Postconcussive symptom severity, risk factors for prolonged recovery, and mental health history: Pathways of influence in a diverse pediatric sample

Design and setting

The interdisciplinary TBI/Concussion Clinic is an outpatient clinic at a children's hospital with a level I trauma center. The team manages mild to moderate TBI in accordance with active rehabilitation principles based on the Centers for Disease Control and Prevention (CDC) guidelines for the management of mild TBI in children.¹⁷ Referrals include sports-related concussion from school athletic trainers, posthospital or emergency department discharge, or injuries managed by primary care physicians if patients are not recovering as expected. Clinic goals include estimating preinjury baseline functioning to assess individual recovery and clear children to return to high-risk activities once appropriate. The team includes specialized medical providers, athletic trainers, and neuropsychologists. Initial visits are focused yet comprehensive (2 h) with rapid assessment of developmental/academic history, current performance, risk factors for prolonged recovery, a validated symptom scale called the Post-Concussive Symptom Inventory (PCSI), neurological examination including balance testing, and neurocognitive screening. For predominantly Spanish-speaking families, the intake paperwork and history forms are provided in Spanish and visits are conducted with the aid of a certified hospital interpreter or bilingual provider. The patient completes a neurological assessment with the medical provider, balance testing with the ahletic trainer, and neurocognitive screening (either computerized or with a neuropsychologist, when appropriate). Interdisciplinary interpretation of all assessments directs an individualized plan of care focusing on active rehabilitation (e.g., symptom management, academic accommodations, and safe, noncontact physical activity). Based on CDC guidelines, an Acute Concussion Education (ACE) care plan is developed and shared with the nurse at the child's school. Follow-up care, including repeat neurocognitive screening, is also individualized based on the stage of recovery and needs of the child and family.

The Institutional Review Board at The University of Texas at Austin approved the prospective database collection of clinical information and the research review of clinical records. Patient data were extracted into a secure, deidentified research database following completion of clinical care.

Participants

Participants (N = 642) were adolescents aged 13.00 to 17.99 years who presented to a multidisciplinary TBI/concussion clinic in the southwest United States between July 2017 and February 2024. Inclusion criteria included: (1) received a Glasgow Coma Scale score ≥13, (2) had normal brain imaging findings or imaging not obtained, (3) presented to the clinic within 8 weeks of injury, and (4) followed through clinical recovery (e.g., not lost to follow-up). Participants with a neurological diagnosis (e.g., epilepsy) or specific developmental disorders (e.g., autism spectrum disorder or intellectual disability) were excluded. Although an interpreter or bilingual provider was present in approximately 15% of visits to facilitate communication with caregivers and family members, 98% of the adolescents were considered English proficient (e.g., receiving English-only school instruction).

Measures

Analytic strategy

Preliminary analyses were performed to examine frequency and descriptive information for all study variables using SPSS version 29. Correlational analyses were run to evaluate the relation between PCSI baseline scores, PCSI scores at visit 1, number of days from injury to visit 1, number of days from visit 1 to clearance, gender, and six variables related to preinjury history (e.g., history of depression, anxiety, ADHD, LD, migraines, and prior concussions).

Hypotheses were tested using mediational analysis in a structural equational framework (SEM), an approach that offers numerous conceptual and statistical advantages over standard regression methods.²⁰ Analyses were conducted in Mplus (Mplus v.8.4²¹), and missing data were handled using full information maximum likelihood estimation. Given the variability evident across previous research,⁷ the inclusion of predictors was based on results of correlational analyses within this sample. Mediation of the effects of preinjury anxiety and depression on PCSI visit 1 score (through PCSI baseline score) and on days from visit 1 to clearance (through PCSI visit 1 score) were tested using bootstrap confidence intervals (BOOTSTRAP = 5000) to examine the significance of the indirect effects.²²

Preliminary results

Demographic information, injury characteristics, and descriptive statistics for primary study variables are presented in Table 1. A total of 642 adolescents, ages 13.00 to 17.99, diagnosed with concussion were included in the SEM model. Of the 642 patients, 45% identified as female, 54% identified as Hispanic, and 41% received medical coverage through Medicaid. Approximately half of the sample (54%) sustained their injury during participation in organized team sports. A preinjury history of anxiety and depression was reported by 29% and 15%, respectively.

Based on results of correlational analyses (Table 2), history of anxiety, depression, migraines, and gender were included as predictors of PCSI baseline score, PCSI visit 1 score, and days from visit 1 to clearance. History of an LD and ADHD were included as direct predictors of PCSI scores at baseline. Prior concussions, PCSI baseline score, and PCSI score at visit 1 were included as predictors of days from visit 1 to clearance. PCSI baseline score was also included as a predictor of PCSI score at visit 1.

Primary results

The SEM fit the full sample data well, χ² (N = 642; df = 6) = 2.97, p = 0.81, CFI = 1.00, SRMR = 0.01, RMSEA = 0.00 [90% CI: 0.00, 0.03]). Standardized model results are illustrated in Figure 1. Higher PCSI baseline scores were significantly associated with a history of anxiety (β = 0.18, SE = 0.05, p < 0.01), depression (β = 0.22, SE = 0.05, p < 0.01), ADHD (β = 0.09, SE = 0.04, p = 0.01), and headaches (β = 0.24, SE = 0.04, p < 0.01) as well as female gender (β = 0.11, SE = 0.04, p < 0.01). Higher PCSI scores at visit 1 were significantly associated with higher PCSI scores at baseline (β = 0.51, SE = 0.04, p < 0.01), female gender (β = 0.15, SE = 0.03, p < 0.01), and a history of anxiety (β = 0.12, SE = 0.04, p = 0.01). A greater number of days from visit 1 to clearance was associated with lower PCSI baseline scores (β = −0.30, SE = 0.06, p < 0.01), higher PCSI scores at visit 1 (β = 0.64, SE = 0.05, p < 0.01), and a history of migraines (β = 0.11, SE = 0.04, p = 0.02).

Based on results of the mediation analyses (Table 3), there was a significant indirect effect of history of anxiety and depression on PCSI score at visit 1 through PCSI baseline score. Meaning, PCSI scores at baseline fully accounted for the relation between history of depression and PCSI score at visit 1 and only partially accounted for the relation between history of anxiety and PCSI score at visit 1 (e.g., significant direct effect from anxiety to PCSI scores at visit 1). Only history of anxiety (b = 0.08) indirectly contributed to greater days to clearance by means of higher visit 1 PCSI scores.

Factors associated with symptom reporting

Consistent with previous studies,²³ a preinjury history of migraines, anxiety, depression, and female gender was associated with greater endorsement of postconcussive symptoms as well as longer time to clearance based on correlational and post hoc analyses. Given the diversity of this sample in terms of ethnicity, insurance status/type, and mechanism of injury, this study was able to provide a more socioeconomically balanced sample to answer questions posed about the generalizability of these preinjury factors across populations. The majority of the included risk factors for prolonged recovery were primarily associated with increased symptomatology at baseline (e.g., prior to injury). For clinicians, this emphasizes the importance of quantifying “baseline” (rather than “asymptomatic”) in order to determine when a patient has returned to their self-reported preinjury level of functioning. This estimation provides important context for interpreting the severity of postconcussive symptoms. For example, if the patient has a preinjury history of migraines and endorses headaches at baseline, similar endorsement of headaches postinjury would not be considered abnormal. A determination of “back to baseline” should be individually based and informed by injury factors and results of symptom reporting, neurological assessment, balance testing, and neurocognitive screening.

Evidence for the role of prior concussions on recovery from a subsequent concussion has been mixed, but more studies than not have been unable to substantiate the relation.⁷ Though associated with more days to clearance in isolation (e.g., correlational results), prior concussion was not significantly associated with days to clearance above and beyond the influence of other preinjury factors included in the model. In the context of influences from other preinjury factors, history of ADHD, but not an LD, was associated with increases in baseline symptom reporting. The lack of significant association between history of an LD or ADHD with time to clearance is consistent with the vast majority of previous empirical studies.⁷ These results speak to the importance of considering risk factors as a constellation rather than individually to better understand their relative importance for postinjury functioning and recovery.

The role of mental health factors

Rates of self-reported depression (15%) and anxiety (29%) among this sample of patients referred to a specialty clinic were higher than prevalence rates of diagnoses at the national level (6.1% and 10.5% for depression and anxiety, respectively; Ghandour et al.²⁴). However, current study rates are consistent with reports of lifetime prevalence rates of depression (14%) and anxiety (32%) among adolescents ages 13–18 years.²⁵ Previous studies have similarly reported elevated rates of preinjury mental health disorders among a specialty concussion population.²⁶

In the present study, inclusion of mental health factors as primary predictors and simultaneous evaluation with other known risk factors (e.g., one statistical model) allowed for a better understanding of their relative strength as a unique predictor (e.g., above and beyond the influence of other factors). The benefit of the path analysis SEM model allowed for simultaneous consideration of all risk factors without a priori determinations of hierarchical processes. The association between history of anxiety and/or depression on baseline reporting of PCSI symptoms further emphasizes the overlap in symptomatology. For example, fatigue, sadness, feeling slowed down, and difficulty concentrating can all be associated with mental health status without the presence of a head injury.

The previous inclusion of mental health variables in relation to symptom recovery among pediatric populations has primarily been descriptive,¹⁵ which precludes the evaluation of shared variance and relative importance of anxiety and depression. Results of the current study show a direct relation between anxiety and PCSI score at visit 1, above and beyond the influence of anxiety on a higher baseline score. In contrast, depression was significantly and directly related to PCSI scores at baseline but not at visit 1. This difference highlights the unique influence of preinjury anxiety on postconcussive symptom experience above and beyond that of depression. Preinjury history of anxiety had larger total effects across the model than depression, which is consistent with prior work documenting significantly higher rates of preinjury anxiety, but not depression, in a mild TBI population compared with national norms.²⁶ High levels of anxiety may lead to hypervigilance and subsequent overinterpretation and misattribution of benign bodily sensations following concussion (e.g., somatization), which, in turn, elevates report of symptoms and prolongs recovery.

Pathways of influence to clearance

A preinjury history of headaches/migraines demonstrated direct effects on baseline PCSI and was the only preinjury factor with a significant direct effect on days to clearance. Headaches, similar to symptoms of anxiety and depression, are another nonspecific symptom associated with concussion that can be maintained by multiple contributing factors (sleep, stress, predisposition/family history, etc.). These findings were consistent with previous literature describing that history of headaches is a common factor associated with prolonged symptoms.²⁷ Higher PCSI scores at visit 1 were strongly associated with a longer time to clearance, which is also consistent with previous findings that higher acute symptom ratings are associated with longer recovery time.¹² As expected, higher PCSI baseline scores predicted higher PCSI visit 1 scores, but baseline PCSI scores were inversely related to days to clearance. This secondary finding likely reflects that individuals who report few to no symptoms at baseline (e.g., no endorsement of headaches) may take longer to return to their preinjury level of functioning compared with individuals for whom aspects of the postconcussive experience, as captured on the PCSI, are a part of their normative daily experience. Additional investigation and replication of this finding is warranted.

History of both anxiety and depression indirectly contributed to a greater number of days to clearance, emphasizing the importance of considering noninjury factors as integral to concussion management and treatment. Practical implications include routine screening for past and current symptoms of anxiety and depression during clinic visits. If present, this information can influence treatment approach and management strategies, as treatment for depression and/or anxiety can differ from that for concussion alone (e.g., cognitive behavioral therapy for prolonged concussion symptoms^{28, 29}). Neuropsychology involvement in concussion management can help differentiate cognitive symptoms postconcussion from broader symptomatology overlapping with mental health considerations. For patients whose psychological factors are more severe (i.e., new-onset post-traumatic stress following a motor vehicle collision), referral to outpatient mental health services is most likely appropriate. Explaining this association between symptomatology and noninjury factors to patients and their families can require a sensitive approach. Regardless of whether symptoms are thought to be neurological or psychological in origin, feedback should be delivered in a manner that focuses on healthy practices to reduce symptoms and improve overall functioning.

Limitations

In the context of study strengths, including a large, diverse sample in terms of ethnicity, insurance status/type, and mechanism of injury, several limitations must also be acknowledged. Though the current sample represents a strength for all the aforementioned reasons, the convenience sample derived from a specialty clinic may limit the generalizability of results to other populations. Diversity has many meanings, not all of which were well captured in this study's sample (e.g., race; predominantly White sample). Preinjury mental health history was collected as a dichotomous (yes/no) variable based on patient/family report, and further information regarding the age of onset, type, course, severity, and duration was not available. Given the consistently strong support for mental health impacts on concussion recovery in the recent literature,¹⁴ the field would benefit from a more nuanced approach that examines mental health history on a finer scale. A closer examination of the role of mental health intervention services prior to injury and during recovery on symptom reporting and recovery trajectory is also warranted. Further, data collection methods precluded evaluation of causal processes due to an inability to establish temporal precedence (e.g., retrospective report of baseline symptoms were reported in the same visit as current postconcussive symptoms).