Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Am Geriatr Soc. Author manuscript; available in PMC 2017 April 1.
Published in final edited form as:
PMCID: PMC5021441

Health Problems Precede Traumatic Brain Injury in Older Adults

Kristen Dams-O’Connor, Ph.D.,a Laura E Gibbons, Ph.D.,b Alexandra Landau, B.A.,a Eric B Larson, MD, MPH,c and Paul K. Crane, MD, MPHb



To evaluate whether indices of pre-injury health and functioning were associated with risk for incident traumatic brain injury (TBI) with loss of consciousness (LOC), and evaluated health-related factors associated with mortality among those with an incident TBI.


Prospective community cohort study.


Group Health, Seattle Washington.


3,363 individuals aged 65 and older with no self-reported prior TBI with LOC were enrolled and followed every 2 years for an average of 7.5 years (range 0–18 years).


We used Weibull survival models to evaluate baseline and time-varying predictors of incident TBI with LOC, including measures of depression, activities of daily living, cerebrovascular disease, and disease comorbidity.


In an adjusted multivariate model, baseline depression symptoms as measured by CES-D score (hazard ratio (HR) and 95% confidence interval (CI) for 4 points = 1.34 (1.13, 1.58); p<0.05) and baseline impairment in activities of daily living (ADL; HR (95% CI) = 2.37 (1.24, 4.53); p<0.01) were associated with incident TBI. In a model that included time-dependent covariates, cerebrovascular disease at the previous visit (HR (95% CI) = 2.28 (1.37, 3.78); p<0.01), CES-D score the previous visit (HR for 4 points (95% CI) = 1.23 (1.02, 1.49); p<0.05) and baseline impairment in ADL (HR (95% CI) 2.14 (1.11, 4.13); p<0.05) predicted incident TBI. Of factors considered, cerebrovascular disease and ADL impairment were associated with earlier mortality among those with an incident TBI with LOC.


Indices of health, mood, and functional status predict incident TBI with LOC in older adults. These findings may have implications for injury prevention and post-injury clinical management.

Keywords: traumatic brain injury, health, older adults


Traumatic brain injury (TBI) is a leading cause of death and disability. Older adults are more likely than their younger counterparts to sustain TBIs and less likely to survive them. TBI-related hospitalizations and deaths are highest in individuals over the age of 65,1 and TBI rates in older people have exceeded population growth in recent years.2 Older TBI survivors have worse functional outcomes,3 and many experience deterioration in overall health.4 TBI has been called the “silent epidemic”,5 and older adults are the “silent population” within this epidemic.6

Several studies document poor outcomes68 and high costs of care9 for older TBI survivors. Compared to people without a history of TBI, adults over age 55 with TBI report more health problems such as gastrointestinal difficulties, autoimmune conditions, metabolic dysfunction, urologic problems, and musculoskeletal conditions.8 It is possible that the decline seen in many older TBI survivors represents continuation of pre-existing disease processes exacerbated by TBI-related sequelae.

The possibility that accidents leading to TBI in older adults (i.e., falls) may indicate declining health preceding TBI is supported by data demonstrating higher medical costs in the year preceding injury among individuals with mild TBI relative to matched controls.9 Moreover, people who have three or more comorbid medical conditions during hospitalization for TBI are up to four times more likely to die within a year than TBI survivors with few comorbidities.10 In a large case-control study individuals who sustained a mild TBI had more systemic disease hospital admissions in the 15 years preceding injury than individuals who sustained a non-head injury.11 Falls are the most common cause of TBI in older people.1 Acute and chronic health problems are established risk factors for falls.12 Nearly three-quarters of older adults with TBI had at least one chronic medical condition before injury, compared to 28% of younger adults with TBI,13 suggesting that pre-injury health may play a role in injury risk. Fall-related TBI may be uniquely associated with pre-injury health problems; in one study people whose TBI resulted from a fall were more likely to have 3 or more comorbid conditions compared to those whose TBI resulted from a motor vehicle accident.7 The extent to which pre-injury health factors may increase a person’s risk for sustaining a TBI in late life has not been thoroughly investigated.

A better understanding of risk factors for TBI in older adults can inform prevention efforts. We studied whether pre-injury health and function predict incident TBI among older adults. Prior work with data from this study suggested lifetime history of TBI with loss of consciousness (LOC) is associated with up to 4-fold increase in risk of late-life TBI.14 The current investigation focuses on individuals without a history of TBI with LOC. Among older adults who sustain an incident TBI, we will investigate whether indices of pre-injury health and function predict mortality.


Participants were from Adult Changes in Thought (ACT), a longitudinal population-based prospective cohort study of incident dementia and Alzheimer’s disease. Detailed study methods have been published.15 Data for the current analyses were gathered from 4,004 dementia-free individuals aged 65 and older who reported no prior TBI with LOC when joining the ACT study. Dementia status is determined by clinical and neuropsychological evaluation and consensus conference for participants scoring below cutoff values on a cognitive test.15,16 The data freeze included visit dates up to October 3, 2012. Of these, 3,363 individuals completed at least one follow-up visit. Participants were followed every 2 years for an average of 7.5 years (range 1–18 years). Every visit included structured interview that addressed each participant’s history of physical injuries, including specifically whether they had sustained an injury to the head that resulted in LOC: “Have you ever had an injury so severe that you lost consciousness?” If an injury was reported, participants were further queried to determine whether the injury was a head injury (TBI) or other type of injury (e.g., near drowning, electric shock), the duration of unconsciousness, whether medical attention was sought, and the age at injury. Data were collected from participants at every visit on self-reported medical conditions (including cerebrovascular disease and heart disease), self-reported depression symptoms (Centers for Epidemiological Studies-Depression Scale [CES-D]),17 cognitive functioning (as measured by the Cognitive Abilities Screening Instrument [CASI]),16 self-reported alcohol problems, sedentary lifestyle (participants were categorized as sedentary if they reported fewer than three 15-minute periods of exercise per week)18 and physical performance using the Short Physical Performance Battery19 (modified to include a measure of grip strength in addition to walking speed, physical strength, and balance).20 Structured interviews were also conducted to assess activities of daily living (ADLs) using a modified Index of ADL scale.21 Participants’ assessment of their ability to walk around their house, bathe, dress, get out bed or a chair, feed themselves and use a toilet were coded from 0 (no difficulty) to 3 (unable to do the activity). Responses were summed, and those with a total above 2 were considered to have ADL difficulty. Blood samples were drawn and DNA evaluated to determine the number of APOE ε4 alleles. Pharmacy records were used to create an index of medical comorbidity at the time of each study visit (RxRisk).22

Statistical methods

We used Weibull models in Stata 13.1 (StataCorp LP, College Station, TX)23 to evaluate baseline and time-varying predictors of incident TBI among all participants, and of death among people who sustained an incident TBI with LOC. The ACT study collects data on the year in which the TBI was sustained, but not the date. We arbitrarily used June 30, except when June 30 did not fall in the between-visit window in which a TBI was reported. In these cases, the date of TBI was 30 days before the current visit or 30 days after the previous visit, as appropriate. This is a common way to account for interval censoring.24 We censored seven participants with a second TBI with LOC at the time of the second TBI.


Demographic information for 3,363 individuals with baseline TBI data and ≥ 1 follow-up are presented in Table 1. Higher proportions of those without follow-up were male and older with more education (p<0.05). There were 76 people (2.3%) who sustained an incident TBI with LOC over 25,304 person-years of follow-up.

Table 1
Demographic and baseline clinical information, mean (±SD) or N (%) (n=3,363)

In models controlling for age, education, and sex, baseline predictors of incident TBI with LOC included self-reported cerebrovascular disease, CES-D measured depression, difficulty with ADLs defined by a total ADL score >2, and pharmacy-derived chronic disease comorbidity (RxRisk) (see Table 2). Cognitive functioning, sedentary lifestyle, physical performance, self-reported alcohol problems, and presence of one or more APOE ε4 alleles were not associated with incident TBI with LOC (p>0.05). In a multivariate model controlling for age, education, and sex, RxRisk and cerebrovascular disease at baseline were not associated with incident TBI with LOC, but CES-D and higher ADL scores were (Table 3).

Table 2
Individual baseline predictors of incident traumatic brain injury with loss of consciousness, each controlled for age, education, and sex
Table 3
Multivariate predictors of incident traumatic brain injury with loss of consciousness, controlling for age, education, and sex

Several time-varying covariates were associated with incident TBI with LOC. In models using data from the study visit preceding incident TBI with LOC, predictors included better physical performance (HR(95% CI) 1.08 (1.01, 1.16), p=0.03), cerebrovascular disease (HR (95% CI) 2.50 (1.52, 4.09), p<0.01), depression (HR (95% CI) per 4 CES-D points of 1.34 (1.13, 1.58); p< 0.01). When we dichotomized ADLs at 2 points, it was not associated with incident TBI with LOC, but the continuous ADL score was, (HR (95% CI) per 4 points, 1.31 (1.01, 1.71); p=0.04). In the best multivariate model, independent predictors of incident TBI with LOC included previous visit cerebrovascular disease, previous visit CES-D, and baseline limited ADLs (see Table 3). Sedentary lifestyle, cognitive functioning, APOEε4 genotype, and alcohol problems were not associated with incident TBI with LOC.

The 76 individuals who sustained an incident TBI with LOC accrued an additional 419 person-years of follow-up, during which we observed 36 deaths. Adjusting for age, education, and sex, ADL score above 2 (HR (95% CI) 3.71 (1.53, 8.95); p< 0.01) and cerebrovascular disease (HR (95% CI) 2.40 (1.21, 4.75); p=0.01) at the post-TBI visit predicted mortality in a multivariate model.


In this prospective cohort study of older adults without previous TBI with LOC, we found that lower ADL scores, cerebrovascular disease, and higher depression scores predicted incident TBI with LOC. We also found that disease comorbidity predicted incident TBI with LOC. Among those with an incident TBI with LOC, participants with cerebrovascular disease were at greater risk for death, suggesting comorbidity may also impact post-TBI mortality. Previous research has reported worse health and functioning among older adults with TBI, leading to a conceptualization of TBI as a triggering event for chronic disease processes. Our findings indicate health and functional problems may also precede TBI, suggesting complex relationships wherein TBIs may be superimposed on already failing health.

TBI risk rises in late life.7 Although most TBIs in individuals over 65 years of age are mild in severity, the consequences may not be mild. Nearly a third of hospitalized survivors are discharged with moderate or severe disability, and 12% die in the hospital.7 The rate of TBI among older US adults is increasing. A 10-year surveillance study of older people in Oklahoma demonstrated that the TBI rate increased 79% while the case-fatality rate decreased from 32% to 18%. TBIs resulting from falls increased 126%, while those resulting from motor vehicle accidents increased 18% over this time period.25 Another study using Pennsylvania state trauma center data demonstrated incidence of TBI (calculated using U.S. Census estimates) among older people doubled between from 1992–2009, and the oldest age groups had the highest rates and poorest outcomes.26 These findings suggest a growing number of older adults sustain and survive TBIs and live with TBI-related disability, underscoring the need for improved primary prevention efforts.

Factors associated with TBI risk may vary across the lifespan.27,28 The current study is the first to our knowledge to evaluate factors associated with incident TBI in older adults in a population-based cohort study.

Previous studies have shown that older adults with TBI face important challenges to return to full functioning. Our findings suggest that TBI effects may be superimposed on pre-existing cognitive and health problems. Our findings do not contradict studies documenting post-TBI health decline; rather, they suggest a more complicated set of relationships. TBI-related cognitive impairment and functional limitations are superimposed on existing age-related changes which together may pose important challenges to independent health management. Accordingly, older adults with TBI may benefit from chronic disease management models. Identification of older adult TBI risk factors can lead to improved prevention. Prior work in fall prevention suggests that targeted interventions for people with known risk factors can be effective.29,30 Further research on TBI risks among older adults is warranted, as interventions to address specific factors may reduce risk of serious injuries. While fall prevention is already a central focus in geriatric care, our findings underscore the importance of implementing empirically supported prevention programs as a means of reducing risk for incident TBI and its cognitive and health consequences.

Our study has limitations that warrant consideration. Only TBI with LOC was recorded; TBIs not resulting in LOC (i.e., very mild TBIs) were not assessed. Milder TBIs (i.e., those that result in alteration of mental status) are common in older adults and can have important consequences, but were not included here. Cause of TBI was not recorded. Data on gait instability, history of falls, and Parkinsonian features were not routinely collected, and these factors may be associated with fall (and therefore TBI) risk. The results of this study are best generalized to the population we studied: non-demented older individuals with health insurance and with no history of a TBI with LOC. Unmeasured confounders and residual confounding could be present as they could be in any observational study. Methodological strengths of the study should also be considered. Incident TBI with LOC was queried in a structured interview format referencing injuries with LOC, maximizing exposure recall. Although recall of remote injuries may be challenging, the study design ensures that people did not have dementia at the time they were asked about TBI with LOC. This prospective population-based study includes more than 25,000 person years of follow-up data, with up to 18 years of follow-up for many participants.

The current findings have implications for prevention of TBI and its consequences in older adults. Prevention efforts should target individuals in this high-risk group who have specific and known TBI risk factors; these efforts may include therapy for cognitive problems along with empirically supported fall prevention programs. Tertiary prevention of the consequences of TBI in this group could involve heightened medical management and/or TBI-specific chronic disease management models.


In this prospective cohort study, we found several factors associated with incident TBI with LOC, including cerebrovascular disease, depression, inability to perform ADLs, and medical comorbidity. Among those who sustained a TBI with LOC, cerebrovascular disease and ADL problems were associated with higher mortality risk. These results underscore the need to study interventions designed to prevent TBI with LOC in this high risk population.


This project was supported by Grant Numbers #U01 AG006781 (Larson, NIH-NIA), #K01HD074651-01A1 (Dams-O’Connor, NIH-NICHD).

Dams-O’Connor: Employed by Icahn School of Medicine at Mount Sinai, Principal Investigator of Grant #K01HD074651-01A1 (NIH-NICHD), Consultant to the Brain Injury Association of America and Craig Hospital, and provides clinical Expert Testimony in forensic cases.

Gibbons: Employed by the University of Washington.

Landau: Employed by Icahn School of Medicine at Mount Sinai.

Larson: Employed by Group Health and receive grants from NIH. Royalties from UpToDate

Crane: Employed by the University of Washington and has received grants from the NIH, the Paul G. Allen Family Foundation, and the Alzheimer’s Association.


Author Contributions:

Dams-O’Connor: study concept and design, analysis and interpretation of data, and preparation of manuscript. Gibbons: analysis and interpretation of data, and preparation of manuscript. Landau: study concept and design, and preparation of manuscript. Larson: acquisition of subjects and/or data, analysis and interpretation of data, and preparation of manuscript. Crane: study concept and design, acquisition of subjects and/or data, analysis and interpretation of data, and preparation of manuscript. All authors: Revision and approval of final manuscript.

Sponsor’s Role: None.


1. Faul M, Xu L, Wald MM, et al. Traumatic brain injury in the United States: Emergency department visits, hospitalizations, and deaths 2002–2006. Atlanta: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2010.
2. Dams-O’Connor K, Cuthbert JP, Whyte J, et al. Traumatic brain injury among older adults at level I and II trauma centers. J Neurotrauma. 2013;30:2001–2013. [PMC free article] [PubMed]
3. Marquez de la Plata CD, Hart T, Hammond FM, et al. Impact of age on long-term recovery from traumatic brain injury. Arch Phys Med Rehabil. 2008;89:896–903. [PMC free article] [PubMed]
4. Masel BE, DeWitt DS. Traumatic brain injury: A disease process, not an event. J Neurotrauma. 2010;27:1529–1540. [PubMed]
5. Traumatic Brain Injury in the United States: A Report to Congress. Atlanta, GA: Centers for Disease Control and Prevention; 1999.
6. Thompson HJ, McCormick WC, Kagan SH. Traumatic brain injury in older adults: Epidemiology, outcomes, and future implications. J Am Geriatr Soc. 2006;54:1590–1595. [PMC free article] [PubMed]
7. Coronado VG, Thomas KE, Sattin RW, et al. The CDC traumatic brain injury surveillance system: Characteristics of persons aged 65 years and older hospitalized with a TBI. J Head Trauma Rehabil. 2005;20:215–228. [PubMed]
8. Breed ST, Flanagan SR, Watson KR. The relationship between age and the self-report of health symptoms in persons with traumatic brain injury. Arch Phys Med Rehabil. 2004;85:S61–67. [PubMed]
9. Leibson CL, Brown AW, Hall Long K, et al. Medical care costs associated with traumatic brain injury over the full spectrum of disease: A controlled population-based study. J Neurotrauma. 2012;29:2038–2049. [PMC free article] [PubMed]
10. Selassie AW, McCarthy ML, Ferguson PL, et al. Risk of posthospitalization mortality among persons with traumatic brain injury, South Carolina 1999–2001. J Head Trauma Rehabil. 2005;20:257–269. [PubMed]
11. McMillan TM, Weir CJ, Wainman-Lefley J. Mortality and morbidity 15 years after hospital admission with mild head injury: A prospective case-controlled population study. J Neurol Neurosurg Psychiatry. 2014;85:1214–1220. [PubMed]
12. Shumway-Cook A, Ciol MA, Hoffman J, et al. Falls in the Medicare population: Incidence, associated factors, and impact on health care. Phys Ther. 2009;89:324–332. [PubMed]
13. Mosenthal AC, Livingston DH, Lavery RF, et al. The effect of age on functional outcome in mild traumatic brain injury: 6-month report of a prospective multicenter trial. J Trauma. 2004;56:1042–1048. [PubMed]
14. Dams-O’Connor K, Gibbons LE, Bowen JD, et al. Risk for late-life re-injury, dementia and death among individuals with traumatic brain injury: A population-based study. J Neurol Neurosurg Psychiatry. 2013;84:177–182. [PMC free article] [PubMed]
15. Kukull WA, Higdon R, Bowen JD, et al. Dementia and Alzheimer disease incidence: A prospective cohort study. Arch Neurol. 2002;59:1737–1746. [PubMed]
16. Teng EL, Hasegawa K, Homma A, et al. The Cognitive Abilities Screening Instrument (CASI): A practical test for cross-cultural epidemiological studies of dementia. Int Psychogeriatr. 1994;6:45–58. [PubMed]
17. Eaton WW, Muntaner C, Smith C, et al. Center for Epidemiologic Studies Depression Scale: Review and revision (CESD and CESD-R) In: Maruish ME, editor. The use of psychological testing for treatment planning and outcomes assessment. 3. New Jersey: Lawrence Erlbaum Associates; 2004. pp. 363–377.
18. Gray SL, Anderson ML, Hubbard RA, et al. Frailty and incident dementia. J Gerontol A Biol Sci Med Sci. 2013;68:1083–1090. [PMC free article] [PubMed]
19. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994;49:M85–94. [PubMed]
20. Wang L, van Belle G, Kukull WB, et al. Predictors of functional change: A longitudinal study of nondemented people aged 65 and older. J Am Geriatr Soc. 2002;50:1525–1534. [PubMed]
21. Katz S, Ford AB, Moskowitz RW, et al. Studies of Illness in the Aged. The Index of ADL: A Standardized Measure of Biological and Psychosocial Function. JAMA. 1963;185:914–919. [PubMed]
22. Fishman PA, Goodman MJ, Hornbrook MC, et al. Risk adjustment using automated ambulatory pharmacy data: The RxRisk model. Med Care. 2003;41:84–99. [PubMed]
23. StataCorp. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP; 2013.
24. Lindsey JC, Ryan LM. Tutorial in biostatistics methods for interval-censored data. Stat Med. 1998;17:219–238. [PubMed]
25. The epidemiology of severe traumatic brain injury among persons 65 years of age and older in Oklahoma, 1992–2003. Brain Inj. 2007;21:691–699. [PubMed]
26. Ramanathan DM, McWilliams N, Schatz P, et al. Epidemiological shifts in elderly traumatic brain injury: 18-year trends in Pennsylvania. J Neurotrauma. 2012;29:1371–1378. [PubMed]
27. Rockhill CM, Fann JR, Fan MY, et al. Healthcare costs associated with mild traumatic brain injury and psychological distress in children and adolescents. Brain Inj. 2010;24:1051–1060. [PubMed]
28. Nordström A, Edin BB, Lindström S, et al. Cognitive function and other risk factors for mild traumatic brain injury in young men: Nationwide cohort study. BMJ. 2013;346:f723. [PubMed]
29. Buchner DM, Larson EB. Falls and fractures in patients with Alzheimer-type dementia. JAMA. 1987;257:1492–1495. [PubMed]
30. Koepsell TD, Wolf ME, Buchner DM, et al. Footwear style and risk of falls in older adults. J Am Geriatr Soc. 2004;52:1495–1501. [PubMed]