Introduction

Idiopathic scoliosis is a three-dimensional torsional deformity of the spine affecting otherwise healthy individuals and often detected during rapid periods of growth [1]. Over the past few decades, a paradigm shift has occurred in understanding health and disability with increased emphasis on health-related quality of life (HRQoL) measures and understanding disability with regard to functioning after health changes. The impact of scoliosis on quality of life is therefore well documented and may affect psychological well-being through concerns for physical appearance, pain, lower self-esteem, uncertain prognosis, and concerns over peer interactions [2,3,4]. It has been argued that treatment modalities, such as bracing, may further affect HRQoL negatively and in a previous adolescent cohort study, individuals who were undergoing or previously received brace treatment had impaired HRQoL, as estimated with the disease-specific Scoliosis Research Society-22r (SRS-22r) and generic EuroQol 5-Dimensions index (EQ-5D), compared to untreated individuals [5]. The evidence however is conflicting, and not all studies have supported the theory that bracing may be detrimental to HRQoL scores in adolescents undergoing brace treatment [6, 7].

Avoiding long-term compromise to HRQoL is an indication for treatment and therefore of relevance in future research. Surgically treated individuals have been shown to have improved HRQoL, pain, satisfaction, and self-image scores, compared to brace treated individuals [8,9,10]. This is, however, with questionable clinical significance [11] and long-term follow-up has illustrated lower scores in function and self-image domains within the SRS-22r compared to brace treated or untreated individuals [12].

The aim of the current study was to further delineate health-related quality of life in adolescents treated or observed for idiopathic scoliosis, as well as making comparisons with a group of healthy controls.

Methods

Individuals with idiopathic scoliosis who were formerly or currently treated or observed at one of five Swedish Orthopaedic Departments (Karolinska University Hospital, Sahlgrenska University Hospital, Skåne University Hospital, Linköping University Hospital and Sundsvall and Härnösand County Hospital) were invited to participate in this study from September 2010 until October 2018. Individuals were included if they had a diagnosis of Idiopathic Scoliosis on standing radiographs with a curve in the frontal plane of at least 10 degrees, measured according to the methods described by Cobb [13], were between 10 and 18 years of age and were able to understand and read the Swedish language. Individuals were excluded if they had signs or symptoms associated with scoliosis of non-idiopathic origin. In total, 307 individuals with Idiopathic Scoliosis with a mean age (standard deviation) of 15.5 (2.1) years met the inclusion criteria were included in this ongoing study which had the primary purpose to study genetics of idiopathic scoliosis [14, 15]. Figure 1 shows the inclusion process. Of the individuals with Idiopathic Scoliosis, 111 individuals were untreated, 64 individuals had ongoing brace treatment, 47 individuals had previously been braced and 85 individuals were surgically treated. The untreated patient group comprised of individuals who had not reached the Cobb threshold of 25° indicated for brace treatment, individuals who had reached skeletal maturity, those who had declined brace treatment, as well as individuals who were being planned for surgery. The ongoing brace group had been undergoing treatment for a mean time of 1.3 ± 1.6 years and the previously braced group had, on average, ceased wearing the brace 1.8 ± 1.6 years previously. Bracing was recommended to adolescents with curves ranging from 25 to 40 degrees, where further skeletal growth was expected, as suggested by the Scoliosis Research Society [16]. A full-time rigid thoracosacral (TLSO) orthosis was utilised by 38 individuals in the ongoing brace group, and 41 individuals in the previously braced group. Hyperextension night-time brace was utilised by 26 individuals in the ongoing brace group (41.3%) and 6 individuals in the previously braced group (12.8%). Individuals wearing a full-time brace were encouraged to wear the brace for 20 h a day and for those with a night-time brace 8–10 h. Surgical treatment was recommended in larger curves (usually 45° or above), if progression was expected. Time after surgery for this group of 85 individuals was a mean of 1.7 ± 1.7 years. Controls without scoliosis were selected from a previous study investigating normative data for SRS-22r and EQ-5D as described earlier [17]. In total, 80 individuals within the same age category as the adolescents with scoliosis at a mean age of 14.0 (2.2) were included in this study.

Fig. 1
figure 1

Flowchart of the individuals with idiopathic scoliosis in the study

Full size image

Health-related quality of life questionnaires

All participants answered the SRS-22r and EQ-5D questionnaires. The SRS-22r consists of 22 questions distributed over five domains: function, pain, self-image, mental health and satisfaction, and has been translated into Swedish [18]. The domain scores range from 1 (worst) to 5 (best). An index for each domain has been derived as well as a total index for all the domains. The satisfaction question cannot be answered by those not being treated for scoliosis or controls and for that reason, the subscore has been used, which is a total score average of all valued SRS-22r items excluding the satisfaction domain.

EQ-5D 3-level version is one of the most widely used generic instruments for measuring HRQoL [19]. Individuals assess their health in five dimensions including mobility, self-care, usual activities, pain/discomfort and anxiety/depression. For each dimension 1 of 3 options is possible to choose and reflects the responders ability to manage each dimension [20]. An index will be obtained using the Swedish tariff where the index values are between 0.34 (worst) and 0.97 (best) [19]. The EQ-5D has been validated for its use in the AIS population [21] and shown to have moderate to good correlation properties with SRS-22r [17].

Exercise participation

Exercise participation was measured subjectively by answering yes or no to the question; “Do you participate in any exercise or competitive sports?” The duration, intensity or exercise type was not taken into consideration for the purpose of this study.

Radiography

Curve size was assessed according to the Cobb measurement method [13] and was obtained from the radiograph available from the most recent clinical follow-up appointment previous to the date of questionnaire completion. Cobb angles mean (min–max) were for the subgroups as follows: untreated 30 (11–78) degrees, ongoing brace group 34 (21–57) degrees, previously braced group 35 (25–50) degrees, and for the surgically treated group 24 (5–44) degrees.

Statistical analysis

Data are presented as the mean, standard deviation and range or number and percentage. The Chi2 test was used for categorical data. Mann–Whitney U test and analysis of covariance (ANCOVA) were used for continuous data, with adjustments for age, sex, and exercise participation. Descriptive and demographic statistics between the four different scoliosis treatment groups were completed with analysis of variance (ANOVA) with post hoc analysis using Tamhanes test for BMI and curve size as equal variances could not be assumed according to Levenes test. Statistical analysis between these four groups was completed with ANCOVA with adjustments for age, sex, and body mass index (BMI).

For comparisons based on curve severity, individuals with ongoing or previous brace treatment as well as untreated individuals were stratified according to Cobb angles (≤ 30° or > 30°) as well as larger angles (< 45° or ≥ 45°). Surgically treated individuals were stratified according to lowest fusion level (L2 as most caudal vertebra or below). Statistical analyses were performed with ANOVA. All analyses were performed using the Statistical Package for the Social Sciences (SPSS) statistical software for Windows (SPSS V26, IBM Corporation, New York, NY, USA).

Results

Health-related quality of life in adolescents with and without idiopathic scoliosis

Descriptive statistics for the entire scoliosis group compared to healthy controls are shown in Table 1. Group comparisons showed that the idiopathic scoliosis individuals were significantly older, contained a larger proportion of females and were less likely to be currently participating in exercise or sports compared to the healthy control group (p < 0.001, p = 0.023 and p = 0.014, respectively).

Table 1 Descriptive statistics for the scoliosis and healthy control groups
Full size table

The scoliosis group had significantly lower scores compared with the individuals without scoliosis in the SRS-22r subscore (p < 0.001) as well as in the separate domains of function (p = 0.009), pain (p < 0.001) and self-image (p < 0.001), Table 2. The EQ-5D index showed similar findings with the individuals with Idiopathic Scoliosis scoring significantly lower (p = 0.032) compared to the healthy controls.

Table 2 HRQoL comparisons between scoliosis and healthy controls
Full size table

Comparisons between the different scoliosis treatment groups

Descriptive data are presented in Table 3. Significant differences were detected between the groups with regard to proportion of females, age, BMI and curve size (p = 0.005, p < 0.001, p = 0.012 and p < 0.001, respectively).

Table 3 Descriptive statistics for scoliosis treatment groups
Full size table

Comparisons between the four treatment scoliosis groups are presented in Table 4. There was no significant difference between the groups in SRS-22r subscore (p = 0.67). Post hoc analysis of the SRS-22r function domain showed significantly lower scores in the surgically treated group (4.40 ± 0.63) compared to the untreated (4.65 ± 0.40, p = 0.002) and ongoing brace groups (4.68 ± 0.39, p = 0.002). There were no differences in function scores between the remainder of the groups. In the SRS-22r pain domain, the surgically treated group had a significantly lower score (3.95 ± 0.90) compared to the ongoing brace group (4.46 ± 0.48, p = 0.003), and the untreated group also had a significantly lower score (4.17 ± 0.79) compared to the ongoing brace group (p = 0.03). There were no statistical differences between the groups in the domains of self-image (p = 0.35) and mental health (p = 0.51). There were no significant differences in EQ-5D index scores between the groups (p = 0.51).

Table 4 Overall comparisons of treatment groups
Full size table

When comparing the three scoliosis groups who received treatment, significantly higher treatment satisfaction scores were evident in the surgically treated group (4.03 ± 0.92) compared to both ongoing brace (3.73 ± 0.90) and previously braced (3.44 ± 0.81) groups (p = 0.001). No difference between these three groups was seen with regard to the total SRS-22r score with satisfaction taken into consideration.

Curve size characteristics

When stratifying the non-surgically treated scoliosis individuals by curve size (≤ 30° or > 30°), the group with larger curves (n = 112) had a lower SRS-22r subscore (4.08 ± 0.53) compared to those with smaller curves (n = 110, 4.31 ± 0.48, p = 0.001), which was mainly driven by differences in the SRS-22r domains function, pain, and self-image (p = 0.001, p = 0.006 and p < 0.001, respectively). The mean scores for the SRS-22r domains function, pain and self-image were, for the group with larger curves: 4.55 ± 0.43, 4.16 ± 0.78 and 3.66 ± 0.83, respectively, and for the group with smaller curves: 4.71 ± 0.39, 4.38 ± 0.63 and 4.10 ± 0.71, respectively. These results are illustrated in Fig. 2. No significant differences in EQ-5D index score were detected between the groups (p = 0.245).

Fig. 2
figure 2

SRS-22r: comparisons based on curve severity. Mean values are plotted

Full size image

When stratifying the non-surgically treated scoliosis individuals by curve size (< 45º or ≥ 45°), those with larger curves (n = 30) had a lower SRS-22r subscore (3.91 ± 0.58) compared to those with smaller curves (n = 192, 4.24 ± 0.50, p = 0.002), as well as a lower EQ-5D subscore compared to those with larger curves (0.89 ± 0.08 and 0.93 ± 0.06, respectively, p = 0.006).

Surgical characteristics

There were no differences in EQ-5D score, SRS-22r subscore or SRS-22r subdomains when stratifying the surgical group by lowest fusion level (all p ≥ 0.28).

Discussion

This study found an overall reduced health-related quality of life in Swedish adolescents with idiopathic scoliosis compared with healthy controls, as estimated with the disease-specific SRS-22r and the generic EQ-5D. The difference in the SRS-22r was mainly driven by the separate domains of function, pain and self-image, the mental health domain did not differ between the groups.

In a previous study, Mariconda et al. 2016 detected reduced HRQoL pre-operatively in a cohort of adolescents with idiopathic scoliosis (n = 87) compared to age- and sex-matched healthy controls in physical domains of the SF-36 questionnaire as well as reduced SRS-23 scores within the self-image and pain domain [3], which is similar to the results in the current study.

When comparing between the scoliosis subgroups, no differences were detected with regard to HRQoL as demonstrated by EQ-5D or the SRS-22r subscore. In the separate SRS-22r domains, the surgically treated group had lower function and pain scores compared to the ongoing brace group as well as lower function scores compared to the untreated group, which is similar to that of previous studies [22, 23]. In contrast however, these studies also detected superior self-image scores in surgically treated individuals compared to conservatively treated individuals. In the study by Bunge et al. better self-image scores were only detected in surgically treated individuals without prior brace treatment, which may to some extent explain the disparity as the surgical group in the current study also includes 24 individuals (28%) previously treated with a brace.

In a large cohort (n = 652) of adolescents with scoliosis, Cheung et al. detected lower HRQoL scores in adolescents with ongoing and previous brace treatment compared to observed scoliosis adolescents [5]. HRQoL in this study was also estimated using SRS-22 and EQ-5D scores, but the results differed somewhat to the current study. Possible explanations for differences may be due to brace characteristics, time after brace cessation, differences in compliance or cultural differences.

The lower function and pain scores in the surgical group in this study may arguably be due to short-term follow-up (mean 1.7 years) following surgery. Diarbakerli et al. have, in a large cohort study, studied the effect of different scoliosis treatments during adolescence on HRQoL in adulthood [12]. In this study, surgically treated individuals had lower function scores, as well as lower SRS-22r subscore and self-image, compared to their conservatively treated peers. This would suggest that reduced function in surgically treated individuals compared to conservatively treated individuals, is not only temporary, but something that continues to extend into adult life. Those with a more caudal fusion level in Diarbakerli’s study also scored lower on EQ-5D index, SRS-22r subscore as well as subdomains of pain and function [12]. This differs from the results of the current study where no differences were detected, and it may be hypothesised that fusion level affects surgically treated individuals more long term, rather than during adolescence, due to degenerative processes affecting the spine.

Our results show that, despite reduction in function and an increase in pain, the surgical group was more satisfied with their treatment compared to both brace groups. This is in keeping with findings from other studies [12, 22, 23]. Satisfaction with management did not differ between the two brace groups. Possible explanations for improved satisfaction scores in the surgical group may be due to the direct improvement in curve size on the frontal radiograph following surgery and improvements in cosmetic appearance. Similarly to our findings, a previous study detected reduced function (measured using the SRS-24) at a 5-year follow-up in surgically treated scoliosis individuals compared to healthy controls [19]. However, pain, self-image and HRQoL in the surgically treated scoliosis individuals were superior to untreated scoliosis individuals and similar to that of healthy controls [19].

In the current study, an inferior HRQoL was detected in the group with larger curves, illustrated by lower SRS-22r subscore. This is in keeping with the study by Berliner et al. where adolescents (n = 286) with curves smaller than 40 degrees had a higher SRS-22r total score as well as superior scores in the subdomains of pain and self-image [24]. Differences in HRQoL in this study were mainly driven by differences in the SRS-22r domains function, pain, and self-image which is in keeping with previous research [24,25,26]. A large difference was seen within the self-image domain, which is supported by other studies which have shown good correlations between self-image and clinical and radiographic deformity [25, 27]. The current study showed differences in SRS-22r subscore when scoliosis individuals were stratified by curve size at both 30 and 45 degrees. Differences within the EQ-5D were, however, only detected with stratification above and below 45 degrees, indicating that this outcome measure may be less sensitive in detecting HRQoL changes within this population group.

Statistical differences between the adolescents with idiopathic scoliosis and the healthy control group showed that scoliosis individuals had a reduced SRS-22r subscore, more pain and worse self-image scores on SRS-22r which exceeded minimal clinically important difference (MCID) values previously published [28]. No differences in the SRS-22r domains between the scoliosis treatment groups or the curve severity groups reached a level of clinical significance. Previously published MCID values are based on individual changes pre- and post-operatively and not necessarily applicable to comparing data on group level.

To the best of our knowledge, this study is the first of its kind to compare so many different scoliosis treatment groups, including a surgical group, as well as a group of healthy adolescents. The internal validity is however, arguably, somewhat compromised since brace type (full-time or night-time) differs within the ongoing and previously braced groups. The cross-sectional design of this study limits results to descriptive findings as opposed to conclusions with regard to causality. The results from this study are thought to have sound generalisability should therefore be taken into consideration by clinicians, especially in comparison to healthy controls, and as a platform for further research in this field.

Conclusions

In conclusion, adolescents with Idiopathic Scoliosis had a lower health-related quality of life compared to healthy controls as measured by both scoliosis specific and generic outcome measures. Adolescents with idiopathic scoliosis are mostly concerned with pain and self-image. Surgically treated individuals were to a larger extent satisfied with their management, despite having decreased HRQoL otherwise, compared with brace treated individuals. Finally, larger curves were shown to be associated with overall decreased HRQoL.