Anger and Depression in Middle-Aged Men: Implications for a Clinical Diagnosis of Chronic Traumatic Encephalopathy

There are no consensus-based or validated clinical criteria for the clinical diagnosis of chronic traumatic encephalopathy (CTE), although several sets have been proposed (1–4). During the 20th century, CTE was described as a neurological disorder characterized by dysarthria, gait disturbance, tremor, and cognitive impairment—with some cases remaining mild and not progressing and other cases progressing to advanced Parkinsonism and dementia (5–11). In recent years, the clinical features attributed to the neuropathology of CTE have been expanded greatly, and there is a risk that these purported clinical features and their association with specific types of neuropathology will be woven into a Gordian knot. According to Greek and Roman mythology, a Gordian knot is so intricate and complex that it is nearly impossible to disentangle or untie. A major shift in conceptualizing CTE in recent years has been to assert that a broad range of mental health and psychosocial problems that are present prior to death are clinical features of CTE. Examples of the breadth of mental health problems attributed to CTE include depression and anxiety (12–14); substance abuse (1, 12, 13); personality changes, anger control problems, and violence (12–14); and suicidal thinking and death by suicide (13–19). In regard to death by suicide, several reviews (20–23), one retrospective historical case review study (24), and one epidemiological study (25) have concluded that there is minimal or no scientific evidence to support this assertion in former professional athletes. Authors have attributed psychosocial problems, such as poor financial decisions and bankruptcy (12), gambling (1), and marital problems, separation, and divorce (19), to the neuropathology of CTE. Some authors have even asserted that generalized body aches and pain (12), headaches (12, 14, 26), and insomnia (19) are clinical features of CTE.

It is essential to appreciate that many of the symptoms and problems described above are common in the general population, such as depression (27), anxiety (28), anger (29), financial problems (30), marital problems or divorce (31), headaches (32), bodily pain (33), and insomnia (34). Moreover, people with medical, psychiatric, neurological, and neurodegenerative diseases that are etiologically separate from CTE are also expected to have some of these symptoms and problems. A major gap in the literature is the lack of studies focused on examining the specificity of the proposed clinical features of CTE in the general population and in people with known clinical conditions resembling CTE who have no prior exposure to repetitive neurotrauma. The purpose of this study was to examine two of the assumed core clinical features of CTE, anger control problems and depression (1), in middle-aged men from the general population. We hypothesized that a small minority of men in the general population would report problems with anger and depression and that in those who did there would be a strong association between these two psychological problems.

Methods

Participants

Participants in this study were men from the U.S. general population who participated as subjects in the standardization sample for the National Institutes of Health (NIH) Toolbox for the Assessment of Neurological and Behavioral Function (35, 36). This project sampled a community-dwelling population from 10 sites who were capable of following instructions in English or Spanish, could understand the informed consent process, and had adequate visual, auditory, vestibular, or motor functioning to complete some of the items in the test battery or availability of assistance or assist devices to complete tasks. The overall sample had a distribution of demographic characteristics similar to the 2010 U.S. Census.

The NIH Toolbox Emotion database contained 1,630 rows of data, with 1,445 unique participants at the time 1 assessment (i.e., 185 were assessed a second time). From this, 180 participants were men between the ages of 40 and 60 years old. Ten men were excluded due to prior head injury (i.e., variables ATO_Head, ATOPx_Head, and MedHx_TBI), and four were excluded for not completing the Anger-Affect Questionnaire, leaving a total of 166 men for analysis (mean age, 49.1 years [SD=5.9]).

This sample was predominantly Caucasian (80.7%; African American, 10.2%, American Indian/Native American, 3.6%, Asian, 1.8%, Native Hawaiian/Pacific Islander, 0.6%, unknown, 3.6%). Almost one-half (42.8%) identified as Hispanic or Latino. The assessment was administered in English only (65.1%), Spanish only (23.5%), and bilingually in English and Spanish (English primary=3.6%, Spanish primary=7.8%). Level of education in the sample was as follows: less than high school diploma=16.3%, high school diploma or general equivalency diploma=27.1%, some college or an associate’s degree=20.5%, bachelor’s degree=25.3%, master’s or doctoral degree=5.4% (missing=5.4%).

The week before the assessment, 70.5% of participants reported working (69.2% working for pay; 1.2% were working but not for pay, for example, at a family business), 3.0% were employed but not working, 13.2% were looking for work, and 10.2% were not working or looking for work (3.0% missing or prefer not to say). Of the 29.5% (N=49) who were not working the week before the assessment, the reasons they cited for not working were taking care of house/family (N=5), retired (N=5), vacation (N=2), temporary health reasons (N=1), have job/contract but off-season (N=4), laid off (N=8), disabled (N=7), other (N=12), missing/prefer not to say (N=5). Whether some of these men participated in sports in their youth or whether they served in the military was not recorded in the database.

Measures

NIH Toolbox Anger-Affect Fixed Form (version 2.0).

A five-item questionnaire that instructs participants to rate how frequently they have experienced symptoms associated with anger over the past 7 days (e.g., “I was grouchy”; “I felt annoyed”). Each item is rated on a 5-point Likert-type scale (i.e., never, rarely, sometimes, often, always), with a possible score range of 5–25. Internal consistency for this scale was high in this sample (Cronbach’s α=0.86).

NIH Toolbox Anger-Hostility Fixed Form (version 2.0).

A five-item questionnaire that instructs participants to rate how true statements are for them (e.g., “I have become so mad I have broken things”). Each item is rated on a 7-point Likert scale with the anchors “extremely untrue of me” and “extremely true of me,” with a possible score range of 5–35. Internal consistency for this scale was high in this sample (Cronbach’s α=0.82).

NIH Toolbox Perceived Hostility Fixed Form (version 2.0).

An eight-item questionnaire that asks participants to rate how often other people in their life have engaged in hostile behaviors (e.g., “argue with me,” “criticize the way I do things”) over the past month. Each item is rated on a 5-point Likert-type scale (i.e., never, rarely, sometimes, often, always), with a possible score range of 8–40. Internal consistency for this scale was high in this sample (Cronbach’s α=0.90).

NIH Toolbox Anger-Physical Aggression Fixed Form (version 2.0).

A five-item questionnaire that instructs participants to rate how true statements are for them (e.g., “I have become so mad I have broken things”). Each item is rated on a 7-point Likert-type scale with the anchors “extremely untrue of me” and “extremely true of me,” with a possible score range of 5–35. Internal consistency for this scale was adequate in this sample (Cronbach’s α=0.78).

NIH Toolbox Fear-Affect Fixed Form (version 2.0).

A seven-item questionnaire that instructs participants to rate how frequently they have experienced symptoms associated with anxiety over the past 7 days (e.g., “I felt worried”; “I felt tense”). Each item is rated on a 5-point Likert-type scale (i.e., never, rarely, sometimes, often, always), with a possible score range of 7–35. Internal consistency for this scale was high in this sample (Cronbach’s α=0.89).

Patient-Reported Outcomes Measurement Information System (PROMIS) Depression Short-Form 8b (version 1.0).

An eight-item questionnaire that instructs participants to rate how frequently they have experienced symptoms associated with depression over the past 7 days (e.g., “I felt sad”; “I felt worthless”). Each item is rated on a 5-point Likert-type scale (i.e., never, rarely, sometimes, often, always), with a possible score range of 8–40. Internal consistency for this scale was high in this sample (Cronbach’s α=0.93).

Statistical Analyses

Descriptive statistics were examined at the item level for the self-report questionnaires administered in this study. The total raw scores for each scale were correlated with one another in a correlation matrix. For exploratory purposes, participants were dichotomized based on their response to the item “I feel depressed” over the past 7 days from the PROMIS version 1.0 Depression Short-Form 8b (i.e., “possible depression”=sometimes, often, or always; “no depression”=never, rarely). Independent groups t tests compared anger, physical aggression, hostility, perceived hostility, anxiety, and hopelessness between the possible depression and no-depression groups.

Results

Individual item endorsements on the physical aggression and the hostility scales are presented in Table 1. When asked, “If I am provoked enough I may hit another person,” 11.4% endorsed that item as true, and 4.8% endorsed “I have become so mad that I have broken things” as true. The percentage of men who endorsed one or more of the five physical aggression questions as true (i.e., scoring a single question as ≥5 on a 7-point scale) was 17.5%. For the question “I wonder why sometimes I feel so bitter about things,” 11.4% endorsed that as true.

Individual item endorsements on the anger, perceived hostility, anxiety, and depression scales are presented in Table 2. The men were asked about their experience with anger in the past 7 days. For the item “I was irritated more than people knew,” 26.5% of men reported “sometimes” and 3.6% reported “often.” For the item “I felt angry,” 21.1% of men reported “sometimes” and 4.8% reported “often.” Regarding perceived hostility, the men were asked, “In the past month, please describe how often people in your life argue with me”; 35.0% responded “sometimes” and 3.7% responded “often” or “always.” When asked about anxiety over the past 7 days, 27.1% reported “sometimes” feeling anxious and 5.4% reported “often” feeling anxious. When asked whether they felt hopeless in the past 7 days, 13.9% said “sometimes” and 3.0% said “often” or “always.” When asked whether they felt depressed in the past 7 days, 25.9% said “sometimes” and 3.6% said “often” or “always.”

An intercorrelation matrix for the psychological health measures is presented in Table 3. Spearman’s rho (ρ) values are presented because all variables were nonnormally distributed (Shapiro-Wilk p<0.001 for all six variables). The highest correlations with anger were anxiety (ρ=0.61) and depression (ρ=0.55). Hostility (ρ=0.44) and feeling perceived hostility from others (ρ=0.35) were also significantly correlated with anger. The highest correlation with physical aggression was hostility (ρ=0.37). There were small significant correlations between physical aggression and anger (ρ=0.28), anxiety (ρ=0.28), and depression (ρ=0.27).

Participants who responded to the question “I feel depressed” with “sometimes” (25.9%), “often” (3.0%), or “always” (0.6%) were included in the possible depression group (N=49 [29.5%]). We did not consider the group to be clinically depressed; rather, we conceptualized them as having some degree of sad mood in their life at the time of the assessment. Those who responded “never” or “rarely” to this question were included in the no-depression group (N=117 [70.5%]). The possible mild depression group endorsed significantly more anger (Cohen’s d=0.96, large effect size), hostility (Cohen’s d=0.95), physical aggression (Cohen’s d=0.48), perceived hostility toward them by others (Cohen’s d=0.38), and anxiety (Cohen’s d=1.51) than the no-depression group. These results are presented in Table 4.

Discussion

This study illustrates that some degree of anger and aggression are reported by a sizable minority of middle-aged men in the general population. Nearly one out of five middle-aged men (17.5%) endorsed one or more of the five physical aggression questions as true (i.e., scoring a single question as ≥5). In addition, symptoms of anxiety and depression were also reported by a minority of men. Anger and hostility were correlated with depression and anxiety, meaning that all tended to co-occur. The subgroup of men who endorsed at least a mild degree of depression also endorsed much greater anger, hostility, and anxiety than men who did not endorse feelings of depression. The comorbidity of affective dysregulation in men in the general population is important to consider when conceptualizing whether former athletes, civilians, or military veterans might have CTE.

Depression

Before 2005, depression was not considered a core or defining clinical feature of CTE. CTE was considered a neurological disorder, and some neurobehavioral characteristics described in former boxers included child-like behavior and volatility. Severe psychiatric illness and severe substance abuse were described in many of the historic case studies of boxers with dementia pugilistica (5, 6, 9, 10, 37), including problems severe enough to require institutionalization. In the past few years, there has been a much greater emphasis on mental health problems, such as depression and anxiety (12–14), as a proposed core clinical feature of CTE. In many of the postmortem case studies (3), the men were reported by a loved one or collateral to have experienced depression at some point before their death; some were depressed immediately before death. The challenge for clinicians and researchers is to determine the extent to which depression is independently related to the neuropathology believed to be specific to CTE versus neuropathology due mostly or entirely to well-established endogenous factors, exogenous factors, or both.

The lifetime prevalence of depression in men born in the United States is approximately 20%−28% (38). Depression is usually conceptualized as multifactorial in causation, and it is believed to arise from the cumulative interaction (39, 40) of genetics (41), adverse events in childhood (42, 43), and ongoing life stressors (44, 45). Former athletes and military veterans may have other health problems that are associated with depression, independent of their risk of neuropathology suggestive of CTE. Numerous psychiatric, medical, and neurological conditions are associated with increased risk for depression, including chronic pain (46, 47); headaches and migraines (48, 49); chronic insomnia (50); sleep apnea (51); posttraumatic stress disorder (PTSD) (52); generalized anxiety disorder (53, 54); substance abuse (55); obesity (56); diabetes (57); hypothyroidism (58); low testosterone (59); cardiovascular, cerebrovascular, and small vessel ischemic disease (60–63); Parkinson’s disease (64); mild cognitive impairment (65); and Alzheimer’s disease (66). Therefore, the challenge of including depression as a defining clinical feature of CTE is related to specificity. There is a risk that attributing depression mostly or entirely to CTE implies that the depression is a manifestation of a progressive neurodegenerative disease, which may or may not be true in any given case. Research is needed to determine if and the extent to which depression or other relatively common psychiatric disorders (e.g., PTSD, generalized anxiety disorder, impulse control syndromes) should be considered a core clinical feature of CTE.

Anger Control Problems

Anger control problems and violence have been described in some of the historical case studies of chronic brain damage in boxers (6, 9, 10, 67). Some authors have questioned the extent to which some of these tendencies might be longstanding and more characterological (11) versus being entirely new-onset and solely neuropsychiatric in origin. Anger control problems have also been described in recent case studies of former football players who had postmortem evidence of CTE pathology (12, 14, 15, 68–71). There are very few studies of anger control problems in former contact sport athletes, such as football players and boxers. In a phone survey, 1,063 retired National Football League (NFL) players were asked if they had ever had attacks of anger in which they lost control and broke or smashed something worth more than a few dollars, hit or tried to hurt someone, or threatened to hit or hurt someone (72). A substantial percentage of retired athletes between the ages of 30 and 49 (30.7%) and those aged ≥50 (29.3%) said yes to this question. Importantly, however, the authors noted that the rate of anger dyscontrol in men in the U.S. general population for these survey questions is greater than that for the retired players (i.e., 54.8% for men between 30 and 49, and 47.2% for men ≥50 years old).

A subgroup of men in the general population have longstanding anger control problems (29). Twin studies suggest that individual differences in anger expression and coping styles can be explained in part by genetic factors, and there are substantial environmental influences (73). Spielberger’s trait anger scale suggests that anger proneness is related to two factors: anger temperament (i.e., experiencing anger without provocation) and anger reaction (e.g., experiencing anger in response to negative events) (74). There is an association between anger control difficulties and adverse childhood events (i.e., abuse), low socioeconomic status and income, and parental mental health and substance use problems (29). Individuals with clinically significant anger difficulties report higher rates of a variety of psychiatric disorders and substance use disorders, and they are more likely to report using tobacco, alcohol, and drugs than individuals without significant anger difficulties (29). There is also an association between impulsivity (as reflected by gambling) and anger (75).

It is important for clinicians and researchers to appreciate that when irritability and anger control problems emerge or clearly worsen over time, these problems might be associated with life stress, marital problems, sleep disturbance, chronic pain, chronic headaches, substance abuse, depression, a neurological problem, a neurodegenerative disease, or a combination of factors. As noted in the DSM-5, people with depression often have considerable problems with irritability (e.g., persistent anger, a tendency to respond to events with angry outbursts or blaming others, and an exaggerated sense of frustration over minor matters), and family members often notice social withdrawal. Men with depression have higher rates of anger attacks or aggression, substance abuse, and risk taking (76). In addition, people in the general population with anger control problems and PTSD are at increased risk for suicidality (77). As with depression, more research is needed to determine the extent to which anger- or other impulse-control problems are related specifically to the neuropathology of CTE.

Limitations

This study has several limitations. The data were obtained from the normative data set for the NIH Toolbox; information relating to lifetime history of participation in contact or collision sports, history of mild concussions, or prior military service was not included in the database. The subjects were evaluated in person, as part of the standardization of the Toolbox, and those who reported a history of head or brain injury were not included in this study. It is possible, perhaps even likely, that some of the men in our case series experienced one or more concussions during the course of their lives that went unreported at the time of their inclusion in this normative data set. Concussions are very common in men in the general population. It is also likely that some of them played contact sports at least briefly during their lives, because sports participation at the high school level is common. Nonetheless, this sample remains a normative one even if such unreported concussive events were experienced by some of the individuals included in this database.

Conclusions

Without question, there are reasons to be concerned about the long-term brain health and neurobehavioral functioning of athletes, civilians, and military service members who are exposed to repetitive neurotrauma. For decades, this concern has been noted for boxers. Imaging studies of retired boxers show evidence of macrostructural and microstructural changes to the brain (78–80). Studies of retired professional football players reveal both functional changes in brain metabolism (81), as measured by functional magnetic resonance imaging, and microstructural differences in white matter (82, 83), as measured by diffusion tensor imaging, in some athletes. Postmortem examinations of the brains of former professional football players have revealed diverse neurodegenerative changes (14, 84). Surveys have revealed that a substantial minority of former NFL players have mental health problems (38, 72, 85); chronic pain and opioid use is relatively common (86, 87), and those with depression and chronic pain also have greater life stress and financial difficulties (85), although the majority of former athletes appear to be functioning well in their daily lives.

Researchers and clinicians are encouraged to be cautious and circumspect when considering the clinical presentation of former athletes and military veterans; they should not uncritically assert or assume that all or nearly all symptoms and problems experienced by athletes and veterans are caused by specific neuropathology. This represents a perspective that is likely over-simplified. Many people classified as having CTE at autopsy have also shown preclinical forms of neuropathology associated with other neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease, and Lewy body disease (26, 88–91); many have met full neuropathological criteria for a separate disease (14) that presumably accounted for some portion of their clinical features. Moreover, in recent years CTE pathology has been identified in a small number of people with no known participation in collision or contact sports and no known exposure to multiple mild injuries to the brain, including individuals with a history of substance abuse (92), temporal lobe epilepsy (93), amyotrophic lateral sclerosis (94), multiple system atrophy (95), and other neurodegenerative diseases (96).

Former athletes and military veterans might experience mental health problems for reasons similar to people in the general population. Transitioning from being a professional athlete to a retiree in one’s 30s or 40s may present unique challenges, such as a loss of identity, worse self-esteem, changes in income, and reductions in positive health behaviors (97, 98). These challenges may be risk factors for the development of mental health problems. These transition difficulties may be more prominent when retirement occurs unexpectedly or in a forced manner (e.g., related to a career-ending injury) (99). Importantly, there are evidence-based and evidence-informed treatment options available for life stress, anxiety, depression, suicidality, chronic pain, sleep problems, headaches, substance abuse, and gambling. Providing these treatments might reduce symptoms and improve functioning in former athletes and military veterans.