This article was downloaded by: [Colorado College] On: 06 March 2015, At: 13:08 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Clinical Neuropsychologist Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ntcn20
Normative Data for the Neurobehavioral Symptom Inventory (NSI) and PostConcussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples ab
a
Jason R. Soble , Marc A. Silva , Rodney D. Vanderploeg ab
Curtiss , Heather G. Belanger Scott
abcde
, Alison J. Donnell
ae
abcde
, Glenn
& Steven G.
cefg
a
Mental Health and Behavioral Sciences Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA b
Click for updates
Department of Psychiatry & Behavioral Neurosciences, University of South Florida, Tampa, FL, USA c
Health Services Research and Development (HSR&D)/ Rehabilitation Research and Development (RR&D) Center of Excellence: Maximizing Rehabilitation Outcomes, James A. Haley Veterans’ Hospital, Tampa, FL, USA d
Department of Psychology, University of South Florida, Tampa, FL, USA e
Defense and Veterans Brain Injury Center, Tampa, FL, USA
f
Physical Medicine and Rehabilitation Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA g
Department of Internal Medicine, University of South Florida, Tampa, FL, USA Published online: 14 Mar 2014.
To cite this article: Jason R. Soble, Marc A. Silva, Rodney D. Vanderploeg, Glenn Curtiss, Heather G. Belanger, Alison J. Donnell & Steven G. Scott (2014) Normative Data for the Neurobehavioral Symptom Inventory (NSI) and Post-Concussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples, The Clinical Neuropsychologist, 28:4, 614-632, DOI: 10.1080/13854046.2014.894576 To link to this article: http://dx.doi.org/10.1080/13854046.2014.894576
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [Colorado College] at 13:08 06 March 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
The Clinical Neuropsychologist, 2014 Vol. 28, No. 4, 614–632, http://dx.doi.org/10.1080/13854046.2014.894576
Normative Data for the Neurobehavioral Symptom Inventory (NSI) and Post-Concussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples Jason R. Soble1,2, Marc A. Silva1, Rodney D. Vanderploeg1,2,3,4,5, Glenn Curtiss1,2, Heather G. Belanger1,2,3,4,5, Alison J. Donnell1,5, and Steven G. Scott3,5,6,7
Downloaded by [Colorado College] at 13:08 06 March 2015
1
Mental Health and Behavioral Sciences Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA 2 Department of Psychiatry & Behavioral Neurosciences, University of South Florida, Tampa, FL, USA 3 Health Services Research and Development (HSR&D)/Rehabilitation Research and Development (RR&D) Center of Excellence: Maximizing Rehabilitation Outcomes, James A. Haley Veterans’ Hospital, Tampa, FL, USA 4 Department of Psychology, University of South Florida, Tampa, FL, USA 5 Defense and Veterans Brain Injury Center, Tampa, FL, USA 6 Physical Medicine and Rehabilitation Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA 7 Department of Internal Medicine, University of South Florida, Tampa, FL, USA The Neurobehavioral Symptom Inventory (NSI) is a self-report measure of symptoms commonly associated with Post-Concussion Syndrome (PCS) that may emerge after mild traumatic brain injury (mTBI). Despite frequent clinical use, no NSI norms have been developed. Thus, the main objective of this study was to establish NSI normative data using the four NSI factors (i.e., vestibular, somatic, cognitive, and affective) identified by Vanderploeg, Silva, et al. (2014) among nonclinical epidemiological samples of deployed and non-deployed Florida National Guard members as well as a reference sample of Guard members with combat-related mTBI. In addition, NSI subscale profile patterns were compared across four distinct subgroups (i.e., non-deployed-nonclinical, deployed-nonclinical, deployed-mTBI, and deployed-PTSD). The deployed-nonclinical group endorsed greater PCS symptom severity than the non-deployed group, and the mTBI group uniformly endorsed more symptoms than both nonclinical groups. However, the PTSD group endorsed higher symptom severity relative to the other three subgroups. As such, this highlights the non-specificity of PCS symptoms and suggests that PTSD is associated with higher symptom endorsement than mTBI. Keywords: Neurobehavioral Symptom Inventory; Normative data; Post-Concussion Syndrome; Traumatic brain injury.
INTRODUCTION Post-Concussion Syndrome (PCS) refers to the collection of cognitive, emotional, and physical symptoms that sometimes emerge following mild traumatic brain injury (mTBI). PCS diagnostic criteria are formalized in the International Classification of Address correspondence to: Rodney D. Vanderploeg, James A. Haley Veterans’ Hospital Psychology Service (116B), 13000 Bruce B. Downs Blvd. Tampa, FL 33612, USA. E-mail: [email protected] (Received 17 October 2013; accepted 11 February 2014)
© 2014 Taylor & Francis
Downloaded by [Colorado College] at 13:08 06 March 2015
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
615
Diseases-Tenth Edition (ICD-10; World Health Organization, 1992) and Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV; American Psychiatric Association [APA], 1994). However, PCS remains a controversial disorder (Granacher, 2008), and as an entity is absent in the DSM-5 (APA, 2013). Despite its lack of widespread acceptance as a valid clinical syndrome, health care providers continually are faced with evaluating and treating these so-called “post-concussion” cognitive, emotional, and physical symptoms that are reported by some patients with histories of TBI. TBI itself is not uncommon, with an annual incidence rate of 1.7 million in the U.S. (Faul, Xu, Wald, & Coronado, 2010), approximately 75% of which are classified as mild (Centers for Disease Control and Prevention, 2003). However, while the majority of individuals who sustain a mTBI recover from acute cognitive sequelae (Belanger & Vanderploeg, 2005; Rohling et al., 2011), a subset of individuals endorse PCS symptoms for months and/or years after injury (Dikmen, Machamer, Fann, & Temkin, 2010; Ruff, Camenzuli, & Mueller, 1996; Vanderploeg, Belanger, & Curtiss, 2009; Vanderploeg, Curtiss, Luis, & Salazar, 2007). Different diagnostic criteria exist between the ICD-10 (World Health Organization, 1992) and DSM-IV (APA, 1994). For instance, the ICD-10 criteria include headache, dizziness, fatigue, irritability, sleep problems, memory or concentration difficulty, and intolerance of stress, emotion, or alcohol. In contrast, the DSM-IV criteria contain fatigue, sleep problems, dizziness, headaches, irritability or aggression, apathy, personality changes, and anxiety, depression, or affective lability that persist for at least 3 months. While several common symptoms (e.g., headache, sleep problems) are included in both criteria, DSM-IV does not address cognitive problems, whereas ICD-10 speaks little to potential affective and personality changes. Thus, PCS prevalence rates can vary based on the specific diagnostic criteria used. In one prospective study of 178 adults with mild-moderate TBI, Boake et al. (2005) found that 64% met the ICD-10 PCS criteria, whereas only 11% met the DSM-IV criteria. Beyond yielding discrepant prevalence rates, these varying criteria also pose a challenge for accurately measuring PCS. Several PCS symptom measures have been developed, including the Post-Concussion Syndrome Checklist (PCSC; Gouvier, Cubic, Jones, Bentley, & Cutlip, 1992), Beaumont Post-Concussional Index (BPCI; Trahan, Ross, & Trahan, 1997), and Rivermead Post-Concussion Symptoms Questionnaire (RPQ; King, Crawford, Wenden, Moss, & Wade, 1995). Of particular interest to the current study is the Neurobehavioral Symptom Inventory (NSI; Cicerone & Kalmar, 1995), originally published as the Post MTBI Symptom Checklist (Cicerone & Kalmar, 1995; Cicerone et al., 1996), and based in part on symptoms included in a structured interview used in a multicenter mTBI study (Levin et al., 1987). The NSI contains 22 common symptoms that patients rate on a 5-point scale based on how much each symptom has disturbed them during the past two weeks, month, or since time of injury, depending on the version used. Surprisingly, no comprehensive normative data have yet been established for this measure despite the NSI’s routine use by the Departments of Defense and Veterans Affairs (VA), including the VA’s TBI Comprehensive Evaluation, as well as it being identified as a supplemental common outcome measure by the multiagency TBI Outcomes Workgroup (Wilde et al., 2010). While past studies that examined the factor structure of the NSI (i.e., Benge, Pastorek, & Thompson, 2009; Cicerone & Kalmar, 1995; Caplan et al., 2010; Meterko et al., 2012) have produced inconsistent results, overall findings generally indicate that
Downloaded by [Colorado College] at 13:08 06 March 2015
616
JASON R. SOBLE ET AL.
PCS is a multidimensional construct. For instance, Cicerone and Kaplan (1995) identified four clusters and five items that did not fit any cluster. Benge et al. (2009) identified a four-factor solution (sensory, affective, a headache, nausea, and visual disturbance factor, and an unclear factor) and a six-factor solution after controlling for Post-Traumatic Stress Disorder (PTSD) (vestibular, cognitive, affective, and hearingrelated, plus two unclear factors). Caplan et al. (2010) found a two-factor solution (somatic/sensory and psychological) and a three-factor solution (somatic/sensory, affective, and cognitive). Finally, based on exploratory factor analysis, Meterko et al. (2012) reported a three-factor solution (somatic/sensory, cognitive, and affective) and a fourfactor solution (somatic, vestibular, cognitive, and affective). Given these inconsistent findings, a more recent study (Vanderploeg, Silva, et al., 2014) used structural equation modeling to compare the existing NSI models. They found that Meterko et al.’s (2012) four-factor model (vestibular, somatic, cognitive, and affective) minus two items (i.e., hearing problems and appetite disturbance) provided the best overall fit for PCS symptoms in both a clinical and nonclinical sample). This multidimensional, four-factor model also may allow for improved discrimination among different clinical subgroups. Beyond its multidimensional nature, the construct of PCS has been called into question. Many of the core symptoms are nonspecific to mTBI and can be found among healthy individuals with no history of TBI (Chan, 2001; Iverson & Lange, 2003; Vanderploeg et al., 2007, 2009). Not surprisingly, past studies comparing NSI scores of those with mTBI to nonclinical samples found that those with mTBI endorsed higher levels of PCS symptoms (King et al., 2012). However, elevated levels of PCS symptoms also are commonly found in the presence of a host of other medical conditions, such as chronic pain (Iverson & McCracken, 1997) and non-TBI physical trauma (Meares et al., 2008), as well as psychiatric disorders including depression, PTSD, and somatization (Belanger, Kretzmer, Vanderploeg, & French, 2010; Bryant, 2001; Donnell, Kim, Silva, & Vanderploeg, 2012). In fact, King et al. (2012) found higher correlations between PCS symptoms and measures of affective psychopathology than TBI. Donnell et al. (2012) reported that PCS symptoms were more frequently endorsed by veterans diagnosed with psychiatric disorders (i.e., PTSD, generalized anxiety, major depression, somatization) than those with mTBI. Given the multidimensional and nonspecific nature of PCS symptoms, the clinical utility of the NSI is limited without adequate normative data and established symptom profile patterns to guide interpretation.
STUDY RATIONALE AND GOALS No standard yet exists to help clinicians interpret NSI scores. Determining the clinical significance of point-in-time raw scores remains a challenge without knowledge of performance by appropriate references groups. To address this limitation, the primary objective of this study was to establish normative data for the NSI and its four factors as identified by Vanderploeg, Silva, et al. (2014). Moreover, given the NSI’s routine use as part of the VA’s TBI Comprehensive Evaluation, a secondary objective was to also develop reference-group normative data using an epidemiological, rather than clinically presenting, cohort of military personnel with combat-related mTBI, thus permitting more sophisticated interpretation of obtained NSI scores. It is our hope that, by presenting normative data for two nonclinical samples (i.e., non-deployed and deployed) as
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
617
well as a mTBI reference sample, health care providers will more effectively utilize the NSI in clinical practice. Beyond developing normative data, a final objective of this study was to examine NSI score patterns among different subgroups (defined in “Method” below).
METHOD
Downloaded by [Colorado College] at 13:08 06 March 2015
Participants Study data came from the 2009–2010 Florida National Guard survey. The study sample is described elsewhere (Vanderploeg et al., 2012). In brief, the study participants were an unselected, representative sample of the Florida National Guard who completed an anonymous online survey that included the NSI. Of 10,400 letters mailed out inviting members to participate, 4005 individuals responded, yielding a response rate of 41.3%. This current sample consisted of 3098 participants who provided usable data and fully completed the survey (we excluded 423 who did not fully complete the survey, 371 who completed the survey more than once, and 113 who gave inconsistent or implausible responses). Of these 3098 responders, 1443 reported at least one deployment and 1655 reported no deployments. Guard members who deployed returned from deployment up to 95 months prior to survey (M = 31.8; SD = 24.4). Of the 3098, the following subgroups were identified (see Table 1 for demographic characteristics of subgroups). Non-deployed-nonclinical (n = 1453). Among the 53% who did not deploy, 175 reported a history of TBI, 32 screened positive for PTSD, 24 screened positive for other anxiety disorders, and 16 screened positive for depressive disorders (screening procedures are described below in the “Measures” section). These individuals were excluded (categories not mutually exclusive), resulting in 1453 participants who comprised the non-deployed-nonclinical reference group. Deployed-nonclinical (n = 1064). Among the 47% who deployed, 392 reported a mTBI prior to deployment, 144 reported a mTBI during deployment, 100 screened positive for PTSD, 61 screened positive for other anxiety disorders, and 47 screened positive for depressive disorders. These individuals were excluded (categories not mutually exclusive), resulting in 1064 participants for who comprised the deployed-nonclinical group. Deployed-mTBI (n = 108). Deployment-related mTBI was determined by endorsement of a motor vehicle accident, blast exposure, or any other event resulting in loss of consciousness, blacking out, or memory gaps. Those who reported a head injury as indicated above were then asked about the duration of any memory gaps (most were less than an hour and all were less than a day). Participants who endorsed being dazed and confused without memory gaps were excluded from the mTBI group because combat events such as unexpected ambushes, firefights, or nearby blasts may be experienced as confusing, disorienting, or “dazing” although not necessarily indicative of brain injury. While medical documentation of mTBI at time of injury is ideal, medical records from the acute period may be unavailable (if medical attention is sought at all). Consequently, self-report is common in studies of mTBI in the chronic phase. Among the 47% who deployed, 144 endorsed sustaining mTBI during deployment. Of these, 36 screened positive for PTSD and were excluded, as a focus of this study was to
21.2 78.8 16.7 20.2 56.1 7.9 33.4 44.9 21.6 2.3(2.3) 1.3(2.4) 1.2(2.5) 19.0(4.7)
17.0 83.0 14.8
17.3 61.9 6.1
26.3 47.5 26.1
2.6(2.5) 2.7(4.2) 2.4(4.3) 22.6(10.9)
Non-deployed –nonclinical n = 1,453
2.7(2.5) 2.6(3.4) 2.1(3.2) 21.7(6.9)
20.4 49.3 30.3
15.2 65.2 4.5
13.3 86.7 15.0
Deployed-nonclinical n = 1,064
3.3(2.7) 4.7(4.5) 4.2(4.2) 28.7(10.6)
17.6 51.9 30.6
16.7 67.6 3.7
13.0 87.0 12.0
Deployed-mTBI n = 108
4.6(3.2) 14.7(4.8) 15.3(6.3) 63.3(10.2)
28.8 48.1 23.1
19.2 57.7 3.8
21.2 78.8 19.2
DeployedPTSD n = 52
Race/Ethnicity “Other” included participants who identified as Native American, Alaskan Native, Asian, Pacific Islander, or Biracial. Percentages may not sum to 100 due to rounding. AUDIT-C = Alcohol Use Disorder Identification Test-Screen; GAD7 = Generalized Anxiety Disorder Questionnaire; PHQ9 = Patient Health Questionnaire, Depression Module; PCL-C = PTSD Checklist, Civilian Version.
Gender Female (%) Male (%) Race/Ethnicity Black (%) Hispanic(%) White (%) Other (%) Education (years) ≤ 12 (%) 13-15 (%) ≥ 16 (%) Clinical Symptoms AUDIT-C, M(SD) GAD7, M(SD) PHQ9, M(SD) PCL-C, M(SD)
Total sample N = 3,098
Table 1. Sociodemographic and selected clinical characteristics
Downloaded by [Colorado College] at 13:08 06 March 2015
618 JASON R. SOBLE ET AL.
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
619
Downloaded by [Colorado College] at 13:08 06 March 2015
compare participants with PTSD (but no mTBI) with participants with mTBI (but no PTSD). We chose not to exclude the 38 participants who screened positive for depressive and other anxiety disorders, because these symptoms are not uncommon after mTBI (e.g., Dikmen et al., 2010) and are included in DSM-IV PCS symptom criteria (APA, 1994). The final Deployed-mTBI group was composed of 108 participants. Deployed-PTSD (n = 52). Among the 47% who deployed, we excluded 190 who endorsed a predeployment mTBI and 144 who endorsed a deployment mTBI. We chose not to exclude the 36 who screened positive for depressive and other anxiety disorders given that PTSD and depression are highly comorbid as well as findings suggestive of a single underlying symptom dimension between the disorders (e.g., Elhai et al., 2011). The final Deployed-PTSD group was composed of 52 participants. Measures Neurobehavioral Symptom Inventory (NSI). The NSI (Cicerone & Kalmar, 1995) is a 22-item self-report questionnaire designed to assess PCS symptom severity. Individuals rated the severity of common PCS symptoms in terms of “how much they have disturbed you over the past month” on a scale ranging from 0 (none) to 4 (very severe). The NSI was found to have good internal consistency (King et al., 2012; Silva, Barwick, Kretzmer, Vanderploeg, & Belanger, 2013) and 7-day test–retest stability (Silva et al., 2013). In this study we specifically examined the NSI vestibular, somatic, cognitive, and affective factors based on the four-factor, 20-item NSI model described by Vanderploeg, Silva, et al. (2014). This model excludes two items (i.e., hearing problems and appetite disturbance) due to poor fit. NSI-20 scores range from 0 to 80. For the subscales, vestibular scores range from 0 to 12, somatic scores range from 0 to 28, cognitive scores range from 0 to 16, and affective scores range from 0 to 24. Generalized Anxiety Disorder Questionnaire (GAD-7). The GAD-7 (Spitzer, Kroenke, Williams, & Lowe, 2006) is a questionnaire designed to screen for Generalized Anxiety Disorder based on DSM-IV criteria. Scores range from 0 to 21, with higher scores indicating greater symptom endorsement. In this study, participants were classified as having a probable anxiety disorder if they endorsed: (a) nervousness or anxiety more than half the time; (b) three of six other anxiety symptoms more than half the time; and (c) impairment in work, home, or interpersonal functioning at the “very difficult” level or higher (Spitzer et al., 2006). Patient Health Questionnaire Depression Module (PHQ-9). The PHQ-9 (Spitzer, Kroenke, & Williams, 1999; see also Kroenke & Spitzer, 2002) is a questionnaire designed to screen for Major Depressive Disorder based on DSM-IV criteria. Scores range from 0 to 27, with higher scores indicating greater symptom endorsement. In this study, participants were classified as having a probable major depressive disorder, if they met DSM-IV symptom criteria and reported impairment in work, home, or interpersonal functioning at the “very difficult” level or higher (Spitzer et al., 1999). PTSD Checklist-Civilian Version (PCL-C). The PCL (Weathers, Litz, Herman, Huska, & Keane, 1993) is a 17-item self-report measure of DSM-IV PTSD symptoms in which participants rate items on a scale ranging from 1 (not at all) to 5 (extremely). Scores can range from 17 to 85. We operationalized PTSD as a PCL score of greater
620
JASON R. SOBLE ET AL.
Downloaded by [Colorado College] at 13:08 06 March 2015
than 50 and meeting DSM-IV symptom criteria for PTSD by endorsing as moderate (3) or higher at least one re-experiencing symptom, at least three avoidance symptoms, and at least two hyperarousal symptoms, as per recommended guidelines (National Center for PTSD, 2012). Consistent with DSM-IV criteria for PTSD, participants were also required to report that their symptoms caused impairment in work, home, or interpersonal functioning at the “very difficult” level or higher. Alcohol Use Disorders Identification Test-Consumption Questions screening test (AUDIT-C). The AUDIT-C (Bush, Kivlahan, McDonell, Fihn, & Bradley, 1998) is a three-item measure designed to screen for heavy drinking and alcohol use disorders. Possible scores range from 0 to 12 with higher scores associated with greater risk of problematic drinking. In this study, AUDIT-C scores are presented for descriptive purposes, but are not used as exclusion criteria.
Procedures After establishing the above groups, subscale scores for each of the four factors identified by Vanderploeg, Silva, et al. (2014) were formed by summing the raw scores on the items deemed to load on the factor. Two total scores were also formed: the NSI22, which was calculated by summing raw scores on all items comprising the NSI and the NSI-20, which was calculated by summing raw scores on all items except hearing problems and appetite problems.
Data analyses NSI reliability analyses were computed for the non-deployed-nonclinical normative sample (n = 1453) and for the deployed nonclinical sample (n = 1064). Normative data were then derived for both of these groups. Reference comparison data were also derived for the deployed-mTBI sample (n = 108). Frequency distributions were examined and, due to skew and kurtosis of the distributions, normative data were derived for various percentile ranges. Last, multivariate analysis of variance (MANOVA) was used to investigate whether the four NSI scales were useful in profiling difference across subgroups. Scale mean scores were used to correct for differences in number of items comprising each scale.
RESULTS Descriptive statistics Indices of central tendency and scale score distributions are presented in Tables 2 and 3. These tables show that the NSI total, subscale, and item level scores are positively skewed with means and medians near zero for most NSI items. Therefore, an attempt to transform the data into a normal distribution would not be appropriate in this case. Rather, normative tables should be presented in a categorical manner.
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
621
Downloaded by [Colorado College] at 13:08 06 March 2015
Table 2. Neurobehavioral Symptom Inventory (NSI) descriptive statistics and reliability indices, Florida National Guard non-deployed nonclinical sample (n = 1453)
M(SD)
Percentile ranks 25/50/75
Range
NSI-20 Vestibular 1-Dizzy 2-Balance 3-Coordination
0.2(0.8) 0.1(0.4) 0.1(0.3) 0.1(0.3)
0/0/0 0/0/0 0/0/0 0/0/0
0–7 0–3 0–3 0–2
α = .74 .59 .65 .51
.48 .52 .51
NSI-20 Somatic 4-Headache 5-Nausea 6-Vision Problems 7-Light Sensitivity 9-Noise Sensitivity 10-Numbness/Tingling 11-Taste/Smell Changes
0.8(1.7) 0.3(0.7) 0.1(0.3) 0.1(0.4) 0.1(0.4) 0.1(0.3) 0.1(0.5) 0.0(0.2)
0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–15 0–4 0–4 0–4 0–3 0–3 0–4 0–4
α = .68 .50 .49 .39 .44 .35 .40 .38
.49 .40 .41 .41 .41 .42 .37
NSI-20 Cognitive 13-Poor Concentration 14-Forgetfulness 15-Diff Decision 16-Slow Thinking NSI-20 Affective 17-Fatigue 18-Sleep Disturbance 19-Anxious/Tired 20-Sad/Depressed 21-Irritable 22-Frustration Intolerance
0.5(1.5) 0.2(0.5) 0.2(0.5) 0.1(0.3) 0.1(0.4) 1.2(2.5) 0.2(0.5) 0.3(0.7) 0.2(0.6) 0.2(0.5) 0.2(0.5) 0.2(0.5)
0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–12 0–4 0–4 0–3 0–4 0–18 0–3 0–4 0–4 0–3 0–4 0–4
α = .85 .70 .75 .64 .74 α = .86 .61 .57 .73 .68 .56 .74
NSI-20 Total
2.8(5.3)
0/0/3
0–46
α = .90
NSI-22 Total 8-Hearing Difficulty 12-Appetite Disturbance
3.0(5.7) 0.1(0.4) 0.1(0.2)
0/0/4 0/0/0 0/0/0
0–50 0–3 0–4
α = .90 – –
Corrected item-scale correlation (subscale)
Corrected item-scale correlation (NSI-20)
.61 .61 .59 .62 .68 .57 .66 .63 .53 .70
– –
NSI-20 = 20-item Neurobehavioral Symptom Inventory, which omits items 8 (hearing difficulty) and 12 (appetite disturbance), based on Vanderploeg, Silva, et al. (2014) factor analysis. NSI-22 = Original 22-item Neurobehavioral Symptom Inventory.
Psychometric properties We calculated internal consistency reliability coefficients for each of the four scales, as well as the NSI-20 for the non-deployed and deployed nonclinical groups (see Tables 2 and 3). The NSI-22, NSI-20, and the four NSI-20 subscales demonstrated good internal consistency, similar to previous studies of the NSI-22 (e.g., King et al., 2012; Silva et al., 2013). The somatic subscale had the lowest but still acceptable inter-item consistency (alphas in the high .60s). For the deployed-mTBI sample, internal consistency reliability coefficients were also high: vestibular, α = .86; somatic, α = .77; cognitive, α = .90; affective, α = .89; NSI-20, α = .93.
622
JASON R. SOBLE ET AL.
Downloaded by [Colorado College] at 13:08 06 March 2015
Table 3. Neurobehavioral Symptom Inventory (NSI) descriptive statistics and reliability indices, Florida National Guard deployed nonclinical sample (n = 1064)
M(SD)
Percentile ranks 25/50/75
Range
NSI-20 Vestibular 1-Dizzy 2-Balance 3-Coordination
0.3(0.9) 0.1(0.4) 0.1(0.4) 0.1(0.4)
0/0/0 0/0/0 0/0/0 0/0/0
0–12 0–4 0–4 0–4
α = .75 .56 .67 .50
.48 .46 .47
NSI-20 Somatic 4-Headache 5-Nausea 6-Vision Problems 7-Light Sensitivity 9-Noise Sensitivity 10-Numbness/Tingling 11-Taste/Smell Changes
1.9(2.6) 0.6(0.9) 0.1(0.4) 0.2(0.6) 0.3(0.7) 0.2(0.7) 0.3(0.7) 0.1(0.3)
0/1/3 0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–17 0–4 0–4 0–4 0–4 0–4 0–4 0–3
α = .69 .36 .30 .47 .54 .47 .41 .36
.43 .38 .42 .46 .44 .43 .33
NSI-20 Cognitive 13-Poor Concentration 14-Forgetfulness 15-Diff Decision 16-Slow Thinking
1.3(2.4) 0.4(0.8) 0.4(0.8) 0.2(0.6) 0.3(0.6)
0/0/2 0/0/1 0/0/1 0/0/0 0/0/0
0–14 0–4 0–4 0–4 0–3
α = .88 .75 .75 .74 .79
.71 .68 .64 .67
NSI-20 Affective 17-Fatigue 18-Sleep Disturbance 19-Anxious/Tired 20-Sad/Depressed 21-Irritable 22-Frustration Intolerance
2.7(3.9) 0.5(0.8) 0.6(1.0) 0.5(0.8) 0.3(0.7) 0.4(0.8) 0.4(0.8)
0/1/4 0/0/1 0/0/1 0/0/1 0/0/0 0/0/1 0/0/1
0–22 0–4 0–4 0–4 0–4 0–4 0–4
α = .87 .66 .64 .74 .67 .62 .73
.72 .65 .71 .64 .60 .70
NSI-20 Total
6.2(8.2)
0/3/9
0–42
α = .91
NSI-22 Total 8-Hearing Difficulty 12-Appetite Disturbance
6.8(8.9) 0.5(0.8) 0.2(0.5)
0/3/10 0/0/1 0/0/0
0–45 0–4 0–4
α = .91 – –
Corrected item-scale correlation (Subscale)
Corrected item-scale correlation (NSI-20)
– –
NSI-20 = 20-item Neurobehavioral Symptom Inventory, which omits items 8 (hearing difficulty) and 12 (appetite disturbance), based on Vanderploeg, Silva, et al. (2014) factor analysis. NSI-22 = Original 22-item Neurobehavioral Symptom Inventory.
Item–total correlations are presented in Table 2 for the non-deployed-nonclinical sample and Table 3 for the deployed-nonclinical sample. Item–total correlations were high in most cases, indicating that items functioned well in predicting subscale and NSI-20 total scores. One exception was that NSI item measuring nausea. For the deployed nonclinical sample, nausea had a corrected item–total correlation of 0.30 with the NSI somatic subscale score. This index is right at the recommended cutoff of 0.30 (Nunnally & Bernstein, 1994; Traub, 1994), suggesting that endorsement of nausea had relatively lower predictability of NSI-20 somatic subscale scores. Nausea had
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
623
acceptable fit with the NSI-20 total score for this group, and with the NSI somatic subscale and NSI-20 total score for the non-deployed-nonclinical group.
Downloaded by [Colorado College] at 13:08 06 March 2015
Demographic analyses Prior to deriving normative data, we conducted a series of one-way analyses of variance (ANOVAs) on the data collected from the non-deployed-nonclinical and deployed-nonclinical samples to determine if significant differences existed on any of the four NSI scales on the basis of gender, race/ethnicity, or education. As shown in Tables 4 and 5, significant differences were found for gender and education across various scales. However, despite statistically significant differences, further examination revealed minimal effect sizes. Thus, because these demographic variables accounted for such a small portion of variance (i.e., < 4% maximum), we chose not to derive separate norms or corrections on the basis of these demographic characteristics.
Table 4. Neurobehavioral Symptom Inventory (NSI) scale means and standard deviations for relevant demographic variables, Florida National Guard non-deployed-nonclinical sample (n = 1453) NSI-20 Vestibular M(SD)
NSI-20 Somatic M(SD)
NSI-20 Cognitive M(SD)
NSI-20 Affective M(SD)
NSI-20 M(SD)
NSI-22 M(SD)
0.2(0.6) 0.5(1.1) 21.31a
The Clinical Neuropsychologist Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/ntcn20
Normative Data for the Neurobehavioral Symptom Inventory (NSI) and PostConcussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples ab
a
Jason R. Soble , Marc A. Silva , Rodney D. Vanderploeg ab
Curtiss , Heather G. Belanger Scott
abcde
, Alison J. Donnell
ae
abcde
, Glenn
& Steven G.
cefg
a
Mental Health and Behavioral Sciences Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA b
Click for updates
Department of Psychiatry & Behavioral Neurosciences, University of South Florida, Tampa, FL, USA c
Health Services Research and Development (HSR&D)/ Rehabilitation Research and Development (RR&D) Center of Excellence: Maximizing Rehabilitation Outcomes, James A. Haley Veterans’ Hospital, Tampa, FL, USA d
Department of Psychology, University of South Florida, Tampa, FL, USA e
Defense and Veterans Brain Injury Center, Tampa, FL, USA
f
Physical Medicine and Rehabilitation Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA g
Department of Internal Medicine, University of South Florida, Tampa, FL, USA Published online: 14 Mar 2014.
To cite this article: Jason R. Soble, Marc A. Silva, Rodney D. Vanderploeg, Glenn Curtiss, Heather G. Belanger, Alison J. Donnell & Steven G. Scott (2014) Normative Data for the Neurobehavioral Symptom Inventory (NSI) and Post-Concussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples, The Clinical Neuropsychologist, 28:4, 614-632, DOI: 10.1080/13854046.2014.894576 To link to this article: http://dx.doi.org/10.1080/13854046.2014.894576
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [Colorado College] at 13:08 06 March 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions
The Clinical Neuropsychologist, 2014 Vol. 28, No. 4, 614–632, http://dx.doi.org/10.1080/13854046.2014.894576
Normative Data for the Neurobehavioral Symptom Inventory (NSI) and Post-Concussion Symptom Profiles Among TBI, PTSD, and Nonclinical Samples Jason R. Soble1,2, Marc A. Silva1, Rodney D. Vanderploeg1,2,3,4,5, Glenn Curtiss1,2, Heather G. Belanger1,2,3,4,5, Alison J. Donnell1,5, and Steven G. Scott3,5,6,7
Downloaded by [Colorado College] at 13:08 06 March 2015
1
Mental Health and Behavioral Sciences Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA 2 Department of Psychiatry & Behavioral Neurosciences, University of South Florida, Tampa, FL, USA 3 Health Services Research and Development (HSR&D)/Rehabilitation Research and Development (RR&D) Center of Excellence: Maximizing Rehabilitation Outcomes, James A. Haley Veterans’ Hospital, Tampa, FL, USA 4 Department of Psychology, University of South Florida, Tampa, FL, USA 5 Defense and Veterans Brain Injury Center, Tampa, FL, USA 6 Physical Medicine and Rehabilitation Service, James A. Haley Veterans’ Hospital, Tampa, FL, USA 7 Department of Internal Medicine, University of South Florida, Tampa, FL, USA The Neurobehavioral Symptom Inventory (NSI) is a self-report measure of symptoms commonly associated with Post-Concussion Syndrome (PCS) that may emerge after mild traumatic brain injury (mTBI). Despite frequent clinical use, no NSI norms have been developed. Thus, the main objective of this study was to establish NSI normative data using the four NSI factors (i.e., vestibular, somatic, cognitive, and affective) identified by Vanderploeg, Silva, et al. (2014) among nonclinical epidemiological samples of deployed and non-deployed Florida National Guard members as well as a reference sample of Guard members with combat-related mTBI. In addition, NSI subscale profile patterns were compared across four distinct subgroups (i.e., non-deployed-nonclinical, deployed-nonclinical, deployed-mTBI, and deployed-PTSD). The deployed-nonclinical group endorsed greater PCS symptom severity than the non-deployed group, and the mTBI group uniformly endorsed more symptoms than both nonclinical groups. However, the PTSD group endorsed higher symptom severity relative to the other three subgroups. As such, this highlights the non-specificity of PCS symptoms and suggests that PTSD is associated with higher symptom endorsement than mTBI. Keywords: Neurobehavioral Symptom Inventory; Normative data; Post-Concussion Syndrome; Traumatic brain injury.
INTRODUCTION Post-Concussion Syndrome (PCS) refers to the collection of cognitive, emotional, and physical symptoms that sometimes emerge following mild traumatic brain injury (mTBI). PCS diagnostic criteria are formalized in the International Classification of Address correspondence to: Rodney D. Vanderploeg, James A. Haley Veterans’ Hospital Psychology Service (116B), 13000 Bruce B. Downs Blvd. Tampa, FL 33612, USA. E-mail: [email protected] (Received 17 October 2013; accepted 11 February 2014)
© 2014 Taylor & Francis
Downloaded by [Colorado College] at 13:08 06 March 2015
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
615
Diseases-Tenth Edition (ICD-10; World Health Organization, 1992) and Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition (DSM-IV; American Psychiatric Association [APA], 1994). However, PCS remains a controversial disorder (Granacher, 2008), and as an entity is absent in the DSM-5 (APA, 2013). Despite its lack of widespread acceptance as a valid clinical syndrome, health care providers continually are faced with evaluating and treating these so-called “post-concussion” cognitive, emotional, and physical symptoms that are reported by some patients with histories of TBI. TBI itself is not uncommon, with an annual incidence rate of 1.7 million in the U.S. (Faul, Xu, Wald, & Coronado, 2010), approximately 75% of which are classified as mild (Centers for Disease Control and Prevention, 2003). However, while the majority of individuals who sustain a mTBI recover from acute cognitive sequelae (Belanger & Vanderploeg, 2005; Rohling et al., 2011), a subset of individuals endorse PCS symptoms for months and/or years after injury (Dikmen, Machamer, Fann, & Temkin, 2010; Ruff, Camenzuli, & Mueller, 1996; Vanderploeg, Belanger, & Curtiss, 2009; Vanderploeg, Curtiss, Luis, & Salazar, 2007). Different diagnostic criteria exist between the ICD-10 (World Health Organization, 1992) and DSM-IV (APA, 1994). For instance, the ICD-10 criteria include headache, dizziness, fatigue, irritability, sleep problems, memory or concentration difficulty, and intolerance of stress, emotion, or alcohol. In contrast, the DSM-IV criteria contain fatigue, sleep problems, dizziness, headaches, irritability or aggression, apathy, personality changes, and anxiety, depression, or affective lability that persist for at least 3 months. While several common symptoms (e.g., headache, sleep problems) are included in both criteria, DSM-IV does not address cognitive problems, whereas ICD-10 speaks little to potential affective and personality changes. Thus, PCS prevalence rates can vary based on the specific diagnostic criteria used. In one prospective study of 178 adults with mild-moderate TBI, Boake et al. (2005) found that 64% met the ICD-10 PCS criteria, whereas only 11% met the DSM-IV criteria. Beyond yielding discrepant prevalence rates, these varying criteria also pose a challenge for accurately measuring PCS. Several PCS symptom measures have been developed, including the Post-Concussion Syndrome Checklist (PCSC; Gouvier, Cubic, Jones, Bentley, & Cutlip, 1992), Beaumont Post-Concussional Index (BPCI; Trahan, Ross, & Trahan, 1997), and Rivermead Post-Concussion Symptoms Questionnaire (RPQ; King, Crawford, Wenden, Moss, & Wade, 1995). Of particular interest to the current study is the Neurobehavioral Symptom Inventory (NSI; Cicerone & Kalmar, 1995), originally published as the Post MTBI Symptom Checklist (Cicerone & Kalmar, 1995; Cicerone et al., 1996), and based in part on symptoms included in a structured interview used in a multicenter mTBI study (Levin et al., 1987). The NSI contains 22 common symptoms that patients rate on a 5-point scale based on how much each symptom has disturbed them during the past two weeks, month, or since time of injury, depending on the version used. Surprisingly, no comprehensive normative data have yet been established for this measure despite the NSI’s routine use by the Departments of Defense and Veterans Affairs (VA), including the VA’s TBI Comprehensive Evaluation, as well as it being identified as a supplemental common outcome measure by the multiagency TBI Outcomes Workgroup (Wilde et al., 2010). While past studies that examined the factor structure of the NSI (i.e., Benge, Pastorek, & Thompson, 2009; Cicerone & Kalmar, 1995; Caplan et al., 2010; Meterko et al., 2012) have produced inconsistent results, overall findings generally indicate that
Downloaded by [Colorado College] at 13:08 06 March 2015
616
JASON R. SOBLE ET AL.
PCS is a multidimensional construct. For instance, Cicerone and Kaplan (1995) identified four clusters and five items that did not fit any cluster. Benge et al. (2009) identified a four-factor solution (sensory, affective, a headache, nausea, and visual disturbance factor, and an unclear factor) and a six-factor solution after controlling for Post-Traumatic Stress Disorder (PTSD) (vestibular, cognitive, affective, and hearingrelated, plus two unclear factors). Caplan et al. (2010) found a two-factor solution (somatic/sensory and psychological) and a three-factor solution (somatic/sensory, affective, and cognitive). Finally, based on exploratory factor analysis, Meterko et al. (2012) reported a three-factor solution (somatic/sensory, cognitive, and affective) and a fourfactor solution (somatic, vestibular, cognitive, and affective). Given these inconsistent findings, a more recent study (Vanderploeg, Silva, et al., 2014) used structural equation modeling to compare the existing NSI models. They found that Meterko et al.’s (2012) four-factor model (vestibular, somatic, cognitive, and affective) minus two items (i.e., hearing problems and appetite disturbance) provided the best overall fit for PCS symptoms in both a clinical and nonclinical sample). This multidimensional, four-factor model also may allow for improved discrimination among different clinical subgroups. Beyond its multidimensional nature, the construct of PCS has been called into question. Many of the core symptoms are nonspecific to mTBI and can be found among healthy individuals with no history of TBI (Chan, 2001; Iverson & Lange, 2003; Vanderploeg et al., 2007, 2009). Not surprisingly, past studies comparing NSI scores of those with mTBI to nonclinical samples found that those with mTBI endorsed higher levels of PCS symptoms (King et al., 2012). However, elevated levels of PCS symptoms also are commonly found in the presence of a host of other medical conditions, such as chronic pain (Iverson & McCracken, 1997) and non-TBI physical trauma (Meares et al., 2008), as well as psychiatric disorders including depression, PTSD, and somatization (Belanger, Kretzmer, Vanderploeg, & French, 2010; Bryant, 2001; Donnell, Kim, Silva, & Vanderploeg, 2012). In fact, King et al. (2012) found higher correlations between PCS symptoms and measures of affective psychopathology than TBI. Donnell et al. (2012) reported that PCS symptoms were more frequently endorsed by veterans diagnosed with psychiatric disorders (i.e., PTSD, generalized anxiety, major depression, somatization) than those with mTBI. Given the multidimensional and nonspecific nature of PCS symptoms, the clinical utility of the NSI is limited without adequate normative data and established symptom profile patterns to guide interpretation.
STUDY RATIONALE AND GOALS No standard yet exists to help clinicians interpret NSI scores. Determining the clinical significance of point-in-time raw scores remains a challenge without knowledge of performance by appropriate references groups. To address this limitation, the primary objective of this study was to establish normative data for the NSI and its four factors as identified by Vanderploeg, Silva, et al. (2014). Moreover, given the NSI’s routine use as part of the VA’s TBI Comprehensive Evaluation, a secondary objective was to also develop reference-group normative data using an epidemiological, rather than clinically presenting, cohort of military personnel with combat-related mTBI, thus permitting more sophisticated interpretation of obtained NSI scores. It is our hope that, by presenting normative data for two nonclinical samples (i.e., non-deployed and deployed) as
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
617
well as a mTBI reference sample, health care providers will more effectively utilize the NSI in clinical practice. Beyond developing normative data, a final objective of this study was to examine NSI score patterns among different subgroups (defined in “Method” below).
METHOD
Downloaded by [Colorado College] at 13:08 06 March 2015
Participants Study data came from the 2009–2010 Florida National Guard survey. The study sample is described elsewhere (Vanderploeg et al., 2012). In brief, the study participants were an unselected, representative sample of the Florida National Guard who completed an anonymous online survey that included the NSI. Of 10,400 letters mailed out inviting members to participate, 4005 individuals responded, yielding a response rate of 41.3%. This current sample consisted of 3098 participants who provided usable data and fully completed the survey (we excluded 423 who did not fully complete the survey, 371 who completed the survey more than once, and 113 who gave inconsistent or implausible responses). Of these 3098 responders, 1443 reported at least one deployment and 1655 reported no deployments. Guard members who deployed returned from deployment up to 95 months prior to survey (M = 31.8; SD = 24.4). Of the 3098, the following subgroups were identified (see Table 1 for demographic characteristics of subgroups). Non-deployed-nonclinical (n = 1453). Among the 53% who did not deploy, 175 reported a history of TBI, 32 screened positive for PTSD, 24 screened positive for other anxiety disorders, and 16 screened positive for depressive disorders (screening procedures are described below in the “Measures” section). These individuals were excluded (categories not mutually exclusive), resulting in 1453 participants who comprised the non-deployed-nonclinical reference group. Deployed-nonclinical (n = 1064). Among the 47% who deployed, 392 reported a mTBI prior to deployment, 144 reported a mTBI during deployment, 100 screened positive for PTSD, 61 screened positive for other anxiety disorders, and 47 screened positive for depressive disorders. These individuals were excluded (categories not mutually exclusive), resulting in 1064 participants for who comprised the deployed-nonclinical group. Deployed-mTBI (n = 108). Deployment-related mTBI was determined by endorsement of a motor vehicle accident, blast exposure, or any other event resulting in loss of consciousness, blacking out, or memory gaps. Those who reported a head injury as indicated above were then asked about the duration of any memory gaps (most were less than an hour and all were less than a day). Participants who endorsed being dazed and confused without memory gaps were excluded from the mTBI group because combat events such as unexpected ambushes, firefights, or nearby blasts may be experienced as confusing, disorienting, or “dazing” although not necessarily indicative of brain injury. While medical documentation of mTBI at time of injury is ideal, medical records from the acute period may be unavailable (if medical attention is sought at all). Consequently, self-report is common in studies of mTBI in the chronic phase. Among the 47% who deployed, 144 endorsed sustaining mTBI during deployment. Of these, 36 screened positive for PTSD and were excluded, as a focus of this study was to
21.2 78.8 16.7 20.2 56.1 7.9 33.4 44.9 21.6 2.3(2.3) 1.3(2.4) 1.2(2.5) 19.0(4.7)
17.0 83.0 14.8
17.3 61.9 6.1
26.3 47.5 26.1
2.6(2.5) 2.7(4.2) 2.4(4.3) 22.6(10.9)
Non-deployed –nonclinical n = 1,453
2.7(2.5) 2.6(3.4) 2.1(3.2) 21.7(6.9)
20.4 49.3 30.3
15.2 65.2 4.5
13.3 86.7 15.0
Deployed-nonclinical n = 1,064
3.3(2.7) 4.7(4.5) 4.2(4.2) 28.7(10.6)
17.6 51.9 30.6
16.7 67.6 3.7
13.0 87.0 12.0
Deployed-mTBI n = 108
4.6(3.2) 14.7(4.8) 15.3(6.3) 63.3(10.2)
28.8 48.1 23.1
19.2 57.7 3.8
21.2 78.8 19.2
DeployedPTSD n = 52
Race/Ethnicity “Other” included participants who identified as Native American, Alaskan Native, Asian, Pacific Islander, or Biracial. Percentages may not sum to 100 due to rounding. AUDIT-C = Alcohol Use Disorder Identification Test-Screen; GAD7 = Generalized Anxiety Disorder Questionnaire; PHQ9 = Patient Health Questionnaire, Depression Module; PCL-C = PTSD Checklist, Civilian Version.
Gender Female (%) Male (%) Race/Ethnicity Black (%) Hispanic(%) White (%) Other (%) Education (years) ≤ 12 (%) 13-15 (%) ≥ 16 (%) Clinical Symptoms AUDIT-C, M(SD) GAD7, M(SD) PHQ9, M(SD) PCL-C, M(SD)
Total sample N = 3,098
Table 1. Sociodemographic and selected clinical characteristics
Downloaded by [Colorado College] at 13:08 06 March 2015
618 JASON R. SOBLE ET AL.
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
619
Downloaded by [Colorado College] at 13:08 06 March 2015
compare participants with PTSD (but no mTBI) with participants with mTBI (but no PTSD). We chose not to exclude the 38 participants who screened positive for depressive and other anxiety disorders, because these symptoms are not uncommon after mTBI (e.g., Dikmen et al., 2010) and are included in DSM-IV PCS symptom criteria (APA, 1994). The final Deployed-mTBI group was composed of 108 participants. Deployed-PTSD (n = 52). Among the 47% who deployed, we excluded 190 who endorsed a predeployment mTBI and 144 who endorsed a deployment mTBI. We chose not to exclude the 36 who screened positive for depressive and other anxiety disorders given that PTSD and depression are highly comorbid as well as findings suggestive of a single underlying symptom dimension between the disorders (e.g., Elhai et al., 2011). The final Deployed-PTSD group was composed of 52 participants. Measures Neurobehavioral Symptom Inventory (NSI). The NSI (Cicerone & Kalmar, 1995) is a 22-item self-report questionnaire designed to assess PCS symptom severity. Individuals rated the severity of common PCS symptoms in terms of “how much they have disturbed you over the past month” on a scale ranging from 0 (none) to 4 (very severe). The NSI was found to have good internal consistency (King et al., 2012; Silva, Barwick, Kretzmer, Vanderploeg, & Belanger, 2013) and 7-day test–retest stability (Silva et al., 2013). In this study we specifically examined the NSI vestibular, somatic, cognitive, and affective factors based on the four-factor, 20-item NSI model described by Vanderploeg, Silva, et al. (2014). This model excludes two items (i.e., hearing problems and appetite disturbance) due to poor fit. NSI-20 scores range from 0 to 80. For the subscales, vestibular scores range from 0 to 12, somatic scores range from 0 to 28, cognitive scores range from 0 to 16, and affective scores range from 0 to 24. Generalized Anxiety Disorder Questionnaire (GAD-7). The GAD-7 (Spitzer, Kroenke, Williams, & Lowe, 2006) is a questionnaire designed to screen for Generalized Anxiety Disorder based on DSM-IV criteria. Scores range from 0 to 21, with higher scores indicating greater symptom endorsement. In this study, participants were classified as having a probable anxiety disorder if they endorsed: (a) nervousness or anxiety more than half the time; (b) three of six other anxiety symptoms more than half the time; and (c) impairment in work, home, or interpersonal functioning at the “very difficult” level or higher (Spitzer et al., 2006). Patient Health Questionnaire Depression Module (PHQ-9). The PHQ-9 (Spitzer, Kroenke, & Williams, 1999; see also Kroenke & Spitzer, 2002) is a questionnaire designed to screen for Major Depressive Disorder based on DSM-IV criteria. Scores range from 0 to 27, with higher scores indicating greater symptom endorsement. In this study, participants were classified as having a probable major depressive disorder, if they met DSM-IV symptom criteria and reported impairment in work, home, or interpersonal functioning at the “very difficult” level or higher (Spitzer et al., 1999). PTSD Checklist-Civilian Version (PCL-C). The PCL (Weathers, Litz, Herman, Huska, & Keane, 1993) is a 17-item self-report measure of DSM-IV PTSD symptoms in which participants rate items on a scale ranging from 1 (not at all) to 5 (extremely). Scores can range from 17 to 85. We operationalized PTSD as a PCL score of greater
620
JASON R. SOBLE ET AL.
Downloaded by [Colorado College] at 13:08 06 March 2015
than 50 and meeting DSM-IV symptom criteria for PTSD by endorsing as moderate (3) or higher at least one re-experiencing symptom, at least three avoidance symptoms, and at least two hyperarousal symptoms, as per recommended guidelines (National Center for PTSD, 2012). Consistent with DSM-IV criteria for PTSD, participants were also required to report that their symptoms caused impairment in work, home, or interpersonal functioning at the “very difficult” level or higher. Alcohol Use Disorders Identification Test-Consumption Questions screening test (AUDIT-C). The AUDIT-C (Bush, Kivlahan, McDonell, Fihn, & Bradley, 1998) is a three-item measure designed to screen for heavy drinking and alcohol use disorders. Possible scores range from 0 to 12 with higher scores associated with greater risk of problematic drinking. In this study, AUDIT-C scores are presented for descriptive purposes, but are not used as exclusion criteria.
Procedures After establishing the above groups, subscale scores for each of the four factors identified by Vanderploeg, Silva, et al. (2014) were formed by summing the raw scores on the items deemed to load on the factor. Two total scores were also formed: the NSI22, which was calculated by summing raw scores on all items comprising the NSI and the NSI-20, which was calculated by summing raw scores on all items except hearing problems and appetite problems.
Data analyses NSI reliability analyses were computed for the non-deployed-nonclinical normative sample (n = 1453) and for the deployed nonclinical sample (n = 1064). Normative data were then derived for both of these groups. Reference comparison data were also derived for the deployed-mTBI sample (n = 108). Frequency distributions were examined and, due to skew and kurtosis of the distributions, normative data were derived for various percentile ranges. Last, multivariate analysis of variance (MANOVA) was used to investigate whether the four NSI scales were useful in profiling difference across subgroups. Scale mean scores were used to correct for differences in number of items comprising each scale.
RESULTS Descriptive statistics Indices of central tendency and scale score distributions are presented in Tables 2 and 3. These tables show that the NSI total, subscale, and item level scores are positively skewed with means and medians near zero for most NSI items. Therefore, an attempt to transform the data into a normal distribution would not be appropriate in this case. Rather, normative tables should be presented in a categorical manner.
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
621
Downloaded by [Colorado College] at 13:08 06 March 2015
Table 2. Neurobehavioral Symptom Inventory (NSI) descriptive statistics and reliability indices, Florida National Guard non-deployed nonclinical sample (n = 1453)
M(SD)
Percentile ranks 25/50/75
Range
NSI-20 Vestibular 1-Dizzy 2-Balance 3-Coordination
0.2(0.8) 0.1(0.4) 0.1(0.3) 0.1(0.3)
0/0/0 0/0/0 0/0/0 0/0/0
0–7 0–3 0–3 0–2
α = .74 .59 .65 .51
.48 .52 .51
NSI-20 Somatic 4-Headache 5-Nausea 6-Vision Problems 7-Light Sensitivity 9-Noise Sensitivity 10-Numbness/Tingling 11-Taste/Smell Changes
0.8(1.7) 0.3(0.7) 0.1(0.3) 0.1(0.4) 0.1(0.4) 0.1(0.3) 0.1(0.5) 0.0(0.2)
0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–15 0–4 0–4 0–4 0–3 0–3 0–4 0–4
α = .68 .50 .49 .39 .44 .35 .40 .38
.49 .40 .41 .41 .41 .42 .37
NSI-20 Cognitive 13-Poor Concentration 14-Forgetfulness 15-Diff Decision 16-Slow Thinking NSI-20 Affective 17-Fatigue 18-Sleep Disturbance 19-Anxious/Tired 20-Sad/Depressed 21-Irritable 22-Frustration Intolerance
0.5(1.5) 0.2(0.5) 0.2(0.5) 0.1(0.3) 0.1(0.4) 1.2(2.5) 0.2(0.5) 0.3(0.7) 0.2(0.6) 0.2(0.5) 0.2(0.5) 0.2(0.5)
0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–12 0–4 0–4 0–3 0–4 0–18 0–3 0–4 0–4 0–3 0–4 0–4
α = .85 .70 .75 .64 .74 α = .86 .61 .57 .73 .68 .56 .74
NSI-20 Total
2.8(5.3)
0/0/3
0–46
α = .90
NSI-22 Total 8-Hearing Difficulty 12-Appetite Disturbance
3.0(5.7) 0.1(0.4) 0.1(0.2)
0/0/4 0/0/0 0/0/0
0–50 0–3 0–4
α = .90 – –
Corrected item-scale correlation (subscale)
Corrected item-scale correlation (NSI-20)
.61 .61 .59 .62 .68 .57 .66 .63 .53 .70
– –
NSI-20 = 20-item Neurobehavioral Symptom Inventory, which omits items 8 (hearing difficulty) and 12 (appetite disturbance), based on Vanderploeg, Silva, et al. (2014) factor analysis. NSI-22 = Original 22-item Neurobehavioral Symptom Inventory.
Psychometric properties We calculated internal consistency reliability coefficients for each of the four scales, as well as the NSI-20 for the non-deployed and deployed nonclinical groups (see Tables 2 and 3). The NSI-22, NSI-20, and the four NSI-20 subscales demonstrated good internal consistency, similar to previous studies of the NSI-22 (e.g., King et al., 2012; Silva et al., 2013). The somatic subscale had the lowest but still acceptable inter-item consistency (alphas in the high .60s). For the deployed-mTBI sample, internal consistency reliability coefficients were also high: vestibular, α = .86; somatic, α = .77; cognitive, α = .90; affective, α = .89; NSI-20, α = .93.
622
JASON R. SOBLE ET AL.
Downloaded by [Colorado College] at 13:08 06 March 2015
Table 3. Neurobehavioral Symptom Inventory (NSI) descriptive statistics and reliability indices, Florida National Guard deployed nonclinical sample (n = 1064)
M(SD)
Percentile ranks 25/50/75
Range
NSI-20 Vestibular 1-Dizzy 2-Balance 3-Coordination
0.3(0.9) 0.1(0.4) 0.1(0.4) 0.1(0.4)
0/0/0 0/0/0 0/0/0 0/0/0
0–12 0–4 0–4 0–4
α = .75 .56 .67 .50
.48 .46 .47
NSI-20 Somatic 4-Headache 5-Nausea 6-Vision Problems 7-Light Sensitivity 9-Noise Sensitivity 10-Numbness/Tingling 11-Taste/Smell Changes
1.9(2.6) 0.6(0.9) 0.1(0.4) 0.2(0.6) 0.3(0.7) 0.2(0.7) 0.3(0.7) 0.1(0.3)
0/1/3 0/0/1 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0 0/0/0
0–17 0–4 0–4 0–4 0–4 0–4 0–4 0–3
α = .69 .36 .30 .47 .54 .47 .41 .36
.43 .38 .42 .46 .44 .43 .33
NSI-20 Cognitive 13-Poor Concentration 14-Forgetfulness 15-Diff Decision 16-Slow Thinking
1.3(2.4) 0.4(0.8) 0.4(0.8) 0.2(0.6) 0.3(0.6)
0/0/2 0/0/1 0/0/1 0/0/0 0/0/0
0–14 0–4 0–4 0–4 0–3
α = .88 .75 .75 .74 .79
.71 .68 .64 .67
NSI-20 Affective 17-Fatigue 18-Sleep Disturbance 19-Anxious/Tired 20-Sad/Depressed 21-Irritable 22-Frustration Intolerance
2.7(3.9) 0.5(0.8) 0.6(1.0) 0.5(0.8) 0.3(0.7) 0.4(0.8) 0.4(0.8)
0/1/4 0/0/1 0/0/1 0/0/1 0/0/0 0/0/1 0/0/1
0–22 0–4 0–4 0–4 0–4 0–4 0–4
α = .87 .66 .64 .74 .67 .62 .73
.72 .65 .71 .64 .60 .70
NSI-20 Total
6.2(8.2)
0/3/9
0–42
α = .91
NSI-22 Total 8-Hearing Difficulty 12-Appetite Disturbance
6.8(8.9) 0.5(0.8) 0.2(0.5)
0/3/10 0/0/1 0/0/0
0–45 0–4 0–4
α = .91 – –
Corrected item-scale correlation (Subscale)
Corrected item-scale correlation (NSI-20)
– –
NSI-20 = 20-item Neurobehavioral Symptom Inventory, which omits items 8 (hearing difficulty) and 12 (appetite disturbance), based on Vanderploeg, Silva, et al. (2014) factor analysis. NSI-22 = Original 22-item Neurobehavioral Symptom Inventory.
Item–total correlations are presented in Table 2 for the non-deployed-nonclinical sample and Table 3 for the deployed-nonclinical sample. Item–total correlations were high in most cases, indicating that items functioned well in predicting subscale and NSI-20 total scores. One exception was that NSI item measuring nausea. For the deployed nonclinical sample, nausea had a corrected item–total correlation of 0.30 with the NSI somatic subscale score. This index is right at the recommended cutoff of 0.30 (Nunnally & Bernstein, 1994; Traub, 1994), suggesting that endorsement of nausea had relatively lower predictability of NSI-20 somatic subscale scores. Nausea had
NEUROBEHAVIORAL SYMPTOM INVENTORY NORMS
623
acceptable fit with the NSI-20 total score for this group, and with the NSI somatic subscale and NSI-20 total score for the non-deployed-nonclinical group.
Downloaded by [Colorado College] at 13:08 06 March 2015
Demographic analyses Prior to deriving normative data, we conducted a series of one-way analyses of variance (ANOVAs) on the data collected from the non-deployed-nonclinical and deployed-nonclinical samples to determine if significant differences existed on any of the four NSI scales on the basis of gender, race/ethnicity, or education. As shown in Tables 4 and 5, significant differences were found for gender and education across various scales. However, despite statistically significant differences, further examination revealed minimal effect sizes. Thus, because these demographic variables accounted for such a small portion of variance (i.e., < 4% maximum), we chose not to derive separate norms or corrections on the basis of these demographic characteristics.
Table 4. Neurobehavioral Symptom Inventory (NSI) scale means and standard deviations for relevant demographic variables, Florida National Guard non-deployed-nonclinical sample (n = 1453) NSI-20 Vestibular M(SD)
NSI-20 Somatic M(SD)
NSI-20 Cognitive M(SD)
NSI-20 Affective M(SD)
NSI-20 M(SD)
NSI-22 M(SD)
0.2(0.6) 0.5(1.1) 21.31a
