A recent Cochrane systematic review evaluated the efficacy of bracing in adolescent idiopathic scoliosis (AIS) [1]. The full text of 128 studies was reviewed, but only one randomized controlled trial [2] and one prospective cohort study [3] could be included. From the results of these studies it was concluded that the evidence in favour of using braces is very low quality, making generalization difficult [1]. The authors recommended that future research should focus on short- and long-term patient-centred outcome. Unfortunately, prospective studies that started > 20 years ago did not use a randomized design, which limits their ability to evaluate causal relationships between bracing, curve progression, and Health Related Quality of Life (HRQL). The most comprehensive studies on long-term outcome including HRQL measures have been published by Danielsson and Nachemson and present outcome 22 years after treatment [4–7]. In a prospective study we recently reported the results in 109 patients treated with Boston brace at an average of 19.2 years earlier [8]. We found that the major curve size was not different from the pre-brace curve. Eighty percent worked full time, 88% had delivered a baby, and the average HRQL-scores were comparable with the results reported by Danielsson and Nachemson and in the same range as in the normal population. Scoliosis Research Society - 22 (SRS - 22) scores were in the same range as in a study evaluating 109 patients 10-years after surgery using third-generation instrumentation [9]. In particular, the mean score for self-image was identical, while patient satisfaction was slightly worse.

Weigert et al. compared results in 49 patients with Boston brace only, 41 patients with surgery, and 33 patients treated with brace and surgery at minimum 2 years follow-up [10]. They reported that surgically treated patients had better scores for post-treatment self-image and satisfaction, while the brace only treated patients had a higher level of general activity. They also found that the double-treated group scored highest in most of the domains including satisfaction, and concluded that brace treatment did not have any long-lasting negative effect on HRQL, causing no more psychological harm to the teenager than providing benefit.

The aim of the present study was to evaluate progression of the scoliotic curve and to report HRQL, using both validated scoliosis specific and generic questionnaires, in patients with idiopathic scoliosis 16 years and more after Boston brace treatment. We particularly wanted to compare results in patients presenting at an early age (late-onset juvenile scoliosis, 7-9 years) with patients presenting after 10 years (AIS), and in patients with curve progression to ≥ 45° and < 45°, respectively, with the results in those who had surgery after bracing.

Long-term results in 109 of 138 patients with AIS who were treated with Boston brace at Sophies Minde Orthopaedic Hospital (Orthopaedic Department, Oslo University Hospital, Rikshospitalet) in Oslo, Norway from 1976-88 and who had their last earlier follow-up no longer than 2 years after brace weaning, have been published previously [8]. These patients were not included in the present study.

We carefully evaluated the medical journals of the remaining 480 patients who had been braced for scoliosis and prospectively followed during the same period, and identified 369 patients with idiopathic scoliosis. These patients were invited for long-term follow-up. We included both patients with late-onset juvenile and AIS. Late-onset juvenile idiopathic scoliosis is discovered at 7 to 10 years, while AIS is discovered at 10 years of age or older [11].

In general we followed the SRS committee criteria, which recommend bracing for curves that measure between 25° and 40° and surgery for curves that are greater than 45° during the growth phase and that the brace should be prescribed until growth of the spine has stopped http://www.srs.org/professionals/conditions_and_treatment/adolsescent_idiopathic_scoliosis/treatment.htm. Accordingly, the indication for bracing was a major scoliotic curve ≥ 20° with an observed progression > 5° after 4 months and Risser sign < 4. Prior to bracing standing radiographs were taken in the front and lateral projections. Patients had follow-up with clinical and radiological examination at 4 months intervals throughout the brace treatment period. Wearing of the brace was assessed by one orthopaedic surgeon (JEL) and reported as used as prescribed, irregular, or aborted. Patients were recommended to use the brace for 23 hours daily. Wearing of the brace < 20 hours daily was described as irregular.

Physiotherapy was not prescribed in addition to bracing, but in general patients were recommended to participate in physical activity at school and leisure time. They were advised not to carry heavy loads, and in case they had a long walk to school they were given two sets of schoolbooks.

Brace weaning was carried out either 2 years after menarche or at Risser sign 5, in some patients at Risser sign 4. After weaning all patients had follow-up examination at 6, 12, and 24 months.

A standardised form was used to obtain clinical and radiological data. Radiological measurements were performed by an orthopaedic surgeon (JEL) and controlled by an experienced radiologist (RG). Both used the Cobb method, digital measurements were used at long-term follow-up. The intra-observer error for the Cobb angle was about 3° in a recent study using manual and digital measurements, and < 5° in a previous study [12, 13]. In the present study the measurement error was within these limits as evaluated by the reproducibility of radiographic readings of repeated measurements of all radiographs from 10 patients at regular intervals. In patients with double thoraco-lumbar curves the largest curve prior to bracing was defined as the major curve.

Results at mean 24.7 years after treatment with Boston brace for AIS are in agreement with previous studies on bracing [8, 23]. Curve progression was < 6° in 69% of the patients and the average major curve size was almost identical to the pre-brace curve. The percentage that had later surgery was in the lower range compared with a recently published systematic review reporting that the percentage of brace treated patients with later surgery ranges from 1% to 43% [24]. HRQL was slightly worse in operated patients, but these patients had larger curves at weaning. About half of them were operated with Harrington instrumentation, but any inference about the results of current methods of surgery for idiopathic scoliosis cannot be made from the results of the present study. Our results are slightly different from Weigert et al. [10], who reported better post-treatment self-image and satisfaction in surgical patients, but results are not directly comparable because most patients had Cotrel-Dubousset instrumentation and follow-up was much shorter. However, previous studies have reported that HRQL at long-term follow-up after Harrington instrumentation for AIS is comparable with the normal population [25, 26].

Except for self-image and the pain domain on EuroQol the reported HRQL was not related to curve size at follow-up. Some of the patients with large curves at follow-up had refused surgery, but most braced only patients had a curve size < 45° at maturity with an average of 37.3°.

Results suggest that average curve progression and HRQL are satisfactory at long-term, although a higher percentage of patients report moderate and severe pain and discomfort and difficulties with usual activities as compared with 40-49 year old women in the United Kingdom national questionnaire survey using the EQ-5D questionnaire [27]. They reported that the population percentage reporting any problem in this age and gender category was 11% for mobility, 4% for self-care, 12% for usual activity, 27.5% for pain and discomfort, and 21% for anxiety and depression. Figure 4 shows that from 60 to 90% of the patients in the present study reported to have either moderate or severe pain. Although the average SRS-22 pain score was low and that the three response alternatives for EQ pain and discomfort do not allow for reporting some or occasional pain, the present study indicates that the majority of patients with idiopathic scoliosis are not pain free at long-term. Despite some pain, most patients have little functional limitations and are working.

The design of the present study does not allow for causal inference. With the lack of control group and a randomized design we cannot exclude that curve development and HRQL would have been in the same range without bracing. The results of one prospective study that included a control group and one randomised trial that compared soft and rigid braces suggest an advantage of bracing on curve progression [3, 28]. Nevertheless, based on these studies, a recent Cohrane systematic review concluded that there is very low evidence in favour of using braces [1]. Results from ongoing randomized controlled trials are expected to improve the knowledge in this field [29, 30]. Because both bracing and surgery is demanding for the adolescent girl, the long-term prognosis in untreated patients must be considered including curve progression, HRQL, and costs. Recently, Danielsson et al. compared long-term outcome after bracing and observation only [31]. They found no difference in curve progression and long-term HRQL, but the cohorts were different at baseline regarding the number of patients who had an observed curve progression of > 5°. According to their conclusion many patients with idiopathic scoliosis are over-treated, and if bracing is limited to patients with documented curve progression the number who are possibly over-treated will be reduced.

Our results are in agreement with Aulisa et al. reporting that conservative treatment does not severely impact on HRQL in AIS patients [32]. However, results are not directly comparable because they included much younger patients (mean age 15.4 years at follow-up) and applied different braces. Thus, they evaluated the effects of brace treatment in young adolescents while the patients in the present study were middle aged. We cannot exclude that bracing had negative impact on some patients, and the scores for satisfaction with treatment suggest that bracing was demanding, but we did not find a negative impact on mental health at long-term. In contrast with the results in the current study, Aulisa et al. reported better results in boys.

The percentage reporting pain during pregnancy is comparable to women without AIS and in agreement with the results of a large previously published case-control study [5]. Questions about pregnancy and delivery are often raised in AIS. In agreement with previous studies our results indicate that patients can be reassured that scoliosis does not affect pregnancy or delivery.

A meta-analyses concluded that bracing is effective, but slightly less effective in patients with juvenile onset [33]. The present study did not include patients with early-juvenile onset, but there was no difference in curve development during bracing or at long-term or in HRQL at long-term between patients with late-juvenile and adolescent onset. Thus, similar results should be expected in these groups.

We conclude that long-term results were satisfactory in most of 272 patients with late-onset juvenile and AIS treated with Boston brace in average 24.7 years earlier. Nine percent had surgery and 13% had curve progression to ≥ 45° at follow-up. HRQL was slightly decreased in these patients, SRS-22 scores for self-image was significantly lower in both groups. Self-report indicates that future patients can be reassured that scoliosis does not affect pregnancy and delivery, and that most patients are expected to work and have HRQL in the normal range at long-term.

The study was supported by grants from the Sophies Minde AS Foundation.