The short- and long-term health effects of playing football at an elite level have been a source of significant speculation and discussion in the medical and lay communities (1,2). A 1988 article cited reports suggesting that former National Football League (NFL) players often die younger than the general population (3). In 1994, the National Institute for Occupational Safety and Health (NIOSH) investigated concern expressed by representatives of the National Football League Players Association that NFL players were at increased risk for premature death (4). NIOSH’s health hazard evaluation found that professional football players who played between 1959 and 1988 with five or more credited seasons had a 46% lower overall mortality rate and a 13% (though not statistically significant) lower cardiovascular mortality rate compared with the reference population, specifically US men with a similar age and race distribution. However, linemen had a 50% greater risk of dying from cardiovascular disease (CVD) than did the general population, and players with the largest body mass index (BMI; ≥32 kg·m−2) had a sixfold increase in risk compared with those in the lowest BMI category (<28 kg·m−2).
In their subsequent study that included 16 additional years of follow-up, NIOSH confirmed a decreased overall player mortality (standardized mortality ratio (SMR), 0.53; 95% confidence interval (CI), 0.48–0.59) and decreased cardiovascular mortality (SMR, 0.68; 95% CI, 0.56–0.81) compared with the general male population. Former players with a playing-time BMI of >30 kg·m−2 had twice the risk of CVD mortality compared with other players, and African American players and defensive linemen had higher CVD mortality compared with other players even after adjusting for playing-time BMI (5). The authors suggested that large former NFL players (playing-time BMI ≥30 kg·m−2) are at increased risk for CVD risk factors and mortality, as are their counterparts in the general population.
In contrast to NFL players of prior generations, current players participate in organized team activities, minicamps, and off-season conditioning for several months before preseason camp and the competitive season (6–9). In addition, players engage in physical training in the remaining months of the year to gain or maintain their strength and conditioning (10).
Parallel to the increased focus on conditioning, recent studies have identified an increase in the weight of NFL linemen, tight ends, and linebackers over time (11,12). In response to these observations, NFL Commissioner Paul Tagliabue convened the NFL Subcommittee on Cardiovascular Health in 2005 to explore the health risks of increasing size in NFL players. The Subcommittee undertook a cross-sectional study to examine cardiac risk factors in active players (13). This study found active players to have elevated blood pressure, comparable lipids, lower fasting glucose, and dramatically less cigarette smoking compared with general population controls.
NFL players represent a range of body habitus across player positions (12,14). Furthermore, positions are characterized by varying physiological demands, each with a corresponding resistance and aerobic training regimen. After his playing career, a player is less likely to find himself in an environment that provides structured physical activity and access to nutritional and medical supervision. Thus, there are a multitude of factors that may affect overall and cardiovascular health after the career.
To investigate long-term effects of these influences on mortality in the former NFL player population, the Subcommittee sought to identify the most common causes of death and to explore the risk of overall and cardiovascular mortality among a recent cohort of former NFL players. In addition, this study assessed whether the findings are consistent with those from the earlier NFL cohort in terms of race, BMI, and position group.
Study population and data collection
The study cohort included 9778 NFL players with at least 1 yr in the NFL whose last season was between 1986 and 2012. Exclusions from the initial cohort of 11,202 players consisted of 617 players who played during the 1987 strike year and were considered “pure” replacement players by the League pension office, 29 players missing dates of birth, 1 player with multiple Social Security numbers, 58 players who reactivated after the 2012 season, and 719 players for whom we could not identify race. Of the 719 players with undetermined race, there was no photo available for 705 (98%), the date of the photo or team indicated in the photo did not match the player demographic/anthropometric data provided by the League for 9 players (1.2%), and a photo was found, but race was unable to be determined for 5 players (0.6%).
Player records were obtained from the NFL Player Information Office and pension fund administrator and then matched to the National Death Index (NDI) to determine vital status, date of death, and cause of death. The Player Information Office provided data including playing-time height and weight, position, and college. The pension fund database included information on each player’s name, Social Security number, date of birth, first season played, last season played, retirement date, years in the NFL, last team played for, and last known address. The investigators created a unique study identifier for each player’s record that linked pension fund data to NDI data; otherwise, missing player information was obtained from media guides or other publicly available sources (e.g., NFL.com).
Race was classified by one person (N.F.K.) using available photos on the Associated Press Images Web site, the Getty Images Web site, or annual team media guides provided by the Pro Football Hall of Fame archives department if not otherwise found. Only photos with the player’s face clearly visible were used. Black and white photos were only used when a color photo was unavailable and if the player’s race could be readily determined. Photo captions were checked for consistency with players’ football career identifiers (e.g., last team played for, last season played, player position, and college) provided by the NFL pension fund database and the NFL office of player information. Discrepancies between player identifiers provided by League sources and player information provided in AP Images photo captions were cross-referenced with NFL.com or Pro Football Reference (pro-football-reference.com). Missing demographic and anthropometric data were collected using NFL.com, Pro Football Reference, or the annual NFL media guides.
Player positions were grouped to reflect a combination of aerobic and anaerobic game demands, training regimens, and physical size required to play at the professional level. The group categories were specialists (quarterbacks, kickers, punters), speed (wide receivers, defensive backs, tight ends, running backs, linebackers), and power (offensive and defensive linemen). Special teams players were categorized according to their primary offensive or defensive position.
Vital status assessment
Vital status was determined primarily by matching records to the NDI. A cohort member was considered alive through 2014 if no death record was located. Causes of death provided by NDI were coded according to the NIOSH 119 cause of death categories (15).
In addition to the 218 players determined to be deceased by NDI, 9 additional players were identified as deceased by the pension fund data. The public accounts of these 9 deaths yielded reports that enabled the authors to categorize each case by major cause of death.
Each cohort member accumulated one person-year at risk (PYAR) for each year of life between the date he became eligible for the study (July 1 of his first season played) and the date of death (for deceased cohort members) or the ending date of the study (December 31, 2014, for cohort members not identified in NDI or pension fund categorized as deceased). The PYAR was stratified into 5-yr intervals by age and calendar time and then multiplied by the appropriate US sex, race, and cause-specific mortality rates to calculate the expected number of deaths for that stratum. The resulting expected numbers were summed across strata to obtain cause-specific and total expected number of deaths.
The ratio of observed to expected number of deaths was expressed as the SMR. Ninety-five percent CI values were computed for the SMR assuming a Poisson distribution for observed deaths (16). As with previous studies (5,17,18), mortality risk was analyzed using the NIOSH life table analysis system (LTAS.NET version 4.0) (15,19). Standard US male mortality rates for the period 1960–2007 (with projected rates through 2014) for 119 cause-of-death categories have been created by NIOSH and were used for LTAS analysis (20,21). Standardized rate ratios (SRR) were used to compare strata within player characteristics.
Custom rate files that combine standard categories were created to analyze deaths attributable to CVD, including diseases of heart and circulatory system. To further explore the variability in overall and heart disease mortality within the NFL cohort, Cox proportional hazards models were created to determine the adjusted risk for overall mortality and CVD–related mortality in relation to specific risk factors such as position group, race, playing-time BMI, and years in the NFL. This study was reviewed by the institutional review board of the MedStar Health Research Institute (Baltimore, MD).
The 9778 players composing the cohort contributed 184,016 PYAR with a median contribution of 18.5 yr (range, 0.4–45.5 yr) per player (Table 1). The median age of those alive as of December 31, 2014, was 41 yr (range, 24–68 yr), whereas the median age at death was 38 yr (range, 23–61 yr).
Two percent (n = 227) of players were deceased, with the most common major causes of death being diseases of heart (n = 47; 21%), violence (n = 39; 17%), and transportation injuries (n = 34; 15%) (Table 2). Among the most common major causes of death, violence and transportation injury were more prevalent before age 35 yr, particularly among nonwhites (Fig. 1). Deaths from CVD and cancer were more prevalent among those 35- to 54-yr-olds.
Compared with their US general population male counterparts, former NFL players had a significantly lower overall mortality rate (SMR, 0.46; 95% CI, 0.40–0.52) as well as the five most common causes of death among those younger than 65 yr: diseases of heart (SMR, 0.68; 95% CI, 0.50–0.90), violence (SMR, 0.38; 95% CI, 0.27–0.52), transportation injuries (SMR, 0.65; 95% CI, 0.45–0.91), other injury (SMR, 0.53; 95% CI, 0.34–0.79), and cancer (SMR, 0.41; 95% CI, 0.26–0.62; Table 2). Figure 2 depicts the distribution of ages at death and playing-time BMI for deaths attributable to ischemic heart disease. Contrary to expectation, we did not find values concentrated in the upper ranges of age or BMI.
Overall mortality was significantly lower than the general population across all race, position group, and years in the NFL categories (Table 3). Mortality from CVD was also significantly lower for whites and nonwhites, speed and specialist position groups, and players with 5 or more years in the NFL. Within the player cohort, overall mortality was significantly higher among nonwhites relative to whites (SRR, 1.63; 95% CI, 1.21–2.21) and power position relative to speed position (SRR, 2.16; 95% CI, 1.62–2.86). Players with BMI of 30–34.9 kg·m−2 (SRR, 1.51; 95% CI, 1.10–2.06) and ≥35 kg·m−2 (SRR, 3.84; 95% CI, 2.66–5.54) had higher mortality rates than did players with a BMI of <30 kg·m−2. Players with 5 or more years in the NFL had a mortality rate comparable with players with 1–4 yr in the NFL (SRR, 0.96; 95% CI, 0.70–1.31).
CVD mortality was higher for nonwhites (SRR, 3.00; 95% CI, 1.54–5.87) and for power position players (SRR, 2.99; 95% CI, 1.67–5.37). Players with BMI of 30–34.9 kg·m−2 (SRR, 3.90; 95% CI, 2.00–7.61) and ≥35 kg·m−2 (SRR, 9.03; 95% CI, 4.20–19.41) had higher CVD mortality rates than did players with a BMI of <30 kg·m−2.
The variability in overall and cardiovascular mortality by BMI, race, and years in the NFL is shown in Figures 3A and B, respectively. The proportional hazards model for overall mortality (χ2 = 109.6, df = 8, P < 0.0001) demonstrates greater risk among nonwhites (P = 0.0002), those with higher BMI (P < 0.0001), and players with fewer years in the NFL (P = 0.0012); the proportional hazards model for cardiovascular mortality (χ2 = 72.2, df = 8, P < 0.0001) also demonstrates greater risk among nonwhites (P = 0.0006), those with higher BMI (P < 0.0001), and players with fewer years in the NFL (P = 0.026).
This study was undertaken to explore the risk of overall and cardiovascular mortality among a recent cohort of former NFL players. Despite media reports that NFL players have dramatically shorter life spans (22,23), our findings indicate that players have significantly reduced overall and cause-specific mortality rates relative to the general population. However, those players with the highest playing-time BMI demonstrated elevated overall and cardiovascular mortality risk within the cohort.
Overall, 2% of former NFL players from the 1986 to 2012 era had died by 2014, primarily from heart disease, violence, transportation injuries, and cancer. Consistent with the previous study of NFL players from an earlier era (1959–1988) (5), heart disease was the most common cause of death in this NFL cohort. Heart disease–related deaths were significantly lower than expected from the general population (SMR, 0.68), although this was largely driven by the extremely low number of deaths among players with the lowest playing-time BMI (<30 kg·m−2). Players with a BMI of ≥35 kg·m−2 were the only stratum with a significantly higher risk of CVD mortality (SMR, 2.2), consistent with the elevated risk among those with similar BMI in the general population (24,25).
These initial findings for this study cohort suggest two primary narratives in which the favorable overall and cardiovascular mortality results of the cohort are dampened by the findings in the largest players. The potential negative influences on cardiovascular health in this subpopulation of largest players are several. Hypertension rates were highest among linemen in the Subcommittee’s study of active players and may persist over time (13). Linemen typically have the highest percentage of body fat among all player positions (12), and persistence of this characteristic is associated with health risks (25). The largest players consistently perform the most vigorous resistance training programs during their playing career, and concern about the effect of this training on vascular stiffness of major vessels has been raised (26). Recent data among college linemen demonstrate potentially maladaptive myocardial remodeling over the course of a single season, which raises concern about potential long-term effects on cardiac function (27). The authors remain concerned that a subpopulation of the former players may emerge as being at high risk for premature aging of the cardiovascular system.
One of the primary findings is a significant racial difference, with nonwhites having higher mortality risk, which is reflective of the general population (28). This contrasts with the lack of racial difference in cardiovascular risk factors found among active players (13,29). Further study should address racial differences in health behaviors, family history, and comorbidities within this population as they leave the structured lifestyle of the NFL.
Among deaths attributable to diseases of the heart, ischemic heart disease is the most common underlying cause of heart disease in this young population of former elite football players, accounting for 22 of 47 cases attributed to diseases of the heart. When examining the distribution of these deaths across BMI and age at death, there was no discernible clustering of deaths among the higher BMI or age values, as might be expected for ischemic heart disease.
Compared with the earlier NFL cohort study, our cohort was nearly three times as large and contributed an additional 80,000 PYAR. Our findings were largely consistent in terms of most common cause of death (CVD) and lower overall and cause-specific mortality compared with the general population of US men, and higher mortality risk among nonwhites, power position players (e.g., linemen), and those in the largest BMI category. The consistency of these findings exists despite differences in cohort definition (5+-yr veterans vs 1+-yr veterans), data collection approaches, increased size among linemen in more recent years (11,12), League tenure as measured in years in the NFL rather than credited seasons, and near–year-round training techniques among players in recent years (6,8–10).
Our findings are consistent with the broader literature that evaluated longevity among elite athletes compared with the general population. Comparing the mortality of French participants in the Tour de France (1947–2012) to French men, Marijon et al. (30) found a lower SMR for all causes (0.59; 95% CI, 0.51–0.68) and cardiovascular death (0.67; 95% CI, 0.50–0.88). In a recent meta-analysis that included 10 studies (Baron et al.’s previous NFL cohort, Marijon et al.’s cyclist cohort, and eight additional studies), Garatachea et al. (31) identified a lower overall mortality (SMR, 0.67; 95% CI, 0.55–0.81) and mortality for CVD (SMR, 0.73; 95% CI, 0.65–0.82) among elite athletes.
There are several possible explanations for the decreased overall and cause-specific mortality in the current NFL cohort. Reduced mortality rates may be partly attributable to the Healthy Worker Effect (HWE), whereby active workers healthy enough to have been employed usually exhibit lower overall mortality rates compared with the general population because the severely ill and chronically disabled are typically excluded from employment (32). The likelihood of the HWE on our findings is substantial, given the optimal health required among players to first join an NFL team. Although we have information on players through their often relatively brief playing careers (median years in NFL, 4 yr), we do not have information on their employment activity after football, such as whether they pursued other careers, were unemployed, or were not in the labor force.
Although we did not have information on subjects’ smoking status or other health behaviors, findings from a large cohort of active NFL players collected in 2007 indicated that smoking was virtually nonexistent (0.2%) among active players at that time (13). A University of Michigan report indicated that smoking among former NFL players was less than 8% compared with greater than 20% of men in the general population (33,34). These findings suggest that smoking rates among former NFL players are likely to be dramatically lower than those of the general population. The effect of dramatically reduced smoking prevalence among NFL players could contribute to a decreased risk for heart disease, cancer, and other major causes of morbidity and mortality.
Other possible explanations include the high levels of physical fitness among former NFL players in their young adulthood, relatively high levels of health insurance (34), education attained, and, though potentially brief, high income levels as professional athletes. The effect of income on mortality was recently associated with a difference of 15 yr in life expectancy between American men in the top 1% in income compared with the poorest 1% (35).
Despite the risk of injury while playing in the NFL (36), the number of years in the League was not found to have a significant effect on overall or CVD mortality risk. The presumed benefits related to playing in the NFL (e.g., high levels of physical activity, nutritional guidance, access to health care, economic advantage) may outweigh the long-term musculoskeletal effects of playing professional football. Alternatively, exposures to risk factors before and after a player’s career may have a much greater effect on mortality risk than those associated with their tenure in the NFL.
This study is limited by the lack of information on specific behavioral or clinical factors that may be related to any increased or decreased risks. Some risk factors associated with CVD, including smoking, high blood pressure, decreased HDL cholesterol, and glucose intolerance (prediabetes and diabetes), were not assessed. Information on BMI was limited to heights and weights reported during their playing career and are subject to variability in measurement technique and accuracy. Given the large size and high levels of fitness that many players maintained during their career, the likelihood that players may have gained or lost significant amounts of weight once leaving the game is high, but these data were not obtained. Also, the use of photographs from media guides and other official sources to assess players’ race may have resulted in misclassification.
Although the NDI-match program used by NIOSH is highly effective in obtaining all available deaths from the NDI, studies have consistently found that 3%–5% of known deaths cannot be obtained from the NDI for various reasons, usually because of erroneous information provided by the informants on death certificates (37). There is also the potential for misclassification bias among causes of death.
Further study is needed to explore the contributions of race, health risk behaviors, diet, weight, general health status, and use of preventive health care services to the overall and cardiovascular mortality among active and former NFL players (38,39). We will continue to follow up this cohort over time to explore whether the elevated mortality risk among nonwhite players and players with a BMI of >35 kg·m−2 persists or worsens with additional follow-up. In addition to players’ longevity, study of their quality of life should be initiated that addresses possible impairment and disability resulting from their careers in professional football.
Given that large size is a primary characteristic for some players in the Power position group to be signed by an NFL team, an active strategy may be needed to shift from a resistance training regimen to one that emphasizes aerobic activity and modification of caloric intake to reduce long-term health consequences after their playing career. A number of recent linemen who played at more than 300 lb have demonstrated the ability to lose significant amounts of weight and may serve as role models for other players as well as the public (40).
In summary, findings for this recent NFL cohort showed reduced overall and cardiovascular mortality risk compared with the general US male population, attributable in part to a HWE and less smoking. However, players in the highest playing-time BMI strata exhibited significantly higher risk of cardiovascular mortality.
The authors acknowledge funding support for this research from the National Football League. They thank the NFL pension fund administrator, the NFL Player Information Office, the National Death Index, and Jon Kendle at the Pro Football Hall of Fame for contributions to data collection efforts. The authors thank the other members of the NFL Subcommittee on Cardiovascular Health (Dr. Elliot Pellman, Dr. Thom Mayer, Dr. Lon Castle, Dr. Robert Heyer, Dr. Patrick Strollo, Dr. Peter Wilson, Dr. Henry Black, and Jim Whalen, ATC) and J. Thomas Goglia, who contributed to the development and results of the study. The authors declare no conflicts of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.
1. Jenkins S, Maese R, Clement S. Do no harm: NFL players endure a lifetime of hurt. Washington Post [Internet]
. 2013 [cited 25 April 2017]. Washington (DC). Available from: http://www.washingtonpost.com/sf/feature/wp/2013/05/16/do-no-harm-retired-nfl-players-endure-a-lifetime-of-hurt/?utm:term=.4122d1d5ef50
2. Olson D, Sikka RS, Labounty A, Christensen T. Injuries in professional football: current concepts. Curr Sports Med Rep
3. Utterback B. NFL life signals early death. Chicago Tribute [Internet]
. 1988 [cited 7 Nov 2016]. Chicago. Available from: http://articles.chicagotribune.com/1988-03-04/sports/8804040445_1_life-expectancy-heart-attack-average-lifespan
4. Baron S, Rinsky R. National Football League Players Mortality Study
. Atlanta (GA): CDC/NIOSH; 1994 Health Hazard Evaluation Report.
5. Baron SL, Hein MJ, Lehman E, Gersic CM. Body mass index
, playing position, race
, and the cardiovascular mortality
of retired professional football players
. Am J Cardiol
6. Anding R, Riley D, Wright R. Strength and Conditioning Program (Players Manual)
. Houston Texans [Internet] [cited 7 Nov 2016]. Available from: http://assets.houstontexans.com/assets/fanzone/Texans%20strength%20manual.pdf
7. Bowen M. For NFL players, there is no summer vacation. Bleacher Report [Internet]
. 2014 [cited 7 Nov 2016]. Available from: http://bleacherreport.com/articles/2099803-for-nfl-players-there-is-no-summer-vacation
8. Collective Bargaining Agreement. August 4, 2011. National Football League. Available from: https://nfllabor.files.wordpress.com/2010/01/collective-bargaining-agreement-2011-2020.pdf
9. Gabriel G. How training camps have changed. National Football Post [Internet]
. 2015 [cited 1 Dec 2016]. Available from: http://www.nationalfootballpost.com/how-training-camps-have-changed/
10. Gergen J. In Today’s NFL, tipping the scales at 300 pounds is not big deal. In: Newsday
. 2001 [Internet]. New York. Available from: http://www.newsday.com/sports/sunday-special-heavy-duty-in-today-s-nfl-tipping-the-scales-at-300-pounds-is-not-big-deal-1.795292
11. Anzell AR, Potteiger JA, Kraemer WJ, Otieno S. Changes in height, body weight, and body composition in American football players from 1942 to 2011. J Strength Cond Res
12. Kraemer WJ, Torine JC, Silvestre R, et al. Body size and composition of National Football League players. J Strength Cond Res
13. Tucker AM, Vogel RA, Lincoln AE, et al. Prevalence of cardiovascular disease risk factors among National Football League players. JAMA
14. Allen TW, Vogel RA, Lincoln AE, Dunn RE, Tucker AM. Body size, body composition and cardiovascular disease risk factors in NFL players. Phys Sportsmed
15. Schubauer-Berigan MK, Hein MJ, Raudabaugh WM, et al. Update of the NIOSH Life Table Analysis System: a person-years analysis program for the Windows computing environment. Am J Ind Med
16. Rothman K, Boice J. Epidemiologic Analysis with a Programmable Calculator
. Washington (DC): U.S. Government Printing Office; 1979.
17. Lehman EJ, Hein MJ, Baron SL. Neurodegenerative causes of death among retired National Football League players. Neurology
18. Lehman EJ, Hein MJ, Gersic CM. Suicide mortality
among retired National Football League players who played 5 or more seasons. Am J Sports Med
19. Steenland K, Spaeth S, Cassinelli R, Laber P, Chang L, Koch K. NIOSH life table program for personal computers. Am J Ind Med
20. Robinson CF, Schnorr TM, Cassinelli RT, et al. Tenth revision U.S. mortality
rates for use with the NIOSH Life Table Analysis System. J Occup Environ Med
21. Waxweiler RJ, Beaumont JJ, Henry JA, et al. A modified life-table analysis system for cohort studies. J Occup Med
22. Hargrove T. Heavy NFL players twice as likely to die before 50. ESPN [Internet]
. 2006 [cited 31 Jan 2006]. Available from: http://www.espn.com/nfl/news/story?id=2313476
24. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality
with overweight and obesity using standard body mass index
categories: a systematic review and meta-analysis. JAMA
25. Jiang J, Ahn J, Huang WY, Hayes RB. Association of obesity with cardiovascular disease mortality
in the PLCO trial. Prev Med
26. Kim JH, Sher S, Wang F. Impact of American-style football
participation on vascular function. Am J Cardiol
27. Lin J, Wang F, Weiner RB, et al. Blood pressure and LV remodeling among American-style football
players. JACC Cardiovasc Imaging
28. Deaton A, Lubotsky D. Mortality
, inequality and race
in American cities and states. Soc Sci Med
29. Tucker AM, Lincoln AE, Vogel RA, et al. Lack of blood pressure difference by race
in professional American football players. J Am Soc Hypertens
30. Marijon E, Tafflet M, Antero-Jacquemin J, et al. Mortality
of French participants in the Tour de France (1947–2012). Eur Heart J
31. Garatachea N, Santos-Lozano A, Sanchis-Gomar F, et al. Elite athletes live longer than the general population: a meta-analysis. Mayo Clin Proc
32. McMichael AJ. Standardized mortality
ratios and the “healthy worker effect”: scratching beneath the surface. J Occup Med
33. Dwyer-Lindgren L, Mokdad AH, Srebotnjak T, et al. Cigarette smoking prevalence in US counties: 1996–2012. Popul Health Metr
34. Weir DR, Jackson JS, Sonnega A. National Football League Player Care Foundation Study of Retired NFL Players. Ann Arbor (MI): University of Michigan Institute for Social Research; 2009.
35. Chetty R, Stepner M, Abraham S, et al. The association between income and life expectancy in the United States, 2001–2014. JAMA
36. Lawrence DW, Hutchison MG, Comper P. Descriptive epidemiology of musculoskeletal injuries and concussions in the National Football League, 2012–2014. Orthop J Sports Med
37. Kraut A, Chan E, Landrigan PJ. The costs of searching for deaths: national death index vs social security administration. Am J Public Health
38. Active Projects. The football players health study at Harvard University. [Internet] 2015 [cited 8 Nov 2016]. Available from: https://footballplayershealth.harvard.edu/research/active-projects/
39. Lezar K. NFL players union and Harvard team up on landmark study of football injuries and illness. Boston Globe [Internet]
. 2013 [cited 7 Nov 2016]. Boston. Available from: http://archive.boston.com/lifestyle/health/2013/01/29/nfl-players-union-and-harvard-team-landmark-study-football-injuries-and-illness/aCGnf96h7ptWX2Lnp5MIiP/story.html
40. Garber G. Weight and ex-linemen: when losing wins. ESPN [Internet]
. 2013 [cited 25 April 2017]. Available from: http://www.espn.com/nfl/story/_/id/10042606/losing-weight-winning-former-nfl-offensive-linemen