In this cohort study, male football players who had played in the Swedish top division had a significantly increased risk of neurodegenerative disease compared with population controls. The risk increase was observed for Alzheimer's disease and other dementias but not for other types of neurodegenerative disease, and among outfield players, but not among goalkeepers. Our study expands on the data that can be used to assess and manage risks in the sport.
Of 7386 football players who had played at least one game in the top Swedish division between Aug 1, 1924, and Dec 31, 2019, 182 players were excluded for an unretrievable personal identity number, and 417 were excluded due to their number not being identified in the Total Population Register. After a further exclusion of 780 players and 11 627 controls who were born outside of Sweden and who had immigrated to the country after age 15 years, 6007 football players (510 goalkeepers) were included in the study population along with 56 168 matched controls. During follow-up to Dec 31, 2020, 537 (8·9%) of 6007 football players and 3485 (6·2%) of 56 168 controls were diagnosed with neurodegenerative disease. The risk of neurodegenerative disease was higher among football players than controls (hazard ratio [HR] 1·46 [95% CI 1·33–1·60]). Alzheimer's disease and other dementias were more common among football players than controls (HR 1·62 [95% CI 1·47–1·78]), significant group differences were not observed for motor neuron disease (HR 1·27 [0·73–2·22]), and Parkinson's disease was less common among football players (HR 0·68 [0·52–0·89]). The risk of neurodegenerative disease was higher for outfield players than controls (HR 1·50 [95% CI 1·36–1·65]) but not for goalkeepers versus controls (HR 1·07 [0·78–1·47]), and outfield players had a higher risk of neurodegenerative disease than did goalkeepers (HR 1·43 [1·03–1·99]). All-cause mortality was slightly lower among football players than controls (HR 0·95 [95% CI 0·91–0·99]).
In this cohort study, we identified all male football players (amateurs and professionals) who had played at least one game in Allsvenskan from Aug 1, 1924 to Dec 31, 2019 and excluded players whose personal identity number could not be retrieved or be identified in the Total Population Register, and those who were not born in Sweden and who had immigrated to the country after age 15 years. Football players were matched with up to ten controls from the general population according to sex, age, and region of residence. We used nationwide registers to compare the risk of neurodegenerative disease (diagnoses recorded in death certificates, during hospital admissions and outpatient visits, or use of prescription drugs for dementia) among football players versus controls. We also assessed each type of neurodegenerative disease (Alzheimer's disease and other dementias, motor neuron disease, and Parkinson's disease) separately, and compared the risk of neurodegenerative disease among outfield players versus goalkeepers.
Football (soccer) players might be at increased risk of neurodegenerative disease, which has led to questions regarding the safety of the sport and recent measures introduced by football associations to reduce heading of the ball. We aimed to assess the risk of neurodegenerative disease among male football players in the Swedish top division Allsvenskan, compared with matched controls.
Using nationwide health and administrative registers, our cohort study aimed to assess the risk of neurodegenerative disease among male football players in the Swedish top division compared with matched controls from the general population.
The heightened concern regarding the safety of football has impacted practice guidelines for millions of players worldwide and brought the potential risk of neurodegenerative disease associated with the sport into broad public attention. Although it is established that physical activity and participation in sports have important health benefits, further studies of neurodegenerative disease among football players are needed to inform the management of risks in the sport.
In 2019, a cohort study showed that former professional male football players in Scotland had more than 3·5 times higher risk of death with neurodegenerative disease, including Alzheimer's disease, other types of dementias, motor neuron disease, and Parkinson's disease, compared with general population controls.The large risk increase was observed for all types of neurogenerative disease, and ranged from a 2-fold increase for death with Parkinson's disease to a 5-fold increase for death with Alzheimer's disease. Another study showed that the proportion of male professional football players in France who died from dementia was increased, although comparisons could only be made with the national average for men in the same age group, the number of outcome events was low, and Parkinson's disease and amyotrophic lateral sclerosis could not be analysed due to few events.Studies of amyotrophic lateral sclerosis in Italianand Spanish football playershave also been published, although these studies had substantial limitations including the absence of a control group, or incomplete capture of outcome events. Following the publication of the Scottish study,the Union of European Football Associations—UEFA—and the British football federations updated their guidelines to reduce exposure to heading among youth players.Limits for the number of headers that involve higher forces, such as those following a long pass or from corners kicks, were later introduced for adult amateur players and professional players in England.
ALS in Italian professional soccer players: the risk is still present and could be soccer-specific.
The overall evidence supports the hypothesis that former elite football players are at increased risk of neurodegenerative disease, especially Alzheimer's disease and other dementias, and that the risk increase is limited to outfield players. However, the magnitude of the association might differ across populations of elite football players, including female players, and whether the findings are generalisable to contemporary football players is uncertain.
Our cohort study included 6007 male football players who had played in the Swedish top division in 1924–2019 and 56 168 general population controls who were matched with players based on sex, age, and region of residence. Neurodegenerative disease was identified in nationwide registers of death certificates and diagnoses registered during hospital admissions and outpatient visits and use of prescription drugs. We found that football players were at increased risk of diagnosis of neurodegenerative disease. The risk increase was observed for Alzheimer's disease and other dementias, but not for other types of neurodegenerative disease, and among outfield players but not among goalkeepers. Almost all the neurodegenerative disease events occurred in players who had played elite football during the mid-20th century. Using a comprehensive sample of elite football players and nationwide health registers with high validity, our study provides data that can be used to manage risks in football.
We aimed to identify studies of neurodegenerative disease outcomes in football players. We searched PubMed for articles in any language published from database inception to July 20, 2022, with search terms “football” or “soccer” and “neurodegenerative” or “dementia” or “Alzheimer's disease” or “motor neuron disease” or “amyotrophic lateral sclerosis” or “Parkinson's disease”. We found one study based on a cohort of male former professional football players in Scotland which had found a 3·5-fold increased risk of death with neurodegenerative disease among football players compared with general population controls, and another study based on the same cohort presenting complementary analyses that also accounted for diagnostic codes registered during hospital visits and use of dementia-related prescription drugs. We found one study showing that the proportion of male professional football players in France who died from dementia was increased, although comparisons could only be made with the national average for men in the same age group, the number of outcome events was low, and Parkinson's disease and amyotrophic lateral sclerosis could not be analysed due to few events. We also found analyses of amyotrophic lateral sclerosis in Italian and Spanish male football players, although these studies suffered from substantial limitations including the lack of a control group, or incomplete capture of outcome events.
Concerns have been raised about a potentially increased risk of neurodegenerative disease associated with playing football. The concerns are based on evidence linking traumatic brain injuries, including both concussions and repetitive sub-concussive injuries without symptoms, to an increased risk of neurodegenerative disease, potentially through a specific neurodegenerative pathology known as chronic traumatic encephalopathy.Symptomatic brain injuries are infrequent in football (<0·1 events per 1000 player-hours in most studies),while redirecting the ball using the head constitutes an integral part of the game. The trauma sustained through repeatedly heading a football has been suggested to cause neurodegeneration, although the evidence for such a link is inconsistent, incomplete, and controversial.
Heading in the right direction: a critical review of studies examining the effects of heading in soccer players.
Head injury in soccer: from science to the field; summary of the head injury summit held in April 2017 in New York City, New York.
Head injury in soccer: from science to the field; summary of the head injury summit held in April 2017 in New York City, New York.
Football is the most popular sport in the world. In 2006, The International Federation of Association Football (FIFA) estimated that football was played competitively by 265 million people in more than 200 countries.
The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
HRs with CIs that did not overlap 1 were considered statistically significant. Analyses were done with SAS software (version 9.4).
Finally, in the main analyses players and controls were censored at emigration. Because players might emigrate during their career and then move back to Sweden at an older age, we did an analysis in which players and controls were not censored at emigration.
The National Patient Register, including data on diagnoses during hospital admission, was available nationwide from 1987. We therefore conducted analyses in which we set the earliest cohort entry to Jan 1, 1987, such that the cohort had full coverage with respect to data from both death certificates and hospital admissions during the whole study period. In these analyses, we excluded players who died or emigrated before 1987 and their matched controls, controls who died or emigrated before 1987, and players and controls who were 65 years or older in 1987 (to avoid potential bias introduced by left truncation).
We did pre-specified sensitivity analyses for the primary outcome. The earliest possible cohort entry was in 1969. If players had an increased risk of death with neurodegenerative disease, the exclusion of players who died before 1969 and inclusion of those who survived might introduce left truncation bias due to a depletion of players susceptible to neurodegenerative disease.We therefore conducted analyses including only players and controls who were 64 years or younger at cohort entry. This age cut-off was chosen because death from neurodegenerative disease is rare in those younger than 65 years.
Point: incident exposures, prevalent exposures, and causal inference: does limiting studies to persons who are followed from first exposure onward damage epidemiology?.
We followed players and controls from cohort entry to emigration, outcome event, death, or end of follow-up (Dec 31, 2020). We used Cox regression analysis with age as the time scale to estimate the hazard ratio (HR) with 95% CI for the risk of each study outcome among football players versus controls. The models were adjusted for region of residence, place of birth, and calendar time as a time-varying covariate; variable definitions are shown in the appendix (p 7) . We tested the assumption of proportional hazards by including an interaction term between age of analysis, first as a continuous variable and then as a categorical variable (0–29, 30–50, 51–70, 71–90, and ≥90 years), and a dummy variable for football player status and assessing whether the interaction was significant (p<0·05) using the Wald test. In the additional analyses by player position, we assessed the risk of outcomes among outfield players and goalkeepers, compared with controls, and in separate analyses restricted to football players, we compared outfield players with goalkeepers.
For football players versus controls, we compared the risk of outcomes that might be affected by risk factors for neurodegenerative disease, including all-cause mortality, death without neurodegenerative disease, death with cardiovascular disease, death with chronic obstructive lung disease or lung cancer (as an indicator of smoking), and death with alcohol-related disorders; definitions of these outcomes are provided in the appendix (p 6) . We then performed a post-hoc analysis for the primary outcome which accounted for the competing risk from death without neurodegenerative disease using the Fine and Gray method.
In additional analyses, we assessed the primary outcome and Alzheimer's disease or other dementias, motor neuron disease, Parkinson's disease and all-cause mortality in separate analyses for outfield players and goalkeepers (vs controls), and in analyses comparing outfield players with goalkeepers. These analyses were done because goalkeepers rarely head the ball,but share exposures with outfield players that are specific to football players. We also performed post-hoc analyses of football players and their matched controls by year that the football player played his first season in the top division (1924–39, 1940–49, 1950–59, 1960–69, and 1970–2019).
Secondary outcomes were each type of neurodegenerative disease, as well as death with the neurodegenerative disease composite and with each of the neurodegenerative disease types. The International Classification of Diseases and ATC codes used for the definition of the outcomes are shown in the appendix (pp 4–5 ).
The primary outcome was neurodegenerative disease: a composite of Alzheimer's disease or other dementias, motor neuron disease (including amyotrophic lateral sclerosis), and Parkinson's disease recorded in the Cause of Death Register (primary or contributing causes of death), the National Patient Register (primary or secondary diagnosis during hospital admission or outpatient visits), or the Prescribed Drug Register (filled prescription for drugs used for treatment of dementia).
We obtained information about date of immigration and emigration, birth region, and region of residence from the Total Population Register,which contains data on life events for all individuals living in Sweden, with almost complete coverage. We obtained data on outpatient and emergency department visits and inpatient admissions to all hospitals in Sweden from the National Patient Register. The register has nationwide coverage for inpatient admissions from 1987 onward and for outpatient visits to hospitals from 2001 onward.The Cause of Death Register provided data on the date and cause of death;we used data from this register from 1969 onward. The National Prescribed Drug Registerprovided individual-level data on all drug prescriptions at all pharmacies in Sweden since July, 2005. The register includes the anatomical therapeutic chemical (ATC) code of the dispensed drug, information about the amount of drug dispensed, and the dispensation date.
The new Swedish Prescribed Drug Register—opportunities for pharmacoepidemiological research and experience from the first six months.
The study was approved by the Swedish Ethical Review Authority and informed consent from participants was not required.
From the base cohort, we excluded players and controls who were not born in Sweden and who had their first date of residence in the country after age 15 years, because foreign players are likely to emigrate during or shortly after their football career which would preclude long-term follow-up. Additional information about the data sources and the methods used to establish the study cohort are provided in the appendix (pp 2–3 ).
Each football player was matched with male controls from the general population, in a 1:10 ratio, to form a base cohort. The matching was done without replacement and was based on year of birth, region of residence (to account for factors such as population density, ambient temperature, and socioeconomic factors), and vital status (ie, both the football player and the matched controls had to be alive) on January 1 in the year of the first game played in Allsvenskan or the first day of registered residency in Sweden (if this occurred after the first season in Allsvenskan) and was done in the Total Population Register. For players who played their first season before the start of the Total Population Register (1969), controls were selected based on male sex, age, vital status, and region of residence on Jan 1, 1969. Cohort entry was Jan 1 of the year during which players and controls were matched.
The collected data included the full name, birth date, the number of seasons and games played, teams represented, and player position (outfield or goalkeeper). Using publicly available databases administered by the Swedish Tax Agency and The Federation of Swedish Genealogical Societies, we used the name and birth date to identify the player's personal identity number, a number assigned to all inhabitants in Sweden which enables linkage of individual-level information across data sources ( appendix pp 2–3 ). We excluded players whose personal identity number could not be retrieved because they were not found in the databases (eg, players who had removed themselves from public databases, foreign players who did not receive a personal identity number, or players who had died before the introduction of personal identity numbers) or because there were multiple individuals with the same name and birth date, and it was not possible to identify who among them was the football player. We then excluded players who were not registered in the Total Population Register, which started in 1969.
In this cohort study, we included all male football players who had played at least one game in the Swedish top division Allsvenskan, from Aug 1, 1924 to Dec 31, 2019. Information about football players was collected using data sources of all current and former football players in Allsvenskan compiled by the Swedish Association of Football Historians and Statisticians. Although many of the football players competed at the highest international level, professional contracts were not allowed in the Swedish top division until the late 1960s. By the late 1990s almost all top division players had football as their full-time occupation.
Sensitivity analyses that included those who were 64 years or younger at cohort entry ( appendix p 10 ), those who were 64 years or younger and entered the cohort in 1987 or later ( appendix p 11 ), and that did not account for emigration yielded HRs which were similar to that of the primary analysis ( appendix p 12 ).
Risk of all-cause mortality (HR 0·95 [95% CI 0·91–0·99]), death without neurodegenerative disease (HR 0·90 [0·86–0·94]), and death with chronic obstructive lung disease or lung cancer (HR 0·82 [0·72–0·94]) were all lower among football players than controls. There were no significant group differences in the risk of death with cardiovascular disease (HR 0·96 [95% CI 0·91–1·02]) and death with alcohol-related disorders (HR 0·87 [0·70–1·08]; table 4 ). In analyses accounting for the competing risk from death without neurodegenerative disease, the subhazard ratio for neurodegenerative disease was 1·44 (95% CI 1·32–1·58).
The risk of neurodegenerative disease was increased among football players versus population controls for football players who played their first season in the top division in 1924–1939, 1940–1949, 1950–1959, and 1960–1969 ( appendix p 9 ). The risk increase was not observed for those who played their first season in 1970–2019, although the analysis of this subgroup included few outcome events and the limited follow-up time meant that it could not provide information about the risk of neurodegenerative disease in older age, when most such events occur.
In the direct comparison, outfield players, compared with goalkeepers, had a significantly higher risk of neurodegenerative disease (HR 1·43 [95% CI 1·03–1·99]). Alzheimer's disease and other dementias were significantly higher among outfield players versus goalkeepers (HR 1·43 [95% CI 1·01–2·02]). Risk of Parkinson's disease was not significantly different in the two groups (HR 1·28 [95% CI 0·46–3·55]), and motor neuron disease was not assessed as no events occurred among goalkeepers. Risk of all-cause mortality did not differ among outfield players versus goalkeepers (HR 1·12 [95% CI 0·97–1·29]; table 3 ).
In the additional analyses by player position, outfield players (HR 1·50 [95% CI 1·36–1·65]) but not goalkeepers (HR 1·07 [0·78–1·47]) had a significantly higher risk of neurodegenerative disease compared with controls ( table 3 ). Compared with population controls, outfield players (HR 1·67 [95% CI 1·51–1·84]), but not goalkeepers (HR 1·19 [0·85–1·66]), had an increased risk of Alzheimer's disease and other dementias, and both outfield players (HR 0·70 [0·53–0·92]) and goalkeepers (HR 0·53 [0·20–1·41]) had a lower risk of Parkinson's disease, although the reduced risk was significant only for outfield players.
Football players, compared with controls, had a significantly increased risk of death with neurodegenerative disease (HR 1·54 [95% CI 1·37–1·73]) and death with Alzheimer's disease and other dementias (HR 1·69 [1·50–1·92]). No significant risk increase for football players versus controls was observed for death with motor neuron disease (HR 1·41 [95% CI 0·79–2·53]) and death with Parkinson's disease was lower among football players (HR 0·70 [0·48–1·03]), although the difference was not significant. Almost all the neurodegenerative disease events in football players occurred among those born in 1954 or earlier ( appendix p 8 ).
In secondary outcome analyses, football players, compared with controls, had a significantly increased risk of Alzheimer's disease and other dementias (HR 1·62 [95% CI 1·47–1·78]). No significant risk increase for football players versus controls was observed for motor neuron disease (HR 1·27 [95% CI 0·73–2·22]), and the risk of Parkinson's disease was lower among football players (HR 0·68 [0·52–0·89]; table 2 ).
During a mean follow-up time of 27·6 (SD 15·6) years (median 27 years [IQR 14–41]), 537 (8·9%) of 6007 football players and 3485 (6·2%) of 56 168 matched controls had the primary outcome of neurodegenerative disease recorded as a cause of death, a diagnosis registered during a hospital admission or outpatient visit, or use of a prescription drug for dementia. Mean follow-up time was 26·8 years (SD 16·6) for football players and 27·7 years (15·4) for controls. Mean age at the outcome event was 79·7 years (SD 7·9) for football players and 79·8 years (9·0) for controls. The risk of neurodegenerative disease was significantly higher among players versus controls (HR 1·46 [95% CI 1·33–1·60]; table 2 ). The analyses fulfilled the proportional hazards assumption when the interaction with football player status was assessed using age as both a continuous and categorical variable (p>0·05 for all interaction terms).
Primary outcome: a composite of Alzheimer's disease or other dementias, motor neuron disease, and Parkinson's disease diagnoses recorded in death certificates, during inpatient and outpatient hospital visits, and through filled prescriptions for drugs used for the treatment of dementia.
7386 football players had played at least one game in the top Swedish division between Aug 1, 1924, and Dec 31, 2019. 182 players were excluded due to an unretrievable personal identity number, 412 were excluded due to their number not being identified in the Total Population Register, and five were excluded due to a reused identity number or multiple entries with the same number. 6787 football players were matched with 67 795 controls. The number of population controls available per football player ranged between three and ten and 99·2% of the football players had ten matched controls. 780 players and 11 627 controls were excluded due to being born outside of Sweden and not having immigrated to the country before age 15 years, or due to being a control matched to a football player who was excluded for this reason. The study population included 6007 football players (510 of whom were goalkeepers) and 56 168 matched controls ( figure ). The median number of seasons played in the top division among football players was three (IQR 1–7; table 1 ). Follow-up was complete for 59 207 (95·2%) of the 62 175 study participants while 2968 (4·8%) were censored due to emigration.
Cohort entry was Jan 1 in the year of the first season played in the Swedish top division Allsvenskan, Jan 1 in the year that the player was first registered as a resident in Sweden (if this occurred later than his first season in Allsvenskan), or Jan 1, 1969 (if the player played his first season before 1969).
Discussion
In this cohort study, male football players in the Swedish top division had a 1·5-fold increased risk of neurodegenerative disease compared with population controls who were matched on sex, age, and region of residence. In analyses by type of neurodegenerative disease, the risk increase was observed for Alzheimer's disease and other dementias, while risk of motor neuron disease was similar among football players and controls and risk of Parkinson's disease was lower among football players compared with controls. The risk increase for neurodegenerative disease was only observed for outfield players and not for goalkeepers, and outfield players had a significantly higher risk of neurodegenerative disease than did goalkeepers.
29 Nguyen VT
Zafonte RD
Chen JT
et al. Mortality among professional american-style football players and professional american baseball players. , 30 Daneshvar DH
Mez J
Alosco ML
et al. Incidence of and mortality from amyotrophic lateral sclerosis in national football league athletes. , 31 Morales JS
Valenzuela PL
Saco-Ledo G
et al. Mortality risk from neurodegenerative disease in sports associated with repetitive head impacts: preliminary findings from a systematic review and meta-analysis. 19 Ludvigsson JF
Andersson E
Ekbom A
et al. External review and validation of the Swedish national inpatient register. The risk of neurodegenerative disease in participants of sports that entail trauma to the head, including boxing, ice hockey, rugby, American football, and football, has been subject to much concern, but epidemiological studies of neurodegenerative disease outcomes among contact sport athletes have been limited in number and size.By using a comprehensive sample of elite football players and nationwide health registers with high validity,our study expands on the data regarding neurodegenerative disease risk among elite football players.
9 Mackay DF
Russell ER
Stewart K
MacLean JA
Pell JP
Stewart W Neurodegenerative disease mortality among former professional soccer players. , 32 Russell ER
Mackay DF
Stewart K
Maclean JA
Pell JP
Stewart W Association of field position and career length with risk of neurodegenerative disease in male former professional soccer players. 33 Wilson J Inverting the pyramid. The history of football tactics. , 34 Sund B Fotbollens strateger. Spelsystem och ledarskap i ett internationellt och svenskt historiskt perspektiv. In line with the previous study from Scotland,we found a significantly higher risk of neurodegenerative disease diagnosis and neurodegenerative disease mortality among football players than population controls. However, the magnitude of the association was not as large as that observed in the Scottish study, which reported a 3·5-fold increase among football players. Moreover, although the large risk increases were observed for all types of neurodegenerative disease among football players in Scotland, the risk increase was only observed for Alzheimer's disease and other dementias, while risk of motor neuron disease was not significantly different in football players versus controls and risk of Parkinson's disease was lower among football players in our study. It could be hypothesised that the weaker association with neurodegenerative disease observed for football players in Sweden versus Scotland could reflect differences in playstyles, practice routines, and frequency of play that the players were exposed to.Sweden was a prominent football nation during the 20th century and many of the players from the top division were competing at the highest international level. However, due to ideals of sportsmanship and amateurism, football clubs in Sweden were not allowed to pay salaries to their football players until the late 1960s. By the late 1990s, almost all top division players had football as their full-time occupation. Further research is needed to investigate how exposures associated with health outcomes might differ across populations of elite football players.
6 Putukian M
Echemendia RJ
Chiampas G
et al. Head injury in soccer: from science to the field; summary of the head injury summit held in April 2017 in New York City, New York. , 35 Tarzi C
Tarzi G
Walker M
Saarela O
Cusimano MD Medical assessment of head collision events in elite women's and men's soccer. 22 Tierney GJ
Higgins B The incidence and mechanism of heading in European professional football players over three seasons. It has been hypothesised that the repetitive mild head trauma sustained through heading the ball and concussions might cause neurodegenerative disease.Although goalkeepers rarely head the ball,they share exposures with outfield players that are specific to football players. In our study, goalkeepers and outfield players had similar risk of all-cause mortality, indicating that their overall health status was comparable. However, outfield players had an increased risk of neurodegenerative disease compared with population controls, while goalkeepers had a similar risk to population controls. Accordingly, in a direct comparison, outfield players had a higher risk of neurodegenerative disease than goalkeepers. The risk increases for outfield players versus goalkeepers was only observed for Alzheimer's disease and other dementias, while the risk of Parkinson's disease was similar in the two groups and motor neuron disease was not assessed due to no events among goalkeepers.
Although our study indicates that neurodegenerative disease could be a potential occupational hazard among elite football players, its generalisability to female elite players and to male and female amateur and youth players, who constitute most football players in the world, is uncertain. It is probable that exposures related to football are extreme among elite players and that potential associations with neurodegenerative disease, if any, are less pronounced among recreational players. Moreover, football players in our study had a slightly lower risk of all-cause mortality and death with chronic obstructive lung disease or lung cancer, underscoring that elite football is associated with important health benefits.
36 Minns J
Holme P Watch your head: the physics of heading a football. 33 Wilson J Inverting the pyramid. The history of football tactics. , 34 Sund B Fotbollens strateger. Spelsystem och ledarskap i ett internationellt och svenskt historiskt perspektiv. Moreover, as neurodegenerative disease was rare in younger ages, most players who had neurodegenerative disease events in our study played elite football during the mid-20th century. During the past decades, football has changed in many ways which might affect the risk of neurodegenerative disease. Although the regulation weight of the footballs has not changed since the late 19th century, the latter half of the 20th century saw a gradual replacement of leather balls with synthetic balls that do not soak up water.Playstyles and practice routines have also substantially changed.It could be speculated that contemporary elite players, owing to more rigorous training, better equipment and, possibly, playstyles associated with less head trauma, might have a lower risk of neurodegenerative disease than did individuals who played elite football during the 20th century. Conversely, it could also be hypothesised that the risk might be higher among contemporary football players due to their exposure to more intense and frequent games and practice from a young age.
37 Livingston G
Huntley J
Sommerlad A
et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. 38 Ueda P
Pasternak B
Svanström H
et al. Alcohol related disorders among elite male football players in Sweden: nationwide cohort study. 9 Mackay DF
Russell ER
Stewart K
MacLean JA
Pell JP
Stewart W Neurodegenerative disease mortality among former professional soccer players. 39 Russell ER
McCabe T
Mackay DF
et al. Mental health and suicide in former professional soccer players. The higher risk of neurodegenerative disease observed among football players could also be explained by exposures other than trauma to the head sustained through football. Known risk factors for dementia include air pollution, less education, infrequent social contact, excessive alcohol consumption, smoking, hearing impairment, depression, head injury, physical inactivity, and cardiometabolic risk factors (including hypertension, diabetes, and obesity).Although the lack of detailed data on risk factors for neurodegenerative disease constitute a limitation of our study, we performed additional analyses to explore their potential influence on our findings. Compared with controls, football players in our study had a slightly lower risk of all-cause mortality, death without neurodegenerative disease, and death with chronic obstructive lung disease or lung cancer, while there was no significant difference between the groups in the risk of death with cardiovascular disease and death with alcohol-related disorders. In a previous study,we found that elite football players in Sweden had a lower risk of alcohol-related disorders than did general population controls. We also found that they had higher income and were more likely to undergo secondary or short university education, but less likely to undergo long university education. Moreover, at military conscription at age 18, they had higher scores for stress resilience, muscle strength, and cardiorespiratory exercise capacity, and slightly lower scores for cognitive ability than did general population controls. However, data on income, education, or military conscription results were not available for the older birth cohorts, who had experienced most of the neurodegenerative disease outcomes in the present study. Moreover, professional football players in Scotland had a lower risk of all-cause death, death from ischaemic heart disease, death from lung cancer,depression, and alcohol-related disordersthan observed in general population controls. Therefore, differences in cardiometabolic risk factors, smoking, alcohol-related disorders, and overall health status of football players versus controls might not explain their higher risk of Alzheimer's disease and other types of dementia. Conversely, playing elite football requires good physical health, which was not required to be included in the study as a general population control. A more favorable risk factor profile for dementia among football players might have attenuated the observed association.
40 Julian TH
Glascow N
Barry ADF
et al. Physical exercise is a risk factor for amyotrophic lateral sclerosis: convergent evidence from Mendelian randomisation, transcriptomics and risk genotypes. , 41 Liu G
Ou S
Cui H
et al. Head injury and amyotrophic lateral sclerosis: a meta-analysis. , 42 Belbasis L
Bellou V
Evangelou E Environmental risk factors and amyotrophic lateral sclerosis: an umbrella review and critical assessment of current evidence from systematic reviews and meta-analyses of observational studies. , 43 Fang F
Hållmarker U
James S
et al. Amyotrophic lateral sclerosis among cross-country skiers in Sweden. 44 Olsson TT
Svensson M
Hållmarker U
James S
Deierborg T Delayed clinical manifestation of parkinson's disease among physically active: do participants in a long-distance ski race have a motor reserve?. , 45 Bellou V
Belbasis L
Tzoulaki I
Evangelou E
Ioannidis JPA Environmental risk factors and Parkinson's disease: an umbrella review of meta-analyses. 45 Bellou V
Belbasis L
Tzoulaki I
Evangelou E
Ioannidis JPA Environmental risk factors and Parkinson's disease: an umbrella review of meta-analyses. , 46 Jafari S
Etminan M
Aminzadeh F
Samii A Head injury and risk of Parkinson disease: a systematic review and meta-analysis. , 47 Marras C
Hincapié CA
Kristman VL
et al. Systematic review of the risk of Parkinson's disease after mild traumatic brain injury: results of the international collaboration on mild traumatic brain injury prognosis. , 48 Gardner RC
Byers AL
Barnes DE
Li Y
Boscardin J
Yaffe K Mild TBI and risk of Parkinson disease: a Chronic Effects of Neurotrauma Consortium Study. Evidence regarding the risk factors for motor neuron disease, including physical activity and traumatic brain injury, is less established.For Parkinson's disease, high levels of physical activity have been associated with a lower risk.Although data on the association of Parkinson's disease with other environmental exposures are less certain, some studies have shown an association between mild traumatic brain injury and an increased risk of the disease.Football players are generally more physically active than population controls, which could potentially explain the lower risk of Parkinson's disease observed among both outfield players and goalkeepers in our study.
49 Mariosa D
Hammar N
Malmström H
et al. Blood biomarkers of carbohydrate, lipid, and apolipoprotein metabolisms and risk of amyotrophic lateral sclerosis: a more than 20-year follow-up of the Swedish AMORIS cohort. , 50 Rizzuto D
Feldman AL
Karlsson IK
Dahl Aslan AK
Gatz M
Pedersen NL Detection of dementia cases in two Swedish health registers: a validation study. , 51 Feldman AL
Johansson ALV
Gatz M
et al. Accuracy and sensitivity of parkinsonian disorder diagnoses in two swedish national health registers. Other limitations of our study include the lack of data on the number and characteristics of heading exposures among players, the risk of exposure misclassification as some population controls might have played football in lower divisions, and the lack of data on exposures to traumatic head injuries or exposure to other collision sports (although the risk of neurodegenerative disease in the control group should include the effects of head injuries entailed in a control scenario of no exposure to elite football). Because elite football players might have played in lower divisions and in other countries, we did not assess the association between career length in the top division and the risk of neurodegenerative disease. The register-based outcome definitions did not provide information regarding underlying neuropathology or how the diagnosis was made; however, when validated against gold-standard clinical investigations, most ICD-codes for neurodegenerative disease have shown high specificity and positive predictive value.No ICD-codes exist for chronic traumatic encephalopathy, although this condition is not routinely evaluated when assessing cause of death.
In this nationwide cohort study, male football players who had played in the Swedish top division were at increased risk of neurodegenerative disease compared with general population controls. The risk increase was observed for Alzheimer's disease and other dementias (but not for other types of neurodegenerative disease) and among outfield players but not among goalkeepers. Although our study confirms that former elite football players are at increased risk of neurodegenerative disease, the association was smaller in our study than that observed in a previous Scottish study and was only observed for Alzheimer's disease and other types of dementia, but not for motor neuron disease or Parkinson's disease.