secondary indices of CV strain, e.g., peak HR (highest value achieved during the game) and the percentage of total game time spent in HR zones corresponding to low (zone 1: < 65% HR max ), moderate (zone 2: 65-85% HR max ), and high (zone 3: > 85% HR max ) levels of exertion, were also investigated. HR zones and age-predicted HR max (207 -[0.7 × age]) [33] were predefined by the Zephyr ™ Performance System.

Mechanical-to-physiological intensity ratio
This ratio represents a measure of efficiency between the cardiovascular and musculoskeletal systems used to quantify the potential mismatch between the demands of officiating and referee physical fitness [3]. A ratio of 1.0 suggests physiological responses match mechanical output; < 1.0 suggests low efficiency (i.e., low mechanical output but high physiological strain); > 1.0 suggests high efficiency (i.e., high mechanical output but low physiological strain).

Missing data
Consented officials were required to wear the Zephyr ™ Performance System but could opt out of T core monitoring and/ or the urine sample (one referee opted out of both procedures). Six referees experienced equipment issues, resulting in missing HR and GPS data (see supplemental material, Table S1, for details).

Data analysis
Data configuration for HR and GPS data were performed with OmniSense ™ software (Version 5.0, Zephyr Technology, Medtronic, Annapolis, MD, USA). Analyses were performed using SPSS v. 28.0 (SPSS Inc., Chicago, IL, USA). Descriptive statistics are summarized as M ± SD (unless otherwise stated). Normality was assessed using boxplots and the Shapiro-Wilk test. Limited by our small and unequal sample sizes across positions, we used one-way analysis of variance to explore mean differences (MD and 95% confidence interval [CI]) between head referees and umpires (coded = −1) versus all other positions (linesman and back-, side-, field-, and line judges; coded = 1). The magnitude of difference by position is quantified using Cohen's d (small = 0.2, moderate = 0.5, and large = 0.8) [34]. Bivariate correlations (r) were used to evaluate relationships among study variables and multiple regression models were used to identify unique predictors of cardiovascular (average HR), thermal (peak T core ), and overall physiological (training intensity) strain. Briefly, independent variables were selected based on the correlational analysis and previously published research involving workers in physically demanding occupations. Multiple regression models were then generated using the backward elimination method to identify the best combination of predictor variables. Models were checked for collinearity and fit. Although this is an exploratory study, an alpha level of 0.05 was used for all significance tests.

Referee characteristics
Eighty football referees were recruited for this study; 46 (58%) agreed to participate and were consented. Table 1 summarizes baseline characteristics for the total sample and by officiating role/position. On average, referees were middle-aged, non-Hispanic White/Caucasian (72%) men, with obesity (BMI ≥ 30 -kg·m −2 ), and very poor eCRF despite more than half (59%) identifying as regular exercisers [13]. Referees had ≥ 2 CVD risk factors (59% had ≥ 3); age, hypertension, and obesity were the most prevalent. Only 11% of the referees were considered normal weight (BMI of 18.5-24.9 kg·m −2 ). Across all positions, referees had ≥ 4 years of experience, officiated ≥ 1 games per week during the study, and about half officiated > 1 sport annually (≈55% were also baseball umpires). Nearly all referees (89%) reported working full-time jobs outside of officiating (5 were retired). Head referees/umpires were older, more obese (higher BMI), and had lower eCRF compared to all other positions (all p ≤ 0.04).

Environmental conditions
Games lasted ≈2.5 h and were played in the evening, under typical weather conditions for the Southeastern United States: 20.8 ± 6.7 °C, 18.9 ± 6.0 °C, and 78.2% ± 12.1% for dry bulb, WBGT, and relative humidity, respectively (see Table S4 for environmental condition summary and monthly climate normals). A game-by-game summary of WBGT values recorded at the start, end, and averaged over the game overlayed with WBGT exposure limit recommendations for working and exercising in the heat endorsed by various organizations [35,36] is depicted in Figure 1. Chronological (temporal) order of the games officiated was strongly related to environmental conditions (e.g., average WBGT r = −0.68 and average humidity r = 0.62, both p < 0.001) (see Figure 1 and Table S3). Higher average WBGT was inversely related to thermal strain (r = −0.33, p = 0.03), while higher average humidity was inversely related to overall physiological strain (training intensity: r = −0.32, p = 0.04) (Table S2).

Cardiovascular strain
Officiating elicited average and peak HRs of 124 ± 16 beats/ min and 159 ± 17 beats/min, corresponding to 72% and 92% HR max , respectively, with no differences across positions ( Table 2). Officials' average HR was in HR zone 2 (65-85% HR max ) for more than half of the total game time. Compared to all other positions, head referees/umpires spent less time in HR zone 1 (< 65% HR max ) (p = 0.005); time spent in other HR zones were not different ( Table 2). Greater cardiovascular strain (average HR response, expressed as a percentage of HR max ) was associated with BMI (r = 0.51, p < 0.001), number of CVD risk factors (r = 0.50, p = 0.002), not participating in regular exercise (r = −0.43, p = 0.01), and with games occurring earlier in the season (r = −0.35, p = 0.02) (Tables S2 and S3). Secondary indices of cardiovascular strain (e.g., time spent at > 85% HR max and < 65% HR max ) were also associated with BMI, number of CVD risk factors, habitual exercise, and chronological (temporal) order of the games officiated. Indices of cardiovascular strain were largely unrelated to kinematic activity outcomes. Average HR response was strongly and inversely correlated with mechanical-to-physiological intensity ratio (r = −0.80, p < 0.001), while time spent in the highest and lowest HR zones correlated with distance traveled by speed zone, mechanical intensity, and mechanical-to-physiological intensity ratio (Table S3).

Kinematic activity
Officials covered 5.2 ± 1.2 km at an average speed of 1.8 ± 0.3 km/h ( Table 2). In general, officials covered more distance at slower rather than faster speeds (speed zone 1: 3.1 ± 1.0 km [60% of total distance] versus speed zones 2 and 3: 2.1 ± 0.8 km [40% of total distance], p < 0.001, respectively), although this differed by position ( Table 2). Head referees/ umpires achieved lower peak speed (p = 0.009), covered more distance at slower speed (p = 0.004), and less distance at higher speed (p = 0.009) compared to all other positions ( Table 2). Kinematic activity outcomes correlated with several of the same individual factors--age, BMI, eCRF, and/or habitual exercise--but showed no association with indices of thermal and cardiovascular strain (Tables S2 and S3). Table 2. Markers of physiological strain, kinematic activity, and officiating performance measured pre-, during, and post-game for the total sample and by referee field position.

Total
Head referee/umpire Other field positions a Differences by position b
In addition to a strong and positive correlation with physiological intensity (r = 0.95, p < 0.001), training intensity was also related to many of the same correlates: chronological (temporal) order of games officiated (r = −0.39, p = 0.01), BMI (r = 0.36, p = 0.02), peak T core (r = 0.38, p = 0.01), and pre-game U SG (r = 0.32, p = 0.04). Training intensity did not differ by position and it was also largely unrelated to kinematic activity outcomes (including mechanical intensity) (Tables S2 and S3).

Multiple linear regression analysis
The final exploratory models for cardiovascular (average HR response), thermal (peak T core ), and overall physiological (training intensity) strain are summarized in Table 3 and Figure 2. Obesity, not participating in regular exercise, and officiating at lower mechanical intensity predicted greater cardiovascular strain in our sample (Figure 2(a)). White/Caucasian race/ethnicity, younger age, and obesity predicted greater thermal strain among officials (Figure 2(b)). Both physiological and mechanical intensity predicted training intensity (overall physiological strain), accounting for 100% of the observed variance. However, physiological (β = 0.993), not mechanical (β = 0.318) intensity was a much stronger predictor of training intensity, substantiating the 'mismatch' or inefficiency observed in our sample.

Discussion
The primary finding of this study is that American high school football referees experienced similar levels of cardiovascular and thermal strain, regardless of their position. One potential explanation for this is the inefficiency we observed (mismatch between the cardiovascular and musculoskeletal systems). Additionally, we observed that individual rather than officiating-specific or environmental factors were primary predictors of increased thermal, cardiovascular, and physiological strain. In this context, our exploratory hypothesis was correct; however, we thought this relationship would be mediated, at least in part, by kinematic activity and environmental factors. The CVD risk profile of our sample (middle-aged, male, overweight/obese, and low eCRF) is consistent with reports of American football officials at the collegiate level [37][38][39]. The lack of research involving amateur and high school sports officials makes comparing our findings to other sports difficult; however, based on research involving amateur and professional sports officials, this profile is drastically different (e.g., soccer, rugby, and basketball officials tend to be younger, have normal body weight, and higher CRF) [40][41][42]. Given the relationship among unaccustomed, acute bouts of vigorous-intensity exercise, underlying CVD risk factors, and sudden cardiac events [3][4][5][6]10,11], high school football officials appear to be at increased risk for triggering an acute cardiac event.

Cardiovascular strain
Not engaging in regular exercise, higher BMI, and lower mechanical intensity were predictors of higher cardiovascular strain. Our findings pertaining to physical inactivity and obesity are consistent with existing research involving other physically demanding occupations [2,3,10,11,43], including sports officials [37][38][39]. Additionally, BMI was positively and strongly correlated with several indices of cardiovascular and physiological strain, suggesting that the heightened physiological responses to officiating duties were partially mediated by obesity, even among duties of minimal or low physical exertion. Indeed, our multiple regression model revealed that officials who achieved lower mechanical intensity during the game experienced higher cardiovascular strain. While unexpected, we posit that differences in kinematic activity and CVD risk profiles between positions and the officials who occupy them may explain this negative relationship. For example, head referees/umpires were older, had higher BMI, lower eCRF, covered more distance at slower speed, less distance at faster speed, and achieved lower mechanical intensity compared to all other positions (Table 3). It seems that mechanical intensity achieved during the game period is related to the health and fitness of the official, i.e., officials with more favorable CVD risk and fitness profiles can achieve and tolerate greater mechanical intensity with less strain imposed on the cardiovascular system. Although officials with these suboptimal profiles tend to occupy positions requiring lower mechanical intensity, they also seem to have a lower tolerance for higher levels of physical exertion.

Thermal strain
We found that White/Caucasian race/ethnicity, younger age, and higher BMI were unique predictors of higher thermal strain. Although we observed significant differences by race/ ethnicity, we are hesitant to over-interpret our results given the lack of representation in our sample (72% White/ Caucasian). Human thermoregulation is considerably adaptable, and while racial/ethnic differences have been suggested, newer, carefully designed experimental studies provide strong evidence to the contrary [44]. The racial/ethnic differences we observed are likely attributable to differences in kinematic activity while officiating. For instance, Black/African American officials covered less total distance (−0.9 km, 95% CI: −1.7, −0.1) and less distance at faster speeds (−0.4 km, 95% CI: −0.8, −0.1) compared to White/Caucasian officials, likely resulting in lower metabolic heat production and heat storage that resulted in lower T core . We offer a similar explanation for the age-related differences observed, as declines in thermoregulatory function with aging are well established [45]. Indeed, increasing age was associated with lower peak and average speed (r= −0.38, p = 0.01), and lower mechanical intensity (r= −0.34, p = 0.03) (Table S2). Older officials also occupied the more senior, less physically (mechanically) demanding positions (i.e., head referee/umpire).
Our findings concerning excess body weight and thermal strain are consistent with the well-documented effects in the literature. Compared to leaner individuals, those with overweight and obesity must work at higher metabolic intensity to achieve the same absolute workload, resulting in greater metabolic heat production, heat storage, and increased T core , independent of environmental conditions [2,36,45]. Reduced body surface area-to-mass ratio among obese versus lean individuals and obesity-induced thermoregulatory impairments (e.g., reduced cutaneous vasodilation resulting from altered sympathetic control in the vasculature) also contribute to greater heat storage [2,36,45].

Physiological, mechanical, and training intensities
Consistent with our other bivariate and multiple regression analyses, our third model revealed that physiological--not mechanical--intensity was the predominant source of strain in our sample. Indeed, the average mechanical-to-physiological intensity ratio pointed to a low level of efficiency between officials' cardiovascular and musculoskeletal systems, suggesting a mismatch between their physical fitness and the demands of officiating. Only five officials in our sample achieved a ratio of 1.0 or higher (1.4 ± 0.4 vs. 0.5 ± 0.2); 88% of the sample experienced higher physiological intensity despite similar or lower mechanical (kinematic) output. In fact, only the percentage of total distance covered in speed zone 1 differed between officials with high versus low ratios (−12.2%, p = 0.03). Conversely, officials with ratios < 1.0 were older, had more experience, achieved higher peak and average HR responses (all p ≤ 0.02), and tended to have lower eCRF and higher peak T core (both p ≤ 0.09). Collectively, the results suggest that the overall strain officials experienced during the game was largely determined by the cardiovascular and not the musculoskeletal system, which is likely linked to their poor CVD risk and fitness profiles.

Environmental conditions
Surprisingly, environmental conditions were not associated with indices of physiological strain. Hot and humid weather conditions during August and September in the Southeastern United States can be particularly taxing because they impede evaporative heat loss--the primary mechanism of heat dissipation during exercise. Hence, we expected officials to experience greater heat stress, fluid loss, and physiological strain during games with more extreme environmental conditions. WBGT and humidity both correlated with chronological (temporal) order of the games officiated, suggesting that greater environmental heat stress occurred earlier rather than later in the season. These relationships may also reflect improved game-specific fitness or thermoregulatory function resulting from repeated bouts of officiating in the heat (i.e., acclimatizing naturally 'on the job') [2], or some combination of factors. Even so, only a couple games exceeded WBGT exposure limits for performing continuous heavy physical work in the heat [36]; no games exceeded the limits for continuous moderate physical work [36], nor were they considered 'very high risk' for exercising in the heat [35,36] (Figure 1). The lack of findings related to environmental conditions and indices of physiological strain led us to speculate that environmental conditions, when averaged over the course of the season, were not 'extreme' enough to elicit significant physiological strain or decrements in performance, especially in our sample of referees who were long-term residents of this geographical region. Hence, they were likely well acclimatized, having developed cardiovascular and thermoregulatory adaptations to sufficiently cope with the challenges of this climate [2,36,44,45].

Limitations and strengths
A limitation of this study was that several baseline measures were collected in the field (outside of the laboratory) using self-report questionnaires or non-exercise methods. Devicemeasured outcomes generally exhibit the highest levels of validity and reliability; however, adequate levels have been reported for self-report methods and non-exercise prediction equations [15,16]. The approach we used allowed us to study American football referees' performance in their natural environment within the context of a 'live' contest. Another limitation is that we did not measure psychological stress or officiating performance (e.g., degree of accuracy, position in relation to the play)--important factors to consider when attempting to account for potential stressors placed on sports officials and the impact they have on occupation-specific tasks [12]. We cannot determine whether officials occupying the more senior but less physically demanding positions (i.e., head referee/umpire) experienced heightened psychological stress, and if so, to what extent it impacted physiological strain. This relationship warrants further investigation.

Conclusion
The results of this study demonstrate that high school American football referees experience elevated levels of physiological strain while officiating, with individual factors--not environmental conditions or kinematic activity--modulating the magnitude of strain. Our study offers some of the first data characterizing the health, kinematic activity, and physiological strain of this understudied group. This information can be used to better inform occupation-specific recommendations regarding medical clearance, appropriate fitness testing procedures, and training regimes to cultivate a safer work environment for officials, with a special emphasis on reducing obesity and improving cardiovascular health.

Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request (hvmacdonald@ua.edu).

Disclosure statement
No potential conflict of interest was reported by the author(s).