Neurodevelopmental Disorders and Risk of Concussion: Findings from the National Collegiate Athletic Association Department of Defense Grand Alliance Concussion Assessment, Research, and Education (NCAA-DOD CARE) Consortium (2014–2017)

Abstract

Evidence suggests neurodevelopmental disorders (NDs) may be associated with an increased incidence of concussion, but no studies have cross-sectionally and longitudinally assessed the associations of NDs and sex with concussion in collegiate athletes. We sought to assess the odds and relative risk (RR) of concussion in athletes self-reporting a diagnosis of attention deficit/hyperactivity disorder (ADHD), learning disability (LD), and ADHD+LD. Data from the Concussion Assessment, Research and Education (CARE) Consortium (2014–2017) were used to evaluate the likelihood of concussion for male and female athletes with ADHD, LD, and ADHD+LD, relative to controls. Odds ratios (ORs) of concussion history prior to enrollment and relative risk ratios for incurring a concussion following enrollment, with and without concussion history were calculated for all groups. Athletes with self-reported diagnosis of ADHD, LD, and ADHD+LD were more likely to report a single concussion (OR range = 1.528 to 1.828) and multiple concussions (OR range = 1.849 to 2.365) prior to enrollment in the CARE Consortium, irrespective of sex compared with control athletes. While enrolled in CARE, male athletes with ADHD, LD, and ADHD+LD had greater risk of incurring a concussion (RR range = 1.369 to 2.243) than controls, irrespective of concussion history. Male athletes with ADHD+LD with concussion history (RR = 2.221) and without concussion history (RR = 1.835) had greater risk of incurring a concussion than controls. These results suggest NDs may be associated with increased odds of single and multiple concussions, irrespective of sex. However, when we accounted for concussion history, it appears only male athletes with ADHD+LD had greater risk than respective controls. There were no significant differences between females and males with ADHD, LD, or ADHD+LD for either odds of concussion history or risk for incurring concussion.

Introduction

Sport-related concussions are a growing public health concern in the United States, with over 3.4 million cases being treated in U.S. emergency departments between 2001 and 2012.¹ Current guidelines call for the identification of factors that may increase an individual's risk for sustaining a concussion.^2,3 Evidence suggests neurodevelopmental disorders (NDs) may be associated with greater risk of bodily injury over a lifetime, including concussion.^3–12 The two most common ND's among college students are attention deficit/hyperactivity disorder (ADHD) and specific learning disorder (LD).¹³

ADHD occurs in 4 to 10% of college athletes^4,14 and is characterized by inattentiveness, impulsivity, hyperactivity, and risk-taking behaviors.¹⁵ Accordingly, several studies in collegiate athletes report that individuals with ADHD have greater overall proportion and odds (odds ratio [OR] = 1.63) of lifetime concussion incidence.^4,16,17 Therefore, having ADHD may be a significant factor that increases the odds of concussion history and relative risk (RR) of incurring a concussion.

Specific learning disorder (LD), present in 11% of college students,¹⁸ is defined as any cognitive difficulty resulting in lower than expected academic achievement despite intelligence level, impacting reading (dyslexia), mathematics (dyscalculia), or others abilities such as written language (dysgraphia).¹⁵ There is some evidence suggesting that athletes with LD are more likely to have a lifetime concussion history than controls.^16,19 Therefore, having LD may also be a significant factor that increases the risk of incurring a concussion.

Although evaluated separately, ADHD and LD are common comorbidities.²⁰ It is presumed that comorbid ADHD+LD would lead to even greater concussion history and subsequently risk of incurring a concussion; however, one study systematically examined concussion likelihood in collegiate athletes with ADHD, LD, and ADHD+LD.¹⁶ The authors aggregated two studies for a total of 8056 college athletes and reported greater odds of three or more concussions among those with ADHD (OR = 2.88), LD (OR = 1.81), and ADHD+LD (OR = 3.38) relative to controls. Therefore, ADHD and LD may have a synergistic effect, resulting in even greater incidence of concussion among this oft-neglected population. However, this study relied on retrospective analyses and was primarily focused on male athletes (97% of sample), precluding causative associations to be made and limiting the generalizability of its findings.

Further limiting our understanding of how NDs influence concussion incidence are their potential interactions with other factors such as sex and sport-type (i.e., contact, non-contact, etc.). Several studies have reported that females incur concussions more frequently than males; however, given that females are more likely to report their symptoms to coaches and parents than males,²¹ it remains unclear whether this is a true difference or an artifact of reporting bias. Additionally, studies evaluating both male and females have reported that athletes in sports with greater contact level incur a greater number of concussions.^21,22 Thus, NDs, sex, and sport-type may interact to influence the odds and relative risk of incurring a concussion.

Accordingly, the current study sought to gain a more clear and comprehensive understanding of the influence of NDs on concussion likelihood in a large cohort of collegiate athletes (n = 34,634). Our primary aim was to evaluate the independent and combined influenced of ADHD and LD, and our secondary aim was to evaluate these NDs within the context of sex and sport contact level. We hypothesized that both male and female athletes with NDs would have greater history of multiple concussions at the time of enrollment and be at greater risk for incurring a concussion during their sporting season relative to controls. Further, we predicted that athletes with ADHD+LD would exhibit the greater odds and relative risk of concussion, irrespective of sex and across sport-type.

Methods

De-identified data were provided by the National Collegiate Athletic Association Department of Defense Grand Alliance (NCAA-DOD) Concussion Assessment, Research, and Education (CARE) Consortium from 2014-2017. The CARE consortium consists of multi-year, multi-modal assessment of the natural history of concussion described in greater detail elsewhere.²³ Participants signed a site-specific institutional review board–approved consent form that was also approved by the U.S. Army Medical Research and Materiel Command Human Research Protection Office.²³ All participants completed a comprehensive demographics and health history questionnaire including self-reported ADHD or LD diagnoses.

Participants were categorized according to the self-reported responses of having been diagnosed with ADHD or LD into four groups: 1) Controls; 2) ADHD; 3) LD; and 4) ADHD+LD (athletes responding an affirmative to both ADHD and LD). All athlete demographics and self-reported diagnoses of NDs and history of concussion were taken from evaluation Year 1 only. We chose, any, one, and two or more concussions for analyses due to a small number of athletes with three or more concussions in our groups of interest. Control athletes who reported having no prior diagnoses of NDs but reported having repeated a year of school were excluded from analyses to control for undiagnosed or unreported NDs (n = 1,239). Athletes with non-responses to sex also were excluded (n = 8; Fig. 1). For categorical contact analyses, we utilized the National Collegiate Athletic Association (NCAA) contact categories such as contact, limited-contact, non-contact sports, and non-NCAA athletes (i.e., intramurals and military service academy cadets).²⁴ Herein, we acknowledge that concussions occurring prior to CARE enrollment are all self-reported and were included for retrospective analyses. Prospective analyses included athletes incurring a concussion following CARE enrollment and injuries which were identified and verified by team medical personnel using the evidence-based DoD criteria.²⁵

FIG. 1.

Selection criteria.

Statistical analysis

Initially, frequencies were performed to identify athletes with any history of concussion, one prior, and two or more prior concussions among groups and by sex (Table 1). We used contingency tables (2 × 2) to calculate crude odds and crude relative risk ratios for all analyses. Retrospective odds ratios were then calculated to determine the odds for having a history of single, and multiple concussions for athletes with ADHD, LD, and ADHD+LD, relative to controls (Table 2). To examine the influence of NDs within sexes, retrospective odds ratios were calculated for single, and multiple concussions for female and male athletes with ADHD, LD, and ADHD+LD, relative to sex-specific controls. To examine the influence of NCAA contact category, retrospective odds ratios were calculated to determine the odds for having history of single, and multiple concussions by NCAA contact category for athletes with ADHD, LD, and ADHD+LD, relative to controls (Table 3). To examine the influence of NCAA contact categories between sexes, retrospective odds ratios were calculated to determine the odds of concussion history by NCAA contact category for females with ADHD, LD, and ADHD+LD, relative to males with ADHD, LD, and ADHD+LD.

Table 1.

Sample Characteristics for Concussion History (Hx)

	Controls		ADHD		LD		ADHD+LD
	No Hx	Hx	No Hx	Hx	No Hx	Hx	No Hx	Hx
Age	19.1 ± 1.4	19.3 ± 1.4	19.3 ± 1.4	19.5 ± 1.4	19.2 ± 1.5	19.4 ± 1.3	19.3 ± 1.4	19.8 ± 1.4
BMI	24.0 ± 3.5	24.8 ± 3.7	24.5 ± 3.9	25.1 ± 4.0	24.7 ± 4.4	25.3 ± 4.5	25.3 ± 4.9	26.3 ± 4.9
Overall sample	24,256	6866	981	532	295	134	212	111
Female (n = 12,559)	9205	2506	351	175	134	58	91	39
Male (n = 20,828)	15,051	4360	630	357	161	76	121	72
Concussion history by quantity
One prior	24,256	5221	981	386	295	97	212	77
Female (n = 2036)	9205	1841	351	129	134	40	91	26
Male (n = 3745)	15,051	3380	630	257	161	57	121	51
Two+ prior	24,256	1645	981	146	295	37	212	34
Female (n = 742)	9205	665	351	46	134	18	91	13
Male (n = 1120)	15,051	980	630	100	161	19	121	21
Concussion history by contact category
Contact	7026	3204	386	304	118	75	94	73
Female (n = 3172)	1923	969	97	74	33	26	30	20
Male (n = 8108)	5103	2235	289	230	85	49	64	53
Limited contact	5400	1284	259	109	87	39	56	24
Female (n = 3769)	2834	662	122	46	46	19	26	14
Male (n = 3489)	2566	622	137	63	41	20	30	10
Non-contact	4101	730	209	64	66	15	43	7
Female (n = 3449)	2664	531	114	45	48	12	30	5
Male (n = 1786)	1437	199	95	19	18	3	13	2
Non-NCAA	7728	1649	127	55	24	5	19	7
Female (n = 2169)	1784	344	18	10	7	1	5	0
Male (n = 7445)	5944	1305	109	45	17	4	14	7

ADHD, attention deficit/hyperactivity disorder; LD, learning disability; BMI, body mass index; NCAA, National Collegiate Athletic Association.

Table 2.

Retrospective Odds (Odds Ratios) of Concussion History Relative to Controls, and Females Relative to Males (F/M)

	Overall history of concussion				One prior concussion				Two or more concussions
	Odds	95% CI	χ2	p	Odds	95% CI	χ2	p	Odds	95% CI	χ2	p
Controls	0.940	0.89-0.99	4.845	< 0.05	0.891	0.84-0.96	13.248	< 0.01	1.110	1.00-1.23	4.005	< 0.05
F/M
ADHD	1.915	1.72-2.14	141.173	< 0.01	1.828	1.62-2.06	97.295	< 0.01	2.195	1.83-2.63	76.118	< 0.01
Females	1.931	1.52-2.21	41.465	< 0.01	1.838	1.49-2.26	33.829	< 0.01	1.814	1.32-2.49	13.924	< 0.01
Males	1.956	1.71-2.24	99.201	< 0.01	1.817	1.56-2.11	62.636	< 0.01	2.438	1.96-3.04	66.643	< 0.01
F/M	0.880	0.70-1.10	1.266		0.901	0.70-1.16	0.677		0.826	0.57-1.20	1.017
LD	1.604	1.31-1.97	20.612	< 0.01	1.528	1.21-1.93	13.075	< 0.01	1.849	1.31-2.61	12.552	< 0.01
Females	1.590	1.17-2.17	8.675	< 0.01	1.493	1.05-2.13	4.906	< 0.05	1.859	1.13-3.06	6.142	< 0.05
Males	1.629	1.24-2.14	12.345	< 0.01	1.577	1.16-2.14	8.738	< 0.01	1.812	1.12-2.93	6.075	< 0.05
F/M	0.917	0.61-1.38	0.171		0.843	0.53-1.34	0.518		1.138	0.57-2.26	0.138
ADHD+LD	1.849	1.47-2.33	28.013	< 0.01	1.687	1.30-2.19	15.604	< 0.01	2.365	1.64-3.41	22.629	< 0.01
Females	1.574	1.08-2.30	5.637	< 0.05	1.429	0.92-2.22	2.566		1.977	1.10-3.56	5.394	< 0.05
Males	2.054	1.53-2.75	24.046	< 0.01	1.877	1.35-2.61	14.498	< 0.01	2.665	1.67-4.26	18.246	< 0.01
F/M	0.720	0.45-1.16	1.845		0.678	0.39-1.17	1.966		0.823	0.44-1.61	0.264

CI, confidence interval; ADHD, attention deficit/hyperactivity disorder; LD, learning disability.

Table 3.

Retrospective odds (Odds Ratios) of Concussion History Relative to Controls, and Females Relative to Males (F/M) within Contact Categories

	NCAA contact				NCAA limited contact				NCAA non-contact				Non-NCAA
	Odds	95% CI	χ2	p	Odds	95% CI	χ2	p	Odds	95% CI	χ2	p	Odds	95% CI	χ2	p
Controls	1.151	1.05-1.26	8.961	<.01	0.964	0.85-1.09	0.355		1.439	1.21-1.72	16.748	<.01	0.878	0.77-1.00	3.831	.050
F/M
ADHD	1.727	1.48-2.02	48.104	<.01	1.770	1.40-2.23	23.843	<.01	1.720	1.29-2.30	13.657	<.01	2.030	1.47-2.80	19.455	<.01
F/M	0.959	0.68-1.36	0.057		0.820	0.52-1.29	0.743		1.974	1.08-3.60	5.008	<.05	1.346	0.58-3.14	0.474
LD	1.394	1.04-1.87	4.995	<.05	1.885	1.29-2.76	10.894	<.01	1.277	0.73-2.25	0.719		0.976	0.37-2.56	0.002
F/M	1.367	0.73-2.55	0.970		0.847	0.40-1.80	0.186		1.500	0.38-5.94	0.413		0.607	0.06-6.44	*	0.58
ADHD+LD	1.703	1.25-2.32	11.692	<.01	1.802	1.11-2.92	5.900	<.05	0.915	0.41-2.04	0.048		1.727	0.73-4.11	1.558
F/M	0.805	0.41-1.58	0.400		1.615	0.61-4.25	0.952		1.083	0.19-6.32	*	0.652	0.000	0.00-0.00	**

^* Indicates Fisher's exact p value is reported.

^** Indicates a cell size of zero and no analysis was completed.

CI, confidence interval; ADHD, attention deficit/hyperactivity disorder; LD, learning disability.

Frequencies were performed to identify athletes who incurred a concussion, without history, and with history of concussion among groups and by sex (Table 4). Prospective relative risk calculations were performed to determine risk for incurring a concussion for athletes with ADHD, LD, and ADHD+LD, relative to controls (Table 5). To examine the influence of NDs and concussion history within sexes, prospective relative risk calculations were performed for athletes with and without a concussion history with ADHD, LD, and ADHD+LD. To examine the influence of NCAA contact category, prospective relative risk calculations were performed by NCAA contact category for athletes with ADHD, LD, and ADHD+LD, relative to controls (Table 6). To examine the influence of NCAA contact categories between sexes, prospective relative risk calculations were performed by NCAA contact category for females with ADHD, LD, and ADHD+LD, relative to males with ADHD, LD, and ADHD+LD. All analyses were performed using SPSS (Version 26; SPSS Inc., Chicago, IL) and an a priori α was set at ≤0.05 for all analyses.

Table 4.

Sample Characteristics for Incurred Concussion (Cx)

	Controls		ADHD		LD		ADHD+LD
	No Cx	Cx	No Cx	Cx	No Cx	Cx	No Cx	Cx
Age	19.2 ± 1.4	18.8 ± 1.2	19.4 ± 1.5	19.2 ± 1.3	19.3 ± 1.5	18.9 ± .8	19.5 ± 1.5	19.4 ± 1.3
BMI	24.2 ± 3.6	24.7 ± 3.9	24.6 ± 3.9	25.6 ± 4.4	24.7 ± 4.2	26.9 ± 5.9	25.4 ± 4.8	26.8 ± 5.5
Incurred concussions	28,576	2546	1360	153	385	44	275	48
Female (n = 1148)	10,645	1066	476	50	175	17	115	15
Male (n = 1643)	17,931	1480	884	103	210	27	160	33
Incurred a concussion by history
No history	22,617	1639	912	69	276	19	190	22
Female (n = 720)	8518	687	330	21	127	7	86	5
Male (n = 1029)	14,099	952	582	48	149	12	104	17
History (N = 1042)	5959	907	448	84	109	25	85	26
Female (n = 428)	2127	379	146	29	48	10	29	10
Male (n = 614)	3832	528	302	55	61	15	56	16
Incurred a concussion by category
Contact	1266		99		32		40
Female (n = 415)	368		25		10		12
Male (n = 1022)	898		74		22		28
Limited contact	370		24		9		3
Female (n = 285)	266		13		5		1
Male (n = 121)	104		11		4		2
Non-contact	209		11		2		1
Female (n = 169)	157		9		2		1
Male (n = 54)	52		2		0		0
Non-NCAA	701		19		1		4
Female (n = 279)	275		3		0		1
Male (n = 446)	426		16		1		3

ADHD, attention deficit/hyperactivity disorder; LD, learning disability; BMI, body mass index; NCAA, National Collegiate Athletic Association.

Table 5.

Prospective Relative Risk (RR) for Incurring a Concussion Relative to Controls, and Females Relative to Males (F/M)

	Overall risk for concussion				Risk without history				Risk with history
	Risk	95% CI	χ2	p	Risk	95% CI	χ2	p	Risk	95% CI	χ2	p
Controls	1.194	1.11-1.29	21.242	< .01	1.180	1.07-1.30	11.73	< .01	1.249	1.11-1.41	12.632	< .01
F/M
ADHD	1.236	1.06-1.44	7.097	< .01	1.041	0.83-1.34	0.11		1.195	0.97-1.47	2.836
Females	1.044	0.80-1.37	0.099		0.802	0.53-1.22	1.08		1.096	0.78-1.55	0.266
Males	1.369	1.13-1.65	10.369	< .01	1.204	0.91-1.59	1.69		1.272	0.99-1.64	3.316
F/M	0.911	0.66-1.26	0.326		0.785	0.48 – 1.29	0.92		1.076	0.71-1.62	0.12
LD	1.254	0.95-1.66	2.420		0.953	0.62-1.48	0.05		1.413	0.99-2.02	3.381
Females	0.970	0.59-1.60	0.014		0.700	0.34-1.45	0.96		1.140	0.64-2.02	0.198
Males	1.494	1.04-2.14	4.694	< .05	1.178	0.68-2.04	0.34		1.630	1.3-2.58	4.048	< .05
F/M	0.777	0.44-1.38	0.742		0.701	0.28-1.79	0.60		0.874	0.35-2.05	0.135
ADHD+LD	1.817	1.40-2.37	18.847	< .01	1.536	1.03-2.29	4.35	< .05	1.773	1.26-2.50	9.841	< .01
Females	1.303	0.76-2.24	0.920		0.736	0.31-1.73	0.51		1.695	0.99-2.92	3.281
Males	2.243	1.64-3.07	24.083	< .01	2.221	1.42-3.47	11.98	< .01	1.835	1.18-2.85	6.731	< .01
F/M	0.675	0.38-1.19	1.898		0.391	0.15-1.02	3.22		1.154	0.58-2.29	0.165

CI, confidence interval; ADHD, attention deficit/hyperactivity disorder; LD, learning disability.

Table 6.

Prospective Relative Risk (RR) for Incurring a Concussion Relative to Controls, and Females Relative to Males (F/M) within Contact Category

	NCAA contact				NCAA limited contact				NCAA non-contact				Non-NCAA
	Risk	95% CI	χ2	p	Risk	95% CI	χ2	p	Risk	95% CI	χ2	p	Risk	95% CI	χ2	p
Controls	1.040	0.93-1.17	0.454		2.332	1.94-3.08	60.241	<.01	1.546	1.14-2.10	7.87	<.01	2.199	1.91-2.54	118.085	<.01
F/M
ADHD	1.159	0.96-1.40	0.299		1.178	0.79-1.76	0.643		0.931	0.51-1.69	0.06		1.396	0.91-2.15	2.252
F/M	1.025	0.67-1.56	0.014		1.407	0.65-3.06	0.750		3.226	0.71-14.65	*	0.09	1.031	0.32-3.31	*	.588
LD	1.340	0.97-1.85	3.072		1.290	0.68-2.44	0.608		0.571	0.14-2.26	*	0.32	0.461	0.07-3.17	*	.352
F/M	1.032	0.52-2.04	0.008		1.173	0.33-4.17	*	0.54	0.000	0.00-0.00	**		0.000	0.00-0.00	**
ADHD+LD	1.935	1.47-2.55	20.050	<.01	0.677	0.22-2.07	*	0.35	0.462	0.07-3.23	*	0.36	2.058	0.83-5.08	0.126
F/M	1.003	0.56-1.81	0.000		0.500	0.05-5.30	*	0.50	0.000	0.00-0.00	**		1.400	0.18-10.79	*	0.6

^* Indicates Fisher's exact p value is reported.

^** Indicates a cell size of zero and no analysis was completed.

NCAA, National Collegiate Athletic Association; CI, confidence interval; ADHD, attention deficit/hyperactivity disorder; LD, learning disability.

Results

Final analyses included 33,387 athletes comprised of 31,122 controls, 1513 athletes with ADHD, 429 athletes with LD, and 323 athletes with ADHD+LD. Retrospective odds can be found in Tables 2 and 3.

Overall odds

Analyses revealed that, irrespective of sex, athletes with self-reported diagnosis of ADHD (OR = 1.915), LD (OR = 1.604), or ADHD+LD (OR = 1.849) had greater odds of concussion history. Specifically, athletes with ADHD (OR = 1.828), LD (OR = 1.528), and ADHD+LD (OR = 1.687) had greater odds of one prior concussion. Further, athletes with ADHD (OR = 2.195), LD (OR = 1.849), and ADHD+LD (OR = 2.365) also had greater odds of two or more prior concussions

Odds within sex

Among females, athletes with ADHD (OR = 1.931), LD (OR = 1.590), and ADHD+LD (OR = 1.574) had greater odds of concussion history. Specifically, female athletes with ADHD (OR = 1.838) and LD (OR = 1.493) had greater odds of one prior concussion, but not ADHDL+LD (OR = 1.429). Further, female athletes with ADHD (OR = 1.814), LD, (OR = 1.859), and ADHD+LD (OR = 1.977) had greater odds of two or more prior concussions.

Among males, athletes with ADHD (OR = 1.956), LD (OR = 1.629), and ADHD+LD (OR = 2.054) had greater odds of concussion history. Specifically, male athletes with ADHD (OR = 1.817), LD (OR = 1.577), and ADHDL+LD (OR = 1.877) had greater odds of one prior concussion. Further, male athletes with ADHD (OR = 2.438), LD (OR = 1.812), and ADHD+LD (OR = 2.665) had greater odds of two or more prior concussions.

Odds between sexes

No significant differences were observed for history of one or multiple concussions between females and males with ADHD, LD, or ADHD+LD (OR ranges = 0.720-1.138; Table 2).

Overall odds by contact category

Analyses revealed that irrespective of sex, athletes in contact sports with self-reported diagnosis of ADHD (OR = 1.727), LD (OR = 1.394), and ADHD+LD (OR = 1.703) had greater odds of concussion history. Athletes in limited contact sports with ADHD (OR = 1.770), LD (OR = 1.885), and ADHD+LD (OR = 1.802) had greater odds of concussion history. Athletes in non-contact sports with ADHD (OR = 1.720) had greater odds of concussion history, but not with LD (OR = 1.277) or ADHD+LD (OR = 0.915).

Odds by contact category between sexes

No significant differences were observed between females and males in contact sports with self-reported diagnosis of ADHD, LD, or ADHD+LD (OR ranges = 0.805 - 1.367) for history of concussion. No significant differences were observed between females and males in limited contact sports with ADHD, LD, or ADHD+LD (OR ranges = 0.820-1.615) for history of concussion. However, in non-contact sports females with ADHD (OR = 1.974) had greater odds of concussion history relative to male athletes with ADHD. No significant differences were observed for history of concussion between non-NCAA females and males with ADHD or LD (OR ranges = 0.61-1.35), no analysis was performed for ADHD+LD due to a cell size of 0 (Table 3).

Overall risk

Overall relative risk estimates can be found in Tables 5 and 6. Analyses revealed that, irrespective of sex, athletes with self-reported diagnosis of ADHD (relative risk [RR] = 1.236) and ADHD+LD (RR = 1.817) had greater risk for concussion, but not LD (RR = 1.254). Specifically, athletes without concussion history but with ADHD+LD (RR = 1.54) had greater risk for concussion, but not ADHD (RR = 1.041) or LD (RR = 0.953). Further, athletes with concussion history and ADHD+LD (RR = 1.773) had greater risk for concussion, but not ADHD (RR = 1.195) or LD (RR = 1.413).

Risk within sex

Among females, no significant differences were observed for risk of concussion, for athletes with or without history of concussion ADHD, LD, or ADHD+LD (RR ranges = 0.700 - 1.695).

Among males, athletes with ADHD (RR = 1.369), LD (RR = 1.494), and ADHD+LD (RR = 2.243) had greater risk for concussion. Specifically, males without concussion history but with ADHD+LD (RR = 2.221) had greater risk for concussion, but not ADHD (RR = 1.204) or LD (RR = 1.178). Further, males with concussion history and LD (RR = 1.630) and ADHD+LD (RR = 1.835) had greater risk for concussion, but not ADHD (RR = 1.272; Table 5).

Risk between sexes

No significant differences were observed for risk of concussions between female and male athletes, with or without concussion history among ADHD, LD, or ADHD+LD (RR ranges = 0.382 – 1.154).

Risk by contact category

Analyses revealed that, irrespective of sex, athletes in contact sports with self-reported diagnosis of ADHD +LD (RR = 1.935) had greater risk for concussion, but not ADHD (RR = 1.159) or LD (RR = 1.340). There were no significant differences for limited contact, non-contact or non-NCAA for ADHD, LD, or ADHD+LD (RR range = 0.500-1.400) and risk for concussion.

Risk by contact category between sexes

No significant differences were observed for risk of concussion between females and males, irrespective of contact category, with ADHD, LD, or ADHD+LD (RR ranges = 0.500-1.400), no analyses were performed for non-contact with LD and ADHD+LD, nor non-NCAA with LD due to cell sizes of 0; Table 6).

Discussion

The primary aim of this study was to cross-sectionally and prospectively investigate the influence of ADHD, LD, and ADHD+LD on the crude odds of concussion history, and crude relative risk of concussion in collegiate athletes. In addition, as prior research focused on male athletes, we investigated whether there were sex differences for odds and risk of concussion in athletes with ADHD, LD, and ADHD+LD. Further, we examined the relative influences of concussion history and NDs by evaluating concussion risk for those with NDs who did and did not have a concussion history. Lastly, we sought to assess odds, and risk across NCAA contact category, to assess whether any sex differences were attributable to characteristics of the types of sports.

Consistent with our hypotheses, we observed that both male and female athletes with self-reported diagnosis of ADHD had significantly greater odds of single and multiple concussions than controls, replicating prior results observed in both high school and collegiate athletes.^4,17 Additionally, we prospectively observed that male athletes with ADHD had greater risk for incurring a concussion during the study period, replicating previous findings evaluating male collegiate athletes.²⁶ Interestingly, this same pattern was observed for both LD and ADHD + LD groups. That is, irrespective of ND status both males and females had an increased odds of concussion or multiple concussions; however, only males with NDs had an increased risk of incurring a concussion during the study. Thus, contrary to our hypothesis, sex does seem to be a critical factor intersecting with NDs to influence concussion risk.

This pattern is particularly interesting as females without an ND exhibited greater odds of having a history of multiple concussions than males, as well as a greater risk of incurring a concussion during the study, which replicates prior findings.²⁷ It should be noted, however, that when decomposing risk based on contact category, the interaction of sex and NDs disappears. Only athletes having LD + ADHD (irrespective of sex) who participated in contact sports exhibited increased risk. This may seem a direct contradiction to our other findings; however, the reduced numbers when stratifying data on contact category and sex by contact category are likely the reason. Closer inspection reveals that sex and sex by ND pattern of results observed across the larger data sex are close to being significant. Thus, although there were no significant differences between females and males with NDs for risk of incurring a concussion in any contact category, this may change with larger sample size and more balanced numbers within contact categories for males and females.

Mechanisms

Although the exact mechanisms that predispose an athlete with a ND towards incurring concussions is unclear, research suggests the inherent characteristics of NDs probably contribute to the greater occurrence of concussion among these athletes.²⁸ It is possible that structural and physiological alterations often observed in athletes with ADHD, such as reductions in axonal integrity,^29–32 glucose metabolism,³³ cerebral blood flow,³³ and catecholamine levels,^34-36 may render these athletes more susceptible to concussion injuries. Thus the “threshold at which concussion occurs” may be lessened in these athletes resulting in a greater incidence and risk for concussive injuries; however, to our knowledge no physiological or kinematic data exists in athletes with ADHD.

Although the physiology is complex, neuro-behavioral mechanism may also be related to the observed increased occurrence of concussion. ADHD is widely acknowledged to be a disorder of executive functions,^37-39 particularly those of inhibition (i.e., impulse control) and attentional focus. Inhibition deficits in athletes may manifest as unnecessary risk taking, ignoring the potential outcomes of negative or dangerous actions, or heightened distractibility. Athletes with ADHD, due to lack of inhibitory control, may place themselves in potentially dangerous situations, such as head-to-head contact. Also, deficits in inhibition may result in the inability to filter out irrelevant information, resulting in distraction and a potentially hazardous situation. Indeed, several studies have suggested that pre-concussion ADHD symptom severity plays a role in both concussion incidence and concussion symptom severity.^40,41 These considerations may contribute to the observed increased prevalence of concussions among athletes with ADHD. However, additional research is warranted.

The heterogeneity of deficits in LD makes identifying and understanding potential causal mechanisms more difficult. Like ADHD, individuals with LD frequently exhibit deficits in executive control (inhibition, working memory, mental flexibility), attention, and perceptual and sensory processing.¹⁵ However, compared with individuals with ADHD, those with LD tend to show no differences in impulse control,⁴² suggesting that differences in concussion risk are unlikely to be explained by impulsivity. Indeed, when accounting for attentional problems, the core deficits of LD appear to be processing speed, temporal processing, and working memory.⁴³ This suggests that instead of the behavioral and self-regulatory concerns of ADHD, individuals with LD have diminished central processing capacity that may be tasked during sport participation. Therefore, it is possible as previously described⁴⁴ that information processing difficulties may make these athletes more prone to concussive injury.

Athletes with co-existing ADHD+LD represent a unique consideration especially while lacking symptomology data. Theoretically, these athletes would share the common deficits of behavioral/self-regulations and attentional focus associated with ADHD concurrently with diminished processing speed and working memory deficits associated with LD. These shared yet divergent deficits may partially explain why the smallest group of interest demonstrated the largest odds for two or more concussions and the largest risk for concussion, irrespective of concussion history. However, the finding that females with ADHD, LD and ADHD+LD did not have increased risk despite having increased odds of concussion highlights the need to further investigate the complex interaction between sex differences, and NDs with regards to concussion incidence.

There are many possible explanations for these findings, including different styles of play between males and females, the sports that males and females participate in may have varying levels of “contact” due to different rulesets, or objective biological differences between males and females. Further, there are behavioral differences between males and females with NDs as well as different “cognitive styles.” Specifically, males with ADHD exhibit greater deficits in impulse control than females with ADHD and males with LD exhibit greater deficits in information processing, trading speed for accuracy across many cognitive domains. When considered with the other aforementioned factors, it is not unreasonable to expect that males with NDs, particularly those engaged in contact sports, would be at the greater risk of incurring single and multiple concussions across their collegiate career. However, further research taking a comprehensive bio-psycho-social approach is needed to further disentangle the confluence of sex and NDs on concussion incidence.

Clinical implications

The current findings have important clinical implications. Noting the diagnoses of NDs in medical history during pre-participation exams may help clinicians identify athletes at greater risk of injury, and who may experience prolonged recovery.^7,44–49 This can be used to help inform coaches to play close attention to these athletes and help guide them in modifying their “style of play” to reduce their probability of injury. These athletes would also be good candidates for increasingly popular and available “sport vision” and sport-specific attention training, which may also reduce their risk for injury.

With respect to clinical assessments, there is evidence to suggest athletes with ADHD demonstrate concussion-like symptoms at baseline, which reflect inherent characteristics of ADHD. Several studies have reported that college athletes without a history of concussion, but with ADHD demonstrated concussion-like symptoms at baseline, such as “problems learning,” “emotional lability,”⁵⁰ “difficulty concentrating,” “difficulty remembering,”^16,51 and vestibulo-ocular dysfunction.⁵² These inherent symptoms may influence neurocognitive testing, which is often used following concussion to assess cognitive performance. As noted above, athletes with NDs typically exhibit deficits in executive functions, which are disproportionately affected by concussive injuries.³⁷ Thus, data from these athletes require careful interpretation, as they are known to deviate from “normative data.” Likewise, NDs are considered modifying factors for managing concussion throughout RTP and return-to-school²; therefore, having athlete specific standard regarding “typical” symptoms and cognitive performance while acknowledging their ND diagnosis may prove beneficial during recovery and management.^42,44,52

Limitations

This study has several methodological limitations. Foremost, we were unable to include exposure rates (i.e., injuries per 1000 h/exposures) in our analysis due to inconsistent or missing data. Therefore, these findings should be considered a general overview of odds and risk for these selected populations. Due to the cross-sectional nature of retrospective self-reported ND diagnoses and prior concussions history, we cannot establish a causal influence of ND on concussion odds. However, because most NDs are diagnosed during early childhood, it is likely they received a diagnosis before sustaining a concussion, although some studies have reported secondary ADHD following a moderate or severe brain injury.²⁸ Additionally, diagnosis of NDs was self-reported, therefore it is possible that some athletes did not report their diagnosis due to concern of stigma associated with NDs. However, the rates of ADHD, LD, and ADHD+LD are relatively consistent with previous studies of collegiate athletes.^16,42,53 Of note, this sample contains sports in which there is a disproportionate number of a single sex participants; as such, it is possible our findings may be subject to sample bias. Indeed, despite having a large initial sample, several analyses were reliant on relatively smaller numbers (Table 2) as evidenced by the large confidence intervals with technically non-significant findings that trended towards significance (i.e., lower confidence intervals of 0.99).

Conclusion

In summary, this is the largest study to date to investigate retrospective odds and prospective risk of concussion while stratifying by sex and contact level among college athletes with ADHD, LD, and ADHD+LD. We found that athletes with ADHD, LD, and ADHD+LD had higher odds of sustaining single and multiple concussions prior to the time of enrollment relative to controls, and irrespective of sex. Once evaluated prospectively, however, only male athletes with ADHD, and ADHD+LD had greater risk for incurring a concussion relative to controls, irrespective of concussion history. Further, female athletes with ADHD, LD, or ADHD+LD did not appear to have greater concussion risk, despite having increased odds of multiple concussions. These are important findings as most studies use odds ratios to understand the influence of NDs to understand concussion incidence, but without retrospective analyses the conclusions drawn appear to be misleading, particularly when considering sex. Thus, data from prior and future research solely relying on odds ratios should be interpreted with caution.

To our knowledge, these are the first findings of its kind and make significant advances to prior research. The current results will help to inform researchers and clinicians and reinforce the need to further examine the differential factors that predisposing male and female athletes to concussion. Hopefully, these data will provide an impetus for additional longitudinal research.

CARE Consortium Investigators

Holly J. Benjamin, University of Chicago, Chicago, IL; Thomas Buckley, University of Delaware, Newark, DE; Kenneth L. Cameron, United States Military Academy, West Point, NY; Sara P.D. Chrisman, University of Washington, Seattle, WA; James R. Clugston, University of Florida, Gainesville, FL; Micky Collins, University of Pittsburgh Medical Center, Pittsburgh, PA; Stefan M. Duma, Virginia Polytechnic Institute and State University, Blacksburg, VA; CDR Carlos Estevez, United States Coast Guard Academy, New London, CT; Luis A. Feigenbaum, University of Miami, Coral Gables, FL; Joshua T. Goldman, University of California, Los Angeles, Los Angeles, CA; Joseph B. Hazzard, Jr., Bloomberg University, Bloomsburg, PA; Megan N. Houston, Keller Army Community Hospital, West Point, NY; April Hoy, Azusa Pacific University, Azusa, CA; Thomas W. Kaminski, University of Delaware, Newark DE; Louise A. Kelly, California Lutheran University, Thousand Oaks, CA; Anthony P. Kontos, University of Pittsburgh Medical Center, Pittsburgh, PA; Laura Lintner, Wake Forest University School of Medicine Winston-Salem, NC;. Christina L. Master, Children's Hospital of Philadelphia, Philadelphia, PA; Gerald McGinty, United States Air Force Academy Colorado Springs, CO; Jason P. Mihalik, University of North Carolina at Chapel Hill, Chapel Hill, NC; Justus Ortega, Humboldt State University, Arcata, CA; Nicholas Port, Indiana University, Bloomington, IN; Margot Putukian, Princeton University, Princeton, NJ; Steve Rowson, Virginia Polytechnic Institute and State University, Blacksburg, VA; Adam James Susmarski, United States Naval Academy, Annapolis, MD.

Funding Information

This work was supported by the Office of the Assistant Secretary of Defense for Health Affairs through the Psychological Health and Traumatic Brain Injury Program under Award No. W81XWH-14-2-0151.

Author Disclosure Statement

Steven Broglio, Michael McCrea, and Thomas McAllister received funding from the National Collegiate Athletic Association and Department of Defense to complete this investigation and cover travel costs related to the study. For the other authors, no competing financial interests exist.