Abstract
Previous studies suggest that health-related quality of life (HRQoL) is impaired after a traumatic spinal cord injury (TSCI) and may be worse with older age. This study determines whether the expectations to achieve normal HRQoL in Canadians after a TSCI is indeed influenced by older age. A prospective observational study was conducted on adult patients admitted acutely at a single level-1 trauma center after a TSCI. We assessed HRQoL using the SF-36 physical and mental component summary (PCS and MCS) scores obtained one year post injury. Using Canadian normative HRQoL data matched for age and sex, we defined normal PCS and MCS as a score within 2 standard deviations with respect to the normative Canadian mean. We then conducted logistic regression models to determine the relationship between age at the time of injury and the likelihood of achieving normal PCS and MCS, while controlling for confounding variables. Overall, 39.3% of individuals displayed normal PCS, whereas 80.4% displayed normal MCS. When adjusted for confounders, older age remained significantly associated with increased likelihood of achieving normal PCS (Odds Ratio: 1.03; 95% Confidence Interval: 1.01–1.06; P = 0.002). We observed no association between age and achieving normal MCS. A significant proportion of individuals can achieve a normal HRQoL similar to their healthy peers following a TSCI, particularly for the mental component. When compared to younger individuals, older individuals are more likely to achieve normal PCS and present a similar likelihood for achieving normal MCS.
Introduction
Traumatic spinal cord injury (TSCI) can be associated with lifelong multi-systemic impairments and major psychosocial challenges. These limitations can impair the patient's participation and overall level of activity, which in turn diminishes quality of life (QoL) and overall well-being. 1 -4 Given that older individuals represent an increasing proportion of patients with acute TSCI, understanding the impact of age on QoL after TSCI is critical in establishing accurate prognosis and planning adequate rehabilitation care for this specific population. 5 While it is assumed that older patients are able to show good resilience after injury, most studies have nonetheless reported that age negatively affects QoL in individuals with TSCI. 6 –11 Associating poor QoL with older age could entail clinical repercussions that, for example, can lead surgeons to become more reluctant to opt for surgical treatments in older patients after TSCI. 12
Unfortunately, previous studies have failed to account for the normal age- and sex-related QoL variations in the general population when interpreting their data. Even among the general population, QoL tends to decline with aging. 13,14 Canadian normative QoL data obtained by Hopman and colleagues 14,15 through a nationwide survey show that older Canadians have on average a lower physical QoL than their younger peers and that women experience slightly lower QoL than men. Therefore, studying the expectations to achieving normal QoL after TSCI requires that each individual's QoL be compared with normative QoL data of the same age group and sex.
To our knowledge, this present study is the first to investigate the likelihood that individuals with TSCI achieve normal QoL when compared with able-bodied peers of the same age and sex. We also aim to determine whether older age affects the likelihood of achieving normal QoL one year after TSCI.
Methods
Participants
A prospective observational cohort study was conducted in patients with TSCI admitted to a single level-1 Canadian trauma center (Hôpital du Sacré-Coeur de Montréal, Montréal, Canada) between April 2010 and February 2020. We included patients aged 18 or older who had sustained a TSCI (including cauda equina syndrome) with an initial American Spinal Injury Association Impairment Scale (AIS) grade A to D with any neurological level of injury (NLI) from C1 to S5. We excluded patients with missing data pertaining to initial neurological examination. We secured approval from our institution's research ethics board, and all patients provided informed and voluntary consent before enrollment.
Data collection
Main independent variable
Age at the time of injury, our main independent variable, was collected on admission and was treated as a continuous variable.
Confounding variables
Data pertaining to sex, initial neurological deficit, trauma severity from the Injury Severity Score (ISS), time since injury, comorbidities from the Charlson Comorbidity Index (CCI), living arrangement (alone vs. not alone), and highest attained education level were collected as potential confounders.
The neurological examination was performed in acute care by a trained physical medicine and rehabilitation physician or spine surgeon following the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI). 16 The NLI and AIS grade were used to describe the initial neurological deficit. The NLI is defined as the most caudal segment of the spinal cord with both preserved motor and sensory function and was grouped as high cervical (C1–C4), low cervical (C5–C8), thoracic (T1–T8), thoracolumbar (T9–L1), and lumbosacral (L2–S5). The AIS grade is determined by the degree of function preservation below the level of injury based on the rectal examination. It ranges from AIS grade A (no motor or sensory function below level of injury) to AIS grade D (≥50% of the key muscles below the level NLI have grade ≥3/5). 17
The ISS assesses the overall severity of sustained trauma and was treated as a continuous variable. 18 Time since injury, which measures the number of days between the date of trauma and the date of follow-up where we assessed QoL, was treated as a continuous variable. The CCI assesses the burden of comorbidity by means of a score based on the number and severity of comorbid medical conditions and was dichotomized as 0 (no comorbidity) or ≥1 (presence of at least one comorbidity). 19 Highest attained education level was dichotomized as either high school or lower, or post-high school education.
Outcome variables
The QoL was assessed using the Short Form 36 (SF-36) instrument, a validated 36-item QoL questionnaire that encompasses eight domains: physical functioning, role physical, bodily pain, general health perceptions, vitality, social functioning, role emotional, and mental health. 20 These domains are summarized in two composite scores: the physical component summary (PCS), which measures the overall physical QoL, and the mental component summary (MCS), which measures the overall mental QoL. 21 We obtained the PCS and MCS scores (the primary outcomes) within the six to 12 month period after the injury.
For each patient, the measured PCS and MCS scores were standardized using the normative PCS and MCS mean and standard deviation of Canadian individuals of the same age group and sex. 15 We calculated the Z-score of the PCS (Z-PCS) and MCS (Z-MCS) of each patient using the mean and standard deviation of their corresponding age group and sex. We subsequently determined whether the Z-PCS or Z-MCS was normal based on the following definition: (1) abnormal if a patient's PCS or MCS was more than two standard deviations below the PCS or MCS mean of the corresponding sex and age group in the general Canadian population (i.e., Z-PCS or Z-MCS < -2), and (2) normal if a patient's PCS or MCS was no more than two standard deviations below the mean of the corresponding sex and age group in the general Canadian population (i.e., Z-PCS or Z-MCS ≥ -2).
Statistical analyses
Following the description of baseline characteristics for the entire cohort, we compared baseline characteristics between patients with normal and abnormal Z-PCS and Z-MCS using the Student t test (continuous variables) and χ 2 tests (categorical variables). We subsequently conducted logistic regression analyses to determine the effect of age on achieving normal Z-PCS and Z-MCS, while including covariates significantly associated with achieving normal Z-PCS or Z-MCS in the bivariate analyses.
We conducted two separate logistic regression analyses to determine the effect of age on achieving normal Z-PCS and Z-MCS. Goodness-of-fit was verified with the Hosmer-Lemeshow test and the area under the receiver operating characteristic curve (AUC). Normal distribution of continuous variables was verified with Kolmogorov-Smirnov tests, and appropriate transformations were applied when required.
For all statistical analyses, we set the level of significance at 0.05 and used IBM SPSS Statistics Version 26 software package.
Results
Of the 477 eligible patients, nine were excluded because of an incomplete acute neurological examination. Of the remaining 468 patients, 171 (36.5%) patients were not administered the SF-36 questionnaire between six to 12 months post-injury, a loss to follow-up ratio in line with previous studies in the TSCI population. 22 –24 An additional 17 patients had an incomplete SF-36 questionnaire, which left a total of 280 patients for final analysis (see Fig. 1). The mean age of the cohort was 50.4 ± 19.0 years, 78.8% were male, and most patients (44.0%) had sustained AIS grade D injuries at admission (Table 1). Because more than 20% of the cohort was lost to follow-up, we compared baseline characteristics between patients with and without available follow-up data. We found that a higher proportion of patients lost to follow-up live alone (Table 1).
Flow diagram of patient selection process. SF-36, Short Form-36.
Description of Baseline Characteristics of the Cohort
FU, follow-up; SD, standard deviation; AIS, American Spinal Injury Association Impairment Scale; NLI, Neurological level of injury; ISS, Injury Severity Score; CCI, Charlson Comorbidity Index.
Bold: Variables significant at level of 0.05.
Variable normalized using Log(ISS) transformation.
Comparison of baseline characteristics between patients with normal and abnormal Z-PCS/Z-MCS is found in Table 2. Overall, a total of 110 (39.3%) patients showed a normal PCS score while 225 (80.4%) patients showed a normal MCS score within the first year after injury. Bivariate analyses showed that patients with normal Z-PCS were significantly older than patients with below normal Z-PCS (55.5 ± 18.1 years vs. 45.3 ± 16.9 years; p < 10−5). The group with normal Z-PCS also sustained less severe TSCI (p < 10−9), higher NLI (p < 10−3), lower trauma severity (20.4 ± 5.8 vs. 23.7 ± 10.2; p = 0.01), and had a lower burden of comorbidities (p = 0.04); these variables were included in multi-variate analysis. Age was not significantly different between normal MCS and abnormal MCS groups (p = 0.33).
Comparison of Baseline Characteristics between Patients with Normal and Below Normal Physical Component Summary Z-Score and Normal and Below Normal Mental Component Summary Z-Score
PCS, Physical Component Summary; MCS, Mental Component Summary; SD, standard deviation; AIS, American Spinal Injury Association Impairment Scale; NLI, Neurological level of injury; ISS, Injury Severity Score; CCI, Charlson Comorbidity Index.
Bold: Variables significant at level of 0.05.
Variable normalized using Log(ISS) transformation.
Results of the logistic regression analysis for achieving normal PCS are shown in Table 3. When adjusting for covariables (AIS grade, NLI, ISS, and CCI), older patients remained significantly more likely to achieve a normal PCS score (odds ratio [OR] 1.03; 95% confidence interval [CI] 1.01–1.06; p = 0.002). Patients with AIS grade D were also more likely to achieve a normal PCS than those with AIS grades A, B, and C. Hosmer-Lemeshow (p = 0.13) and AUC (0.77) showed good fit.
Factors Associated with Achieving Normal Physical Component Summary Z-Score after Traumatic Spinal Cord Injury
OR, odds ratio; CI, confidence interval; AIS, American Spinal Injury Association Impairment Scale; NLI, Neurological level of injury; ISS, Injury Severity Score; CCI, Charlson Comorbidity Index.
Bold: Variables significant at level of 0.05.
Variable normalized using Log(ISS) transformation.
Hosmer-Lemeshow goodness-of-fit test, p = 0.13; area under the receiver operating characteristic curve = 0.77.
We did not conduct logistic regression analysis for achieving normal MCS, because bivariate analysis showed no statistically significantly association between age or any covariates and Z-MCS.
Discussion
To our knowledge, this study is the first to measure the likelihood of achieving normal QoL after TSCI by interpreting QoL with respect to age- and sex-matched normative data. Our results show that 39.3% and 80.4% of patients display normal PCS and MCS, respectively, suggesting that a significant proportion of individuals will achieve normal QoL within a year after TSCI. Interestingly, the overall proportion of patients with normal mental QoL (MCS) in our cohort was more than double that of patients with normal physical QoL (PCS), suggesting that TSCI predominantly affects long-term physical QoL. Accordingly, many authors observed similar inconsistencies between physical disability and mental QoL and hypothesized that patients redefine their expectations and purpose to better reflect their new reality, thus reducing its negative impact on long-term mental well-being. 2,8,10,25
In contrast to previous studies, we conclude that older age is associated with improved physical QoL after adjusting for AIS grade, NLI .and trauma severity. 6 –9 Our results show that older patients are more likely to achieve physical QoL comparable to their able-bodied peers of the same age group and sex (normal Z-PCS). This result reflects that in the general Canadian population, younger individuals have higher mean PCS scores (53.0 ± 7.2 for 25–34 year-olds) than older individuals (42.0 ± 10.3 for ≥75 year-olds). 15 In other words, older patients with TSCI have a higher likelihood for achieving normal Z-PCS, because older Canadians typically have reduced physical QoL when compared with younger individuals.
In addition, patients with TSCI generally receive extensive rehabilitation and can benefit from technical aids, which may contribute to further reduce the difference in physical QoL with able-bodied peers. On the other hand, younger healthy Canadians have a much higher physical QoL than older healthy Canadians, which makes it more difficult for younger patients to achieve a normal physical QoL after TSCI. On the flipside, this also means that on an unstandardized QoL basis, it is expected that younger TSCI patients have a higher raw (unstandardized) PCS score than older TSCI patients, mirroring the tendency in the general population.
Consequently, given that all previous QoL studies have measured outcomes that were not standardized with age- and sex-matched normative data, it is no surprise that they have found QoL to decrease with age. 6 –9 A similar tendency may be found in other outcomes after TSCI. For instance, the general consensus in the TSCI literature is that older patients experience worse functional outcomes than younger patients after injury. 26 –28 This is also to be expected, because all the above studies assessed functional status without previous standardization. If we were to standardize functional outcomes using age-matched normative scores as we have done for QoL in this present study, however, we argue that older patients could show similar or better functional outcomes than younger patients.
Similar to previous studies, AIS grade was negatively associated with achieving normal PCS in patients with TSCI. 2,8,21,25,29 The AIS D patients were three to five times more likely to achieve normal physical QoL than AIS grades A, B, or C patients, with only slight differences in the likelihood of achieving normal physical QoL between A, B, and C TSCI. Surprisingly, NLI was not associated with physical QoL. Overall, these results suggest that age and AIS grade (AIS A, B, and C vs. AIS D) are most important for patient counseling and decision making relative to long-term expectations in physical QoL.
Older age was not a determining factor in achieving normal mental QoL. This may partly be explained by the lower variability in normative MCS scores between younger and older Canadians (50.1 ± 9.6 for 25–34 year-olds vs. 54.5 ± 8.6 for ≥75 year-olds), which translates to smaller differences in the threshold to achieve normal MCS across different age groups. 15
Limitations
One important limitation of our study is the high number of patients (36.5%) lost at the one-year follow-up. This proportion, however, remains in line with previous studies conducted in the TSCI population. 22 –24 In addition, comparisons of baseline characteristics between patients present and absent at follow-up showed no significant differences other than in living arrangement. Thus, we believe that the external validity of this study is still sufficient to apply to a majority of patients with TSCI.
Another potential limitation is that the use of two standard deviations to define a range of normal PCS and MCS for each age group may seem skewed. First, we must note that this quantifiable and objective definition of normalcy was proposed on the basis that the commonly accepted statistical definition of a “reference range” in the medical literature typically encompasses 95% in a sample or population. 30 –33 For instance, in clinical biochemistry, the reference range of laboratory values is determined based on the middle 95% (or two standard deviations) of the measurements from a sufficiently large sample of healthy individuals. 32 In the case of SF-36 composite scores, table 8.3 of the SF-36 User Manual shows that, among the overall US population (all ages combined), respectively 6.5% and 6% of individuals have a PCS and MCS score two standard deviations or more below the mean. 34 In other words, respectively, 93.5% and 94% of the overall population have PCS and MCS within the two standard deviation range, which is very close to the statistical definition of a reference range (95% of a population). Therefore, we argue that the cutoff point of two standard deviations is not as skewed as it may seem and that it can be applied for PCS and MCS scores across all age groups to establish age-specific normal QoL.
Conclusions
We have interpreted QoL in patients with TSCI by comparing patient QoL with normative QoL data for healthy individuals of the same age and sex, which better reflects the likelihood for individuals to achieve normal QoL after TSCI. A significant proportion of individuals with TSCI can achieve normal QoL, particularly for the mental component. When compared with younger individuals, older individuals are more likely to achieve normal PCS and present a similar likelihood for achieving normal MCS.
Footnotes
Acknowledgment
We wish to thank Jean Begin for assistance with statistical analyses. Parts of our abstract was submitted as a digital poster at the 2021 International Spinal Cord Society scientific meeting.
Authors' Contributions
Victor Lim, Andréane Richard-Denis, Antoine Dionne. and Jean-Marc Mac-Thiong conceived and planned the study. All authors contributed to data collection, and Victor Lim and Antoine Dionne performed statistical analyses. All authors participated in the interpretation of the results. Victor Lim took the lead in drafting the manuscript, and all authors provided critical feedback and helped shape the research, analyses, and interpretation.
Funding Information
This study was supported by grants from the Fonds de recherche du Québec – Santé, the Medtronic research chair in spinal trauma at Université de Montréal and the Praxis Spinal Cord Institute.
Author Disclosure Statement
Mr. Dionne has received a scholarship from the Medtronic research chair in spinal trauma at Université de Montréal and a scholarship from the Craig H. Neilsen Foundation. Dr. Richard-Denis has received a scholarship and research grants from the Fonds de recherche du Québec – Santé, an investigator-initiated research grant from Medline Industries, and a research grant from Praxis Spinal Cord Institute. Dr. Mac-Thiong reports grants from Fonds de recherche du Québec - Santé, grants from Medtronic research chair in spinal trauma at Université de Montréal, grants from Praxis Spinal Cord Institute, during the conduct of the study; other from Spinologics Inc., grants from Medtronic, grants from DePuy-Synthes, grants from Canadian Institutes of Health Research, grants from Craig H Neilsen Foundation, grants from New Frontiers in Research Fund, grants from Fonds de recherche du Québec, grants from Canada Foundation for Innovation, grants from U.S. Department of Defense, grants from Medline Industries, grants from Vertex Pharmaceutical, grants from Abbvie, grants from Asahi Kasei Pharma, outside the submitted work.