Advertisement

Trajectories and Predictors of the Course of Mental Health After Spinal Cord Injury

  • Christel M. van Leeuwen
    Affiliations
    Swiss Paraplegic Research, Nottwil, Switzerland

    Rudolf Magnus Institute of Neuroscience and Center of Excellence for Rehabilitation Medicine, University Medical Center Utrecht and Rehabilitation Center De Hoogstraat, Utrecht, The Netherlands
    Search for articles by this author
  • Trynke Hoekstra
    Affiliations
    Department of Health Sciences, Faculty of Earth and Life Sciences, VU University Amsterdam, Amsterdam, The Netherlands

    Department of Epidemiology and Biostatistics and the EMGO Institute for Health and Care Research, VU University Medical Center Amsterdam, Amsterdam, The Netherlands
    Search for articles by this author
  • Casper F. van Koppenhagen
    Affiliations
    Rudolf Magnus Institute of Neuroscience and Center of Excellence for Rehabilitation Medicine, University Medical Center Utrecht and Rehabilitation Center De Hoogstraat, Utrecht, The Netherlands
    Search for articles by this author
  • Sonja de Groot
    Affiliations
    Center for Human Movement Sciences, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

    Reade, Centre for Rehabilitation and Rheumatology, Amsterdam, The Netherlands
    Search for articles by this author
  • Marcel W. Post
    Correspondence
    Reprint requests to Marcel W. Post, PhD, Rehabilitation Center De Hoogstraat, Rembrandtkade 10, 3583 TM, Utrecht, The Netherlands
    Affiliations
    Swiss Paraplegic Research, Nottwil, Switzerland

    Rudolf Magnus Institute of Neuroscience and Center of Excellence for Rehabilitation Medicine, University Medical Center Utrecht and Rehabilitation Center De Hoogstraat, Utrecht, The Netherlands
    Search for articles by this author

      Abstract

      van Leeuwen CM, Hoekstra T, van Koppenhagen CF, de Groot S, Post MW. Trajectories and predictors of the course of mental health after spinal cord injury.

      Objective

      To study the course and predictors of mental health in the period between the start of active spinal cord injury (SCI) rehabilitation and 5 years after discharge. The hypothesis was that different mental health trajectories would be identified.

      Design

      Multicenter prospective cohort study with measurements at the start of active rehabilitation, after 3 months, at discharge, 1, 2, and 5 years after discharge.

      Setting

      Eight Dutch rehabilitation centers with specialized SCI units.

      Participants

      Persons (N=206) with recently acquired SCI aged between 18 and 65 years.

      Interventions

      Not applicable.

      Main Outcome Measure

      The 5-item Mental Health Index with a total score between 0 (lowest mental health) and 100 (highest mental health).

      Results

      Levels of mental health increased between the start of active rehabilitation and 3 months later, remained stable thereafter, and increased again between 2 and 5 years after discharge. Latent class growth mixture modeling revealed 5 trajectories: (1) high scores (above 80) at all time-points (52%), (2) low scores (≤60) at all time-points (4%), (3) early recovery from 40 to scores above 70 (13%), (4) intermediate scores from 60 to scores above 70 (29%), and (5) severe deterioration of scores above 70 to scores below 30 (2%). Pain, sex, and education level were predictors to distinguish between the 5 trajectories.

      Conclusions

      Five different mental health trajectories were identified between the start of active rehabilitation and 5 years after discharge. About one third of the persons with SCI still perceived moderate to severe mental health problems 5 years after discharge. Pain, sex, and education level only predicted a small part of the variance in mental health trajectories.

      Key Words

      List of Abbreviations:

      AIS (American Spinal Injury Association Impairment Scale), BIC (Bayesian information criterion), BLRT (bootstrap likelihood ratio test), MHI-5 (5-item Mental Health Index), SCI (spinal cord injury)
      A SPINAL CORD INJURY (SCI) is a major life event leading to serious physical disability and secondary medical problems.
      • Dijkers M.P.
      Quality of life of individuals with spinal cord injury: a review of conceptualization, measurement, and research findings.
      • Post M.
      • Noreau L.
      Quality of life after spinal cord injury.
      There is evidence that the prevalence of mental health problems, such as depression, is elevated in persons with chronic SCI.
      • Post M.W.
      • van Leeuwen C.M.
      Psychosocial issues in spinal cord injury: a review.
      In the present study, mental health is used as a general term covering mood, emotions, and distress.
      • Post M.W.
      • van Leeuwen C.M.
      Psychosocial issues in spinal cord injury: a review.
      Only a few studies have assessed the course of mental health early after SCI. Two studies in inpatient rehabilitation found a decline of depressed mood within the first 12 to 18 weeks after SCI.
      • Kennedy P.
      • Lude P.
      • Elfström M.L.
      • Smithson E.
      Cognitive appraisals, coping and quality of life outcomes: a multi-centre study of spinal cord injury rehabilitation.
      • Kennedy P.
      • Rogers B.A.
      Anxiety and depression after spinal cord injury: a longitudinal analysis.
      One of these studies also reported an increase in depressed mood with longer duration of rehabilitation.
      • Kennedy P.
      • Lude P.
      • Elfström M.L.
      • Smithson E.
      Cognitive appraisals, coping and quality of life outcomes: a multi-centre study of spinal cord injury rehabilitation.
      In contrast, 2 other studies reported stable levels of depressed mood during inpatient rehabilitation.
      • Hancock K.M.
      • Craig A.R.
      • Dickson H.G.
      • Chang E.
      • Martin J.
      Anxiety and depression over the first year of spinal cord injury: a longitudinal study.
      • Cushman L.A.
      • Dijkers M.P.
      Depressed mood in spinal cord injured patients: staff perceptions and patient realities.
      Figures from the SCI Model Systems showed that the proportion of persons with depressed mood in the clinical range was more or less stable between 1 and 5 years postinjury,
      • Hoffman J.M.
      • Bombardier C.H.
      • Graves D.E.
      • Kalpakjian C.Z.
      • Krause J.S.
      A longitudinal study of depression from 1 to 5 years after spinal cord injury.
      • Richardson E.J.
      • Richards J.S.
      Factor structure of the PHQ-9 screen for depression across time since injury among persons with spinal cord injury.
      although there were large individual changes in depression status over time.
      • Hoffman J.M.
      • Bombardier C.H.
      • Graves D.E.
      • Kalpakjian C.Z.
      • Krause J.S.
      A longitudinal study of depression from 1 to 5 years after spinal cord injury.
      In sum, conflicting research findings have been reported with respect to the course of mental health early after SCI. Moreover, all studies reported average scores or overall proportions. This might obscure individual differences in the course of mental health after an SCI,
      • van Leeuwen C.M.
      • Post M.W.
      • Hoekstra T.
      • et al.
      Trajectories in the course of life satisfaction after spinal cord injury: identification and predictors.
      even in case these overall figures are stable.
      • Hoffman J.M.
      • Bombardier C.H.
      • Graves D.E.
      • Kalpakjian C.Z.
      • Krause J.S.
      A longitudinal study of depression from 1 to 5 years after spinal cord injury.
      In an earlier study, we found 5 distinct trajectories in the course of life satisfaction after an SCI, suggesting that we might be able to identify trajectories in the course of mental health as well.
      Several reviews have looked at predictors of mental health.
      • North N.T.
      The psychological effects of spinal cord injury: a review.
      • Müller R.
      • Peter C.
      • Cieza A.
      • Geyh S.
      The role of social support and social skills in people with spinal cord injury—a systematic review of the literature.
      • van Leeuwen C.M.
      • Kraaijeveld S.
      • Lindeman E.
      • Post M.W.
      Associations between psychological factors and quality of life ratings in persons with spinal cord injury: a systematic review.
      • Peter C.
      • Muller R.
      • Cieza A.
      • Geyh S.
      Psychological resources in spinal cord injury: a systematic literature review.
      An older review
      • North N.T.
      The psychological effects of spinal cord injury: a review.
      concluded that a younger age, being a woman, certain personality traits (eg, locus of control), better perceived social support, and having good communication skills predict favorable psychological outcome. Moreover, pain, medication, isolation, medical complications, and cognitive problems were described as factors affecting the psychological state.
      • North N.T.
      The psychological effects of spinal cord injury: a review.
      Three recent reviews have shown that social support and social skills are positively related to mental health
      • Müller R.
      • Peter C.
      • Cieza A.
      • Geyh S.
      The role of social support and social skills in people with spinal cord injury—a systematic review of the literature.
      and that psychological factors, such as perceived control, sense of coherence, self-worth, hope, and purpose in life, are consistently associated with higher mental health.
      • van Leeuwen C.M.
      • Kraaijeveld S.
      • Lindeman E.
      • Post M.W.
      Associations between psychological factors and quality of life ratings in persons with spinal cord injury: a systematic review.
      • Peter C.
      • Muller R.
      • Cieza A.
      • Geyh S.
      Psychological resources in spinal cord injury: a systematic literature review.
      The first aim of the present study was to examine the course of mental health in the period between the start of active SCI rehabilitation and 5 years after discharge. We hypothesized that different mental health trajectories can be identified. The second aim was to examine predictors of the course of mental health. We assessed demographics, lesion characteristics, physical factors, and social support as predictors of the course of mental health.

      Methods

      Participants

      This study is a follow-up of the Dutch research program Restoration of mobility in the rehabilitation of persons with spinal cord injuries.
      • de Groot S.
      • Dallmeijer A.J.
      • Post M.W.
      • et al.
      Demographics of the Dutch multicenter prospective cohort study ‘Restoration of mobility in spinal cord injury rehabilitation’.
      In this research program, 225 persons admitted for initial inpatient rehabilitation were selected consecutively in 8 rehabilitation centers with specialized SCI units in the Netherlands between August 2000 and July 2003. Subjects were included if they fulfilled the following criteria: (1) a recently acquired SCI; (2) aged between 18 and 65 years (this—arbitrary—age limit was chosen to exclude older adults who were already retired at the onset of SCI, more likely to have comorbidity, and for whom the physical tests in our test protocol might be too burdensome); (3) grades A, B, C, or D on the American Spinal Injury Association Impairment Scale (AIS); and (4) expected permanent wheelchair dependency. Participants were excluded if they had (1) an SCI caused by a malignant tumor, (2) a progressive disease, (3) psychiatric problems, or (4) insufficient command of the Dutch language (unable to understand the goal of the study and test instructions). The research protocol was approved by the Medical Ethics Committee of the Rehabilitation Limburg/Institute for Rehabilitation Research. All persons gave written informed consent.

      Procedure

      In total, there were 6 measurement occasions: at the start of active rehabilitation, defined as the moment that a person could sit for 3 to 4 hours, 3 months after the start of active rehabilitation, at discharge from inpatient rehabilitation, and 1, 2, and 5 years after discharge. The measurements used in the present study comprised a medical examination, an oral interview with a trained research assistant, and a self-report questionnaire.

      Instruments

      Mental health

      Mental health was assessed using the 5-item Mental Health Index (MHI-5), consisting of items concerning nervousness, sadness, peacefulness, mood, and happiness.
      • Ware J.E.
      • Snow K.K.
      • Kosinski M.
      • Gandek B.
      SF-36 Health Survey: manual and interpretation guide.
      Respondents rated the frequency of each item during the previous 4 weeks on a 6-point scale (1 = all of the time, 2 = most of the time, 3 = a good bit of the time, 4 = some of the time, 5 = a little of the time, 6 = none of the time), for example, How much of the time, during the last month, have you felt downhearted and blue? A total score between 0 (lowest mental health) and 100 (highest mental health) was computed.
      • Ware J.E.
      • Snow K.K.
      • Kosinski M.
      • Gandek B.
      SF-36 Health Survey: manual and interpretation guide.
      A cutoff point of 72 or lower refers to mental health problems, and of 60 or lower refers to severe mental health problems.
      • Hoeymans N.
      • Garssen A.A.
      • Westert G.P.
      • Verhaak P.F.
      Measuring mental health of the Dutch population: a comparison of the GHQ-12 and the MHI-5.
      • Perenboom R.
      • Oudshoorn K.
      • van Herten L.
      • Hoeymans N.
      • Bijl R.
      [Life expectancy in good mental health: establishing cut-offs for the MHI-5 and GHQ-12] [Dutch].
      The validity and reliability of the MHI-5 in persons with SCI was good.
      • van Leeuwen Jr, C.M.
      • van der Woude L.H.
      • Post M.W.
      Validity of the mental health subscale of the SF-36 in persons with spinal cord injury.

      Lesion characteristics

      Lesion characteristics were assessed according to the International Standards for Neurological Classification of SCI.
      • Maynard Jr, F.M.
      • Bracken M.B.
      • Creasey G.
      • et al.
      International standards for neurological and functional classification of spinal cord injury.
      Neurologic levels below T1 were defined as paraplegia; neurologic levels at or above T1 were defined as tetraplegia. AIS grades A and B were considered motor complete; AIS grades C and D were considered motor incomplete. Cause of injury was dichotomized in traumatic versus nontraumatic.

      Functional independence

      Functional independence was measured with the motor score of the FIM, consisting of 13 items about self-care, mobility, transfers, and toileting. The FIM motor score is a responsive measure in persons with SCI.
      • Hall K.M.
      • Cohen C.E.
      • Wright J.
      • Call M.
      • Werner P.
      Characteristics of the Functional Independence Measure in traumatic spinal cord injury.
      A high score refers to a high level of functional independence.

      Secondary impairments

      Secondary impairments were split into 2 categories: pain and other secondary impairments. In the pain category, 13 locations were assessed to measure the severity of pain in muscles and joints, including the upper limbs, lower limbs, back, and neck. For each location, a 5-point severity scale was scored, ranging from no pain to very severe pain. To measure the severity of abnormal pain sensations, 9 characteristics (eg, numbness, itching, and tingling) were assessed on a 5-point severity scale. A total pain score was calculated by summing up the total score for severity of pain in muscles and joints, and the total score for severity of abnormal pain sensations. A logarithmic transformation was applied to the total pain score because of its skewed distribution.
      In the other secondary impairments category, the presence of 7 secondary impairments, which were pressure sores, urinary tract infections, pulmonary infections, neurogenic heterotopic ossification, edema, hypotension, and autonomic dysreflexia, was reported as absent (score: 0) or present (score: 1). A total sum score was computed by adding up the 7 individual scores.

      Social support

      Social support was measured with the Social Support List, which is a reliable and valid short version of the Social Support List-Interactions, consisting of 12 items on 3 scales: everyday social support, support in problem situations, and esteem support.
      • Kempen G.I.
      • van Eijk L.M.
      The psychometric properties of the SSL-12-I, a short scale for measuring social support in the elderly.
      Possible item scores range from 1 (seldom or never) to 4 (very often). A total score was calculated by adding up all 12 items (range, 12–48).

      Demographic characteristics

      These characteristics included age, sex, education level (low, middle, high), marital status (being married or living together vs living alone), and having children (yes, no).

      Statistical Analyses

      Random coefficient analysis (multilevel analysis)
      • Twisk J.W.
      Applied longitudinal data analysis for epidemiology A practical guide.
      • Rasbash J.
      • Charlton C.
      • Browne W.J.
      • Healy M.
      • Cameron B.
      MLwiN Version 2.02 manual.
      was used to study the course of mental health between the start of active rehabilitation and 5 years after discharge. Mental health was the dependent variable and time was the only determinant, entered in the model as a set of 5 dummy variables. In order to examine the course of mental health in each time period, different dummy variables were used as a reference. The SD of the general Dutch population (17.4) was used to calculate the effect size of changes in mental health.
      • Aaronson N.K.
      • Muller M.
      • Cohen P.D.
      • et al.
      Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations.
      Trajectories of mental health were determined by fitting a latent class growth mixture model to the data,
      • Hoekstra T.
      • Barbosa-Leiker C.
      • Koppes L.L.
      • Twisk J.W.
      Developmental trajectories of body mass index throughout the life course: an application of latent class growth (mixture) modelling.
      • Jung T.
      • Wickrama K.A.
      An introduction to latent class growth analysis and growth mixture modeling.
      • Muthén L.K.
      • Muthén B.O.
      Mplus user's guide.
      • Muthén B.
      • Muthén L.K.
      Integrating person-centered and variable centered analyses: growth mixture modeling with latent trajectory classes.
      using the Mplus 6.1 software program.
      • Muthén L.K.
      • Muthén B.O.
      Mplus user's guide.
      ,a Latent class growth models are based on regression and structural equation models. The aim is to capture heterogeneity in the course of mental health in an optimal number of subgroups, each with a unique trajectory. Each subgroup has its own growth parameters (ie, intercept, slope, quadratic slope). The optimal model is a model where individuals within a subgroup are most similar to each other and most different to individuals in other subgroups. To determine the optimal number of trajectories, the Bayesian information criterion (BIC) and the bootstrap likelihood ratio test (BLRT) were used. Lower values of the BIC and a nonsignificant P value of the BLRT indicate a better model fit.
      • Nylund K.
      • Asparouhov T.
      • Muthén B.O.
      Deciding on the number of classes in latent class analysis and growth mixture modelling: a Monte Carlo simulation study.
      Participants were assigned to the trajectory to which they had the highest probability of belonging to, by the use of posterior probabilities.
      Once participants were assigned to a trajectory, multinomial regression analysis was applied to determine which predictors could discriminate between the trajectories. The different trajectories were the dependent variable and demographics; lesion characteristics, and physical and social factors were entered as independent predictors. Significance was set at a P<.05. SPSS statistical program for Windows (version 17.0)b was used for the multinomial regression analyses.

      Results

      Respondent Characteristics

      At the start of active rehabilitation, 225 persons with SCI were included in the study. One hundred and fifty-five persons participated 3 months after, 198 at discharge, 156 at 1-year discharge, 99 at 2-years discharge, and 144 persons participated 5 years after discharge. Details have been reported elsewhere.
      • van Leeuwen C.M.
      • Post M.W.
      • Hoekstra T.
      • et al.
      Trajectories in the course of life satisfaction after spinal cord injury: identification and predictors.
      • de Groot S.
      • Dallmeijer A.J.
      • Post M.W.
      • et al.
      Demographics of the Dutch multicenter prospective cohort study ‘Restoration of mobility in spinal cord injury rehabilitation’.
      A total of 206 persons completed at least 2 measurements and were included in the analyses. Participants' characteristics are displayed in table 1.
      Table 1Descriptive Characteristics at the Start of Active Rehabilitation of Participants (N=206)
      Characteristicsn%
      Sex (male)15374.3
      Marital status (together)15474.8
      Children (yes)10852.4
      Education
       Low6833.0
       Middle10249.5
       High3617.5
      Type of injury
       Incomplete paraplegia3818.4
       Complete paraplegia9244.7
       Incomplete tetraplegia2612.6
       Complete tetraplegia5024.3
      Cause of injury (traumatic)15474.8
      MedianIQR
      Age (y)41.028–54.3
      Time since injury (d)75.052–115
      Abbreviation: IQR, interquartile range.

      Course of Mental Health

      Figure 1 shows that mental health increased between the start of active rehabilitation and 3 months later (effect size=.37), remained stable between 3 months after start and 2 years after discharge, and increased again between 2 years and 5 years after discharge (effect size=.27). The effect sizes were both small.
      • Cohen J.
      Statistical power analysis for the behavioural sciences.
      Figure thumbnail gr1
      Fig 1Average course of mental health (N=206). The numbers below the curve refer to the SE of the nonstandardized regression coefficient. The model had a random intercept and 3 time dummies had random slopes. *Significant increase.

      Identifying Mental Health Trajectories

      Table 2 shows the results of several model fit indices comparing a 1 to 6 class solution. The fit indices favored the 6-class solution (see table 2). However, this model consisted of several trajectories with very small subgroups, making the model unstable and difficult to interpret. The main difference between the 4-class and the 5-class models was that the intermediate trajectory was not identified in the 4-class model, which resulted in higher scores for the low trajectory, and lower scores for the high trajectory. Therefore, the 5-class model was chosen as best representing the different trajectories of mental health.
      Table 2Model Fit Indices for the Selection of the Number of Trajectories
      No. of TrajectoriesBICBRLT (P)
      k=17899.093NA
      k=27694.312<.001
      k=37633.316<.001
      k=47621.328<.001
      k=57618.562<.001
      k=67615.608<.001
      Abbreviation: NA, not applicable.
      In the 5-class model (fig 2 and table 3), the first trajectory (comprising 51.5% of the sample) is characterized by stable high scores. The second trajectory shows stable low scores during inpatient rehabilitation and slight improvements, but still low mental health scores, after discharge from inpatient rehabilitation (4.4%). The third trajectory reveals stable intermediate scores during inpatient rehabilitation and slight improvements after discharge from inpatient rehabilitation (29.1%). The fourth trajectory (13.1%) shows a trajectory with rapid improvements during inpatient rehabilitation and stable scores after discharge from inpatient rehabilitation. The last and smallest trajectory (1.9%) is characterized by deterioration in mental health.
      Figure thumbnail gr2
      Fig 2Estimated trajectories in the course of mental health.
      Table 3Five Trajectories, Their Posterior Probabilities, and Mental Health Scores
      Trajectoryn (%)Mean Posterior ProbabilitiesStart3mo AfterDischarge1y After2y After5y After
      High mental health106 (51.5).94480 (72–88)84 (79–92)84 (76–92)80 (76–88)84 (76–89)88 (76–92)
      Low mental health9 (4.4).95842 (30–54)44 (32–46)36 (24–48)32 (26–36)48 —60 (42–66)
      Intermediate60 (29.1).82462 (52–68)58 (49–68)60 (56–68)64 (52–68)60 (52–76)72 (66–80)
      Early recovery27 (13.1).82040 (28–44)68 (56–79)76 (66–80)74 (68–84)72 (64–84)74 (58–86)
      Deterioration4 (1.9).89064 (54–74)72 —76 (70–79)46 (38–48)20 —20 —
      NOTE. Values are median (interquartile range) or as otherwise indicated. The interquartile range cannot be estimated in some cases, because the total number of respondents was too small (marked with —).
      Figure 3 shows the distribution of the mental health scores in the 5 trajectories 5 years after discharge. Around 36.8% of the persons obtained a score lower than or equal to the cutoff point of 72, and 13.9% obtained a score lower than or equal to the cutoff point of 60. Figure 3 also shows how persons with SCI score in comparison with the general Dutch population. Most persons in the high mental health trajectory scored above the average score of the general Dutch population. However, 50% of the persons with SCI scored below the average score of the general Dutch population. Moreover, 36.8% scored below the average score of Dutch persons with 1 chronic disease (eg, diabetes, rheumatoid arthritis, and pulmonary obstructive diseases), and 26.4% scored below the average score of Dutch persons with more than 1 chronic disease.
      Figure thumbnail gr3
      Fig 3Distribution of mental health scores in the 5 trajectories at 5 years after discharge. The 2 red lines are the reference lines for the cutoff points of 72 (indicating mental health problems) and 60 (indicating severe mental health problems). The dotted black lines are the reference lines for the Dutch population: 76.6 (general Dutch population), 75.5 (Dutch persons with 1 chronic condition), and 68.9 (Dutch person with >1 chronic disease). The dot in the intermediate trajectory is an outlier.

      Predictors to Distinguish Between Mental Health Trajectories

      Demographics, lesion characteristics, and physical and social factors for each mental health trajectory are displayed in table 4. Table 5 shows that pain, sex, and education level at the start of active rehabilitation were used as predictors to distinguish between mental health trajectories. The high mental health trajectory was chosen as the reference group. Compared with persons in the high mental health trajectory, persons in the early recovery trajectory and intermediate trajectory were more likely to have higher pain levels and to be a woman. Moreover, compared with persons in the high mental health trajectory, persons in the low mental health trajectory, early recovery trajectory, and deterioration trajectory were more likely to have a lower education level. Pain, sex, and education level only explained a small amount of the variance to distinguish between mental health trajectories (R2=.18).
      Table 4Characteristics Per Mental Health Trajectory (N=206)
      CharacteristicLow (n=9)Early Recovery (n=27)Deterioration (n=4)Intermediate (n=60)High (n=106)
      Demographics
       Male8 (5.2)15 (9.8)4 (2.6)38 (24.8)88 (57.5)
       Female1 (1.9)12 (22.6)NA (0.0)22 (14.4)18 (44.0)
       Living together7 (4.5)23 (14.9)4 (2.6)46 (29.9)74 (48.1)
       Living alone2 (3.8)4 (7.7)NA (0.0)14 (26.9)32 (61.5)
       Children4 (3.7)15 (13.9)2 (1.9)34 (31.5)53 (49.1)
       No children5 (5.1)12 (12.2)2 (2.0)26 (26.5)53 (54.1)
       Education low7 (10.3)11 (16.2)2 (2.9)19 (27.9)29 (42.6)
       Education middle2 (2.0)14 (13.7)2 (2.0)27 (26.5)57 (55.9)
       Education highNA (0.0)2 (5.6)NA (0.0)14 (38.9)20 (55.6)
       Age (y)32 (29–59)45 (28–53)47 (29–55)43 (30–56)41 (37–52)
      Lesion characteristics
       Paraplegia4 (3.1)21 (16.2)3 (2.3)31 (23.8)71 (54.6)
       Tetraplegia5 (6.6)6 (7.9)1 (1.3)29 (38.2)35 (46.1)
       Complete5 (3.5)22 (15.5)4 (2.8)40 (28.2)71 (50.0)
       Incomplete4 (6.3)5 (7.8)NA (0.0)20 (31.3)35 (54.7)
       Traumatic injury7 (5.3)21 (15.9)2 (1.5)19 (14.4)83 (63.9)
       Nontraumatic injury2 (2.7)6 (8.1)2 (2.7)41 (55.4)23 (31.1)
      Physical factors
       Functional independence29.5 (13–49.8)36 (30–59)35.5 (33.5–37.5)37 (24–47)37.5 (28.3–55)
       Pain4 (1–6.5)5 (4–6)4 (3.3–4)4 (3–5)4 (2–5)
       Secondary impairments1 (0.5–2.5)1 (0–2)2 (0.5–2)1 (0–2)1 (1–2)
      Social factors
       Total social support35 (32–42.5)37 (32–40)35.5 (26.8–42)34 (29.8–39.3)36 (32–39)
       Everyday social support12.5 (10.5–13)12 (10–13)12.5 (10.3–14.8)12 (10–13)12 (11–14)
       Support in problem situations13 (10.5–16)13 (12–15)11 (7.5–15.3)11 (9–13.3)12 (10–14)
       Esteem support10.5 (8.3–13.8)11 (8–13)11 (9–12)10 (9–12)11 (10–13)
      NOTE. Values are n (% of total sample) or median (interquartile range). All variables were measured at the start of active rehabilitation.
      Abbreviation: NA, not applicable.
      Table 5Multinominal Regression Analyses to Distinguish Between the 5 Mental Health Trajectories (N=206)
      VariablesLow vs HighEarly Recovery vs HighDeterioration vs HighIntermediate vs High
      DemographicsOdds Ratio95% CIOdds Ratio95% CIOdds Ratio95% CIOdds Ratio95% CI
      Sex (female)0.610.07–5.203.911.57–9.74NSNS2.831.36–5.88
      Age (higher)NSNSNSNSNSNSNSNS
      Children (yes)NSNSNSNSNSNSNSNS
      Marital status (together)NSNSNSNSNSNSNSNS
      Education (higher)0.140.03–0.640.580.30–1.100.380.07–1.991.000.63–1.59
      Lesion characteristics
       Lesion level (paraplegia)NSNSNSNSNSNSNSNS
       Completeness (complete)NSNSNSNSNSNSNSNS
       Cause of injury (traumatic)NSNSNSNSNSNSNSNS
      Physical variables
       Functional independence (higher)NSNSNSNSNSNSNSNS
       Pain (higher)0.920.38–2.202.841.38–5.830.740.23–2.461.510.95–2.40
       Secondary impairments (higher)NSNSNSNSNSNSNSNS
      Social variables
       Total social support (higher)NSNSNSNSNSNSNSNS
       Everyday social support (higher)NSNSNSNSNSNSNSNS
       Support problem situations (higher)NSNSNSNSNSNSNSNS
       Esteem support (higher)NSNSNSNSNSNSNSNS
      Abbreviations: CI, confidence interval; NS, not significant.

      Discussion

      The present study identified 5 distinct mental health trajectories in the period between the start of active SCI rehabilitation and 5 years after discharge: high mental health trajectory, low mental health trajectory, early recovery trajectory, intermediate trajectory, and deterioration trajectory. Pain, sex and education level were found to be predictors to distinguish between mental health trajectories, but explained only a small part of the variance.

      Course of Mental Health

      To our knowledge, this is the first study examining trajectories in the course of mental health in persons with SCI. In general, mental health improved over time, but the results clearly showed that persons differ in their rate and pace of mental recovery after an SCI. At 5 years after discharge, about one third of the persons still had mental health problems.
      The percentage of persons at risk of severe mental health problem decreased from 35.5% to 13.9% between the start of active rehabilitation and 5 years after discharge. This last percentage is somewhat higher than the percentages between 8.8% and 12% of persons with SCI with probable major depression disorder found in studies using the Patient Health Questionnaire-9.
      • Post M.W.
      • van Leeuwen C.M.
      Psychosocial issues in spinal cord injury: a review.
      In contrast, the percentage of 13.9% is somewhat lower than the percentage of persons with SCI with probable major depression disorder 5 years postdischarge found in the SCI Model Systems (17.6%–18.1%).
      • Hoffman J.M.
      • Bombardier C.H.
      • Graves D.E.
      • Kalpakjian C.Z.
      • Krause J.S.
      A longitudinal study of depression from 1 to 5 years after spinal cord injury.
      • Richardson E.J.
      • Richards J.S.
      Factor structure of the PHQ-9 screen for depression across time since injury among persons with spinal cord injury.
      The use of another measurement instrument and cutoff point might explain the difference in outcome.
      In comparison with the Dutch population,
      • Aaronson N.K.
      • Muller M.
      • Cohen P.D.
      • et al.
      Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations.
      50% of the persons with SCI had lower mental health than the general population, and about one third scored lower on mental health than Dutch persons with 1 chronic condition 5 years after discharge. It is remarkable that a substantial proportion of persons with SCI in our cohort showed mental health scores in the same range or scored even higher than the general Dutch population. This may be a result of certain psychological factors, such as the ability to experience posttraumatic growth. Because research in SCI has concentrated more on determinants of depression than on determinants of mental health, further study is necessary to better understand this.

      Mental Health Trajectories in Other Diagnostic Groups

      Our results are in line with studies examining different trajectories in the course of mental health in other diagnostic groups.
      • Kaptein K.I.
      • de Jonge P.
      • van den Brink R.H.
      • Korf J.
      Course of depressive symptoms after myocardial infarction and cardiac prognosis: a latent class analysis.
      • Dew M.A.
      • Myaskovsky L.
      • Switzer G.E.
      • DiMartini A.F.
      • Schulberg H.C.
      • Kormos R.L.
      Profiles and predictors of the course of psychological distress across four years after heart transplantation.
      • Helgeson V.S.
      • Snyder P.
      • Seltman H.
      Psychological and physical adjustment to breast cancer over 4 years: identifying distinct trajectories of change.

      Norton S, Sacker A, Young A, Done J. Distinct psychological distress trajectories in rheumatoid arthritis: findings from an inception cohort. J Psychosom Res 201;71:290-5.

      Studies investigating the course of depressive symptoms after myocardial infarction
      • Kaptein K.I.
      • de Jonge P.
      • van den Brink R.H.
      • Korf J.
      Course of depressive symptoms after myocardial infarction and cardiac prognosis: a latent class analysis.
      and the course of psychological distress after heart transplantation
      • Dew M.A.
      • Myaskovsky L.
      • Switzer G.E.
      • DiMartini A.F.
      • Schulberg H.C.
      • Kormos R.L.
      Profiles and predictors of the course of psychological distress across four years after heart transplantation.
      also found 5 trajectories, which are more or less similar to the trajectories we found. Studies examining psychological and physical adjustment to breast cancer
      • Helgeson V.S.
      • Snyder P.
      • Seltman H.
      Psychological and physical adjustment to breast cancer over 4 years: identifying distinct trajectories of change.
      and psychological distress in rheumatoid arthritis identified 4 trajectories.

      Norton S, Sacker A, Young A, Done J. Distinct psychological distress trajectories in rheumatoid arthritis: findings from an inception cohort. J Psychosom Res 201;71:290-5.

      In these 2 studies, the low life satisfaction trajectory and the intermediate trajectory were combined. The percentages of persons having mental health problems in these 4 studies
      • Kaptein K.I.
      • de Jonge P.
      • van den Brink R.H.
      • Korf J.
      Course of depressive symptoms after myocardial infarction and cardiac prognosis: a latent class analysis.
      • Dew M.A.
      • Myaskovsky L.
      • Switzer G.E.
      • DiMartini A.F.
      • Schulberg H.C.
      • Kormos R.L.
      Profiles and predictors of the course of psychological distress across four years after heart transplantation.
      • Helgeson V.S.
      • Snyder P.
      • Seltman H.
      Psychological and physical adjustment to breast cancer over 4 years: identifying distinct trajectories of change.

      Norton S, Sacker A, Young A, Done J. Distinct psychological distress trajectories in rheumatoid arthritis: findings from an inception cohort. J Psychosom Res 201;71:290-5.

      were in the same range as what we found (32%–57%).
      In sum, it seems that similar mental health trajectories were identified in different diagnostic groups. This interesting finding could imply that studies in the field of adjustment to disability and/or chronic illness might benefit from a more general approach of examining how persons react to a stressful event.

      Predictors of Mental Health Trajectories

      Pain, sex, and education level were found to be predictors to distinguish between trajectories in mental health. This is in line with results of other studies. In a longitudinal study, 70% of the persons with severe pain (about 45% of the total SCI study population) suffered from mental health distress.
      • Anke A.G.
      • Stenehjem A.E.
      • Stanghelle J.K.
      Pain and life quality within 2 years of spinal cord injury.
      Moreover, an earlier review concluded that being a woman and having good communication skills predict favorable psychological outcome.
      • North N.T.
      The psychological effects of spinal cord injury: a review.
      Social support was not a predictor of mental health, which was in contrast with earlier findings.
      • North N.T.
      The psychological effects of spinal cord injury: a review.
      • Müller R.
      • Peter C.
      • Cieza A.
      • Geyh S.
      The role of social support and social skills in people with spinal cord injury—a systematic review of the literature.
      • van Leeuwen C.M.
      • Kraaijeveld S.
      • Lindeman E.
      • Post M.W.
      Associations between psychological factors and quality of life ratings in persons with spinal cord injury: a systematic review.
      • Peter C.
      • Muller R.
      • Cieza A.
      • Geyh S.
      Psychological resources in spinal cord injury: a systematic literature review.
      At the start of active rehabilitation, persons in all 5 trajectories scored high on social support, which might explain why social support did not turn out to be a predictor of mental health. Moreover, lesion characteristics were not a predictor of mental health trajectories. Other studies also found that the impairment itself is not related to low mental health, but the impact of the impairment on activities and participation is related to low mental health.
      • van Leeuwen C.M.
      • Post M.W.
      • Westers P.
      • et al.
      Relationships between activities, participation, personal factors, mental health, and life satisfaction in persons with spinal cord injury.
      However, the core message is that demographics, lesion characteristics, and physical and social factors at the start of rehabilitation cannot distinguish between mental health trajectories after SCI. We found a similar result for predicting life satisfaction trajectories.
      • van Leeuwen C.M.
      • Post M.W.
      • Hoekstra T.
      • et al.
      Trajectories in the course of life satisfaction after spinal cord injury: identification and predictors.
      Psychological factors, such as self-efficacy, neuroticism, and purpose in life, might play a major role in mental health after SCI.
      • Post M.W.
      • van Leeuwen C.M.
      Psychosocial issues in spinal cord injury: a review.
      • Kennedy P.
      • Lude P.
      • Elfström M.L.
      • Smithson E.
      Cognitive appraisals, coping and quality of life outcomes: a multi-centre study of spinal cord injury rehabilitation.
      • van Leeuwen C.M.
      • Kraaijeveld S.
      • Lindeman E.
      • Post M.W.
      Associations between psychological factors and quality of life ratings in persons with spinal cord injury: a systematic review.
      • van Leeuwen C.M.
      • Post M.W.
      • Westers P.
      • et al.
      Relationships between activities, participation, personal factors, mental health, and life satisfaction in persons with spinal cord injury.
      • Chevalier Z.
      • Kennedy P.
      • Sherlock O.
      Spinal cord injury, coping and psychological adjustment: a literature review.
      Unfortunately, psychological factors were not assessed during inpatient rehabilitation in our cohort study.

      Study Limitations

      A limitation to the present study was that only Dutch persons with an SCI and aged between 18 and 65 years with expected permanent wheelchair dependency admitted to a rehabilitation center were included. This influences the representativeness of the population and thereby the degree to which the results of our study can be generalized to the whole population of persons with SCI. Second, persons who were older and had a nontraumatic injury had a higher chance to drop out of the study. The dropout rate was, however, low, and both random coefficient analysis and latent class growth mixture modeling allow the number of observations per person to vary.
      • Twisk J.W.
      Applied longitudinal data analysis for epidemiology A practical guide.
      • Muthén B.
      • Muthén L.K.
      Integrating person-centered and variable centered analyses: growth mixture modeling with latent trajectory classes.
      Third, a limitation of Latent Class Growth Modeling is that there is not a standard method for determining the best number of trajectories. Disagreement between fit indices is common, making the decision for the final model somewhat complicated.
      • Hoekstra T.
      • Barbosa-Leiker C.
      • Koppes L.L.
      • Twisk J.W.
      Developmental trajectories of body mass index throughout the life course: an application of latent class growth (mixture) modelling.
      • Nylund K.
      • Asparouhov T.
      • Muthén B.O.
      Deciding on the number of classes in latent class analysis and growth mixture modelling: a Monte Carlo simulation study.
      However, we followed the advice of Nylund et al
      • Nylund K.
      • Asparouhov T.
      • Muthén B.O.
      Deciding on the number of classes in latent class analysis and growth mixture modelling: a Monte Carlo simulation study.
      in examining multiple fit indices, in combination with clinically interpreting the models. By examining multiple criteria, a more secure decision for the final model can be made. Finally, we did not assess psychological factors at the start of active rehabilitation. Therefore, we could not test whether psychological factors better predict mental health in persons with SCI.

      Conclusions

      Persons with SCI differ in their rate and pace of mental recovery after an SCI. In general, the severity of mental health problems in persons with an SCI seems to decrease somewhat over time. However, about one third of the persons still had moderate to severe mental health problems 5 years after discharge from rehabilitation.
      Persons who have mental health problems at the start of active rehabilitation and do not show recovery in their mental health during rehabilitation are at risk for poor long-term adjustment. Moreover, there is a small group of persons who showed recovery in their mental health during the first 3 months of active rehabilitation, but severe decreases in their mental health after these 3 months. This group is likely to suffer major depression disorder, because mental health levels are dramatically low after discharge from rehabilitation. In another study, we found that low mental health is related to psychological factors, such as a high score on neuroticism and a low score on self-efficacy.
      • van Leeuwen C.M.
      • Post M.W.
      • Westers P.
      • et al.
      Relationships between activities, participation, personal factors, mental health, and life satisfaction in persons with spinal cord injury.
      Finally, for future research, we recommend examining whether psychological factors are better predictors in distinguishing between mental health trajectories, because demographics, lesion characteristics, and physical and social factors can only predict a small part of the variance in mental health.
      • a
        Muthén & Muthén, 3463 Stoner Ave, Los Angeles, CA 90066.
      • b
        SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

      Acknowledgments

      We thank the participating Dutch rehabilitation centers and the research assistants in these centers who collected all data: Rehabilitation Center De Hoogstraat (Utrecht), Reade, Centre for Rehabilitation and Rheumatology (Amsterdam), Rehabilitation Center Het Roessingh (Enschede), Adelante (Hoensbroek), Rehabilitation Center Sint Maartenskliniek (Nijmegen), Center for Rehabilitation–Beatrixoord (Haren), Rehabilitation Center Heliomare (Wijk aan Zee), and Rehabilitation Center Rijndam (Rotterdam).

      References

        • Dijkers M.P.
        Quality of life of individuals with spinal cord injury: a review of conceptualization, measurement, and research findings.
        J Rehabil Res Dev. 2005; 42: 87-110
        • Post M.
        • Noreau L.
        Quality of life after spinal cord injury.
        J Neurol Phys Ther. 2005; 29: 139-146
        • Post M.W.
        • van Leeuwen C.M.
        Psychosocial issues in spinal cord injury: a review.
        Spinal Cord. 2012; 50: 382-389
        • Kennedy P.
        • Lude P.
        • Elfström M.L.
        • Smithson E.
        Cognitive appraisals, coping and quality of life outcomes: a multi-centre study of spinal cord injury rehabilitation.
        Spinal Cord. 2010; 48: 762-769
        • Kennedy P.
        • Rogers B.A.
        Anxiety and depression after spinal cord injury: a longitudinal analysis.
        Arch Phys Med Rehabil. 2000; 81: 932-937
        • Hancock K.M.
        • Craig A.R.
        • Dickson H.G.
        • Chang E.
        • Martin J.
        Anxiety and depression over the first year of spinal cord injury: a longitudinal study.
        Paraplegia. 1993; 31: 349-357
        • Cushman L.A.
        • Dijkers M.P.
        Depressed mood in spinal cord injured patients: staff perceptions and patient realities.
        Arch Phys Med Rehabil. 1990; 71: 191-196
        • Hoffman J.M.
        • Bombardier C.H.
        • Graves D.E.
        • Kalpakjian C.Z.
        • Krause J.S.
        A longitudinal study of depression from 1 to 5 years after spinal cord injury.
        Arch Phys Med Rehabil. 2011; 92: 411-418
        • Richardson E.J.
        • Richards J.S.
        Factor structure of the PHQ-9 screen for depression across time since injury among persons with spinal cord injury.
        Rehabil Psychol. 2008; 53: 243-249
        • van Leeuwen C.M.
        • Post M.W.
        • Hoekstra T.
        • et al.
        Trajectories in the course of life satisfaction after spinal cord injury: identification and predictors.
        Arch Phys Med Rehabil. 2011; 92: 207-213
        • North N.T.
        The psychological effects of spinal cord injury: a review.
        Spinal Cord. 1999; 37: 671-679
        • Müller R.
        • Peter C.
        • Cieza A.
        • Geyh S.
        The role of social support and social skills in people with spinal cord injury—a systematic review of the literature.
        Spinal Cord. 2012; 50: 94-106
        • van Leeuwen C.M.
        • Kraaijeveld S.
        • Lindeman E.
        • Post M.W.
        Associations between psychological factors and quality of life ratings in persons with spinal cord injury: a systematic review.
        Spinal Cord. 2012; 50: 174-187
        • Peter C.
        • Muller R.
        • Cieza A.
        • Geyh S.
        Psychological resources in spinal cord injury: a systematic literature review.
        Spinal Cord. 2012; 50: 188-201
        • de Groot S.
        • Dallmeijer A.J.
        • Post M.W.
        • et al.
        Demographics of the Dutch multicenter prospective cohort study ‘Restoration of mobility in spinal cord injury rehabilitation’.
        Spinal Cord. 2006; 44: 668-675
        • Ware J.E.
        • Snow K.K.
        • Kosinski M.
        • Gandek B.
        SF-36 Health Survey: manual and interpretation guide.
        The Health Institute, New England Medical Center, Boston1993
        • Hoeymans N.
        • Garssen A.A.
        • Westert G.P.
        • Verhaak P.F.
        Measuring mental health of the Dutch population: a comparison of the GHQ-12 and the MHI-5.
        Health Qual Life Outcomes. 2004; 2: 23
        • Perenboom R.
        • Oudshoorn K.
        • van Herten L.
        • Hoeymans N.
        • Bijl R.
        [Life expectancy in good mental health: establishing cut-offs for the MHI-5 and GHQ-12] [Dutch].
        TNO, Leiden2000
        • van Leeuwen Jr, C.M.
        • van der Woude L.H.
        • Post M.W.
        Validity of the mental health subscale of the SF-36 in persons with spinal cord injury.
        Spinal Cord. 2012 Apr 10; ([Epub ahead of print])
        • Maynard Jr, F.M.
        • Bracken M.B.
        • Creasey G.
        • et al.
        International standards for neurological and functional classification of spinal cord injury.
        Spinal Cord. 1997; 35: 266-274
        • Hall K.M.
        • Cohen C.E.
        • Wright J.
        • Call M.
        • Werner P.
        Characteristics of the Functional Independence Measure in traumatic spinal cord injury.
        Arch Phys Med Rehabil. 1999; 80: 1471-1476
        • Kempen G.I.
        • van Eijk L.M.
        The psychometric properties of the SSL-12-I, a short scale for measuring social support in the elderly.
        Soc Ind Res. 1995; 35: 303-312
        • Twisk J.W.
        Applied longitudinal data analysis for epidemiology.
        4th ed. Cambridge Univ Pr, Cambridge2003
        • Rasbash J.
        • Charlton C.
        • Browne W.J.
        • Healy M.
        • Cameron B.
        MLwiN Version 2.02 manual.
        Centre for Multilevel Modelling, Univ of Bristol, Bristol2005
        • Aaronson N.K.
        • Muller M.
        • Cohen P.D.
        • et al.
        Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations.
        J Clin Epidemiol. 1998; 51: 1055-1068
        • Hoekstra T.
        • Barbosa-Leiker C.
        • Koppes L.L.
        • Twisk J.W.
        Developmental trajectories of body mass index throughout the life course: an application of latent class growth (mixture) modelling.
        Longitudinal and Life Course Studies. 2011; 2: 319-330
        • Jung T.
        • Wickrama K.A.
        An introduction to latent class growth analysis and growth mixture modeling.
        Social and Personality Psychology Compass. 2008; 2: 302-317
        • Muthén L.K.
        • Muthén B.O.
        Mplus user's guide.
        6th ed. Muthén and Muthén, Los Angeles2010
        • Muthén B.
        • Muthén L.K.
        Integrating person-centered and variable centered analyses: growth mixture modeling with latent trajectory classes.
        Alcohol Clin Exp Res. 2000; 24: 882-891
        • Nylund K.
        • Asparouhov T.
        • Muthén B.O.
        Deciding on the number of classes in latent class analysis and growth mixture modelling: a Monte Carlo simulation study.
        Structural Equation Modeling. 2007; 14: 535-569
        • Cohen J.
        Statistical power analysis for the behavioural sciences.
        2nd ed. Lawrence Erlbaum Associates, Hillsdale1988
        • Kaptein K.I.
        • de Jonge P.
        • van den Brink R.H.
        • Korf J.
        Course of depressive symptoms after myocardial infarction and cardiac prognosis: a latent class analysis.
        Psychosom Med. 2006; 68: 662-668
        • Dew M.A.
        • Myaskovsky L.
        • Switzer G.E.
        • DiMartini A.F.
        • Schulberg H.C.
        • Kormos R.L.
        Profiles and predictors of the course of psychological distress across four years after heart transplantation.
        Psychol Med. 2005; 35: 1215-1227
        • Helgeson V.S.
        • Snyder P.
        • Seltman H.
        Psychological and physical adjustment to breast cancer over 4 years: identifying distinct trajectories of change.
        Health Psychol. 2004; 23: 3-15
      1. Norton S, Sacker A, Young A, Done J. Distinct psychological distress trajectories in rheumatoid arthritis: findings from an inception cohort. J Psychosom Res 201;71:290-5.

        • Anke A.G.
        • Stenehjem A.E.
        • Stanghelle J.K.
        Pain and life quality within 2 years of spinal cord injury.
        Paraplegia. 1995; 33: 555-559
        • van Leeuwen C.M.
        • Post M.W.
        • Westers P.
        • et al.
        Relationships between activities, participation, personal factors, mental health, and life satisfaction in persons with spinal cord injury.
        Arch Phys Med Rehabil. 2012; 93: 82-89
        • Chevalier Z.
        • Kennedy P.
        • Sherlock O.
        Spinal cord injury, coping and psychological adjustment: a literature review.
        Spinal Cord. 2009; 47: 778-782