Adolescent Idiopathic Scoliosis and Mental Health Disorders: A Narrative Review of the Literature

Adolescent idiopathic scoliosis (AIS) is the most common type of scoliosis. The condition begins in puberty, affects 1–4% of adolescents, and disproportionately affects young women. Our aim was to comprehensively examine the association between AIS and risk for depression, anxiety, eating disorders, psychotic disorders, and personality dysfunctional mechanisms. Methods: Literature review of related articles published in PubMed, Google Scholar, and Scopus up to 15 July 2021. Results: A total of 30 studies were deemed eligible, examining the effects of AIS upon mental health, and using appropriate psychometric inventories. Studies highlighted the association of brace treatment with elevated anxiety. In addition, mental health conditions and traits (e.g., anxiety and depressive symptoms, neuroticism) were detected more frequently amongst AIS patients compared to healthy controls. Conclusions: AIS represents a risk factor for mental health disorders. More longitudinal studies, utilizing accurate psychometric instruments, are warranted, to reveal the current impact of AIS on the mental health of youngsters, along with the tailoring of well-targeted interventions to reduce the burden of mental health issues in adulthood.


Introduction
Scoliosis is defined as a three-dimensional (3D) structural deformity of the spine and for practical reasons it is widely diagnosed radiographically, on the basis of the measurement of the Cobb angle. A Cobb angle is used as a measurement of the degree of spinal curvature; thus, scoliosis is present in case of at least 10 degrees of deviation [1]. In most cases, the etiology of scoliosis is multifactorial, and since 1922 these patients have received the diagnosis of idiopathic scoliosis [2]. Idiopathic scoliosis (IS) is classified into three categories, depending on the peak period of onset: the infantile, which occurs under the age of three; the juvenile, from five to eight years old; and the adolescent, which affects teenagers from 10 to 19 years old [3]. The latter accounts for 80% of idiopathic scoliosis cases, and a meta-analysis on its prevalence has shown a range from 0.47 to 5.2%, after school-screening [4].
Adolescent idiopathic scoliosis (AIS) involves high possibilities of progression, primarily amongst girls during the growth spurt at puberty [5]. From mild to moderate A literature search was performed on 30 May 2021 in three different databases (PubMed, Scopus, and Google Scholar). For the search, various key-words were used: idiopathic scoliosis, teenagers, youngster, adolescence, young adults, mental disorders, psychopathology, eating disorders, anorexia, bulimia, mood disorders, depression, emotional reactions, cyclothymic disorder, premenstrual dysphoric disorder, anxiety, and stress. Those terms were used in pairs, using "AND" or "OR", in order to receive specified results.

Inclusion Criteria
This review aims to investigate the association between AIS and mental health disorders, or indices of psychopathology, in adolescence or in adulthood. For this reason, eligible articles had to meet the following criteria:

1.
Studies had to report on adolescents or young adults with a diagnosis of AIS in their puberty, in terms of retrospective studies.

2.
Studies could be purely based on AIS, present results on a subgroup analyses on AIS, or encompass AIS cases as the majority of the study sample. 3.
Studies had to provide data about the correlation between AIS and mental health disorders, as defined by appropriate psychometric instruments. Any strategy to diagnose AIS was deemed eligible. 4. Prospective cohort/cross-sectional/case-control/retrospective/large-data based epidemiological studies/reliability and validity analyses of questionnaires were included. 5.
The article was written in English language. 6.
There was no restriction in publication year.

Exclusion Criteria
Articles meeting the following criteria were excluded from the review: 1.
Articles evaluating different aspects of patients' well-being and psychological function utilizing generic inventories.
Literature reviews; however, these were screened for relevant resources not found via search terms.

Data Abstraction
Numerous studies were retrieved from the databases. After the first check, a great number was excluded by title or abstract. The studies that were considered eligible were evaluated in full-text and separately tabulated on the basis of the mental health condition they assessed. From each study, analytic characteristics were extracted concerning the study design, the sample size, the percentage of males, study population, assessment of scoliosis, evaluation of mental health disorder, treatment prescribed, severity of the main curve, and the main findings.

Study Characteristics
A total of 30 studies, from 1992 to 2021, were included in the present review . Most of those included were cross-sectional, (15 of 30), including eight from Poland, two from China, two from United Kingdom, two Italian, and one from Spain. The others were mainly prospective cohort studies, including seven ranging from Poland, USA, and Taiwan to UK and Sweden. Among the remaining ones, there were four case-controlled studies, one registered audit from the UK, an Iranian validation study, and one clinical trial from Spain. The percentages of male patients were significantly lower (range: 0-41%) and only one study was exclusively based on male subjects with scoliosis [50]. Most cases of idiopathic scoliosis detected were adolescent type, with an overall moderate severity (20-40 • ) treated by brace. With regards to mental health disorders, nine studies examined stress/anxiety, three studied depressions, eight studied jointly the aforementioned conditions, four examined eating disorders, five studied personality pathology, and two studied schizophrenia. The study of Oh et al. [43] was added in both the latter categories. Demographic characteristics are represented in Table 1.
The tables present the studies on the association between AIS and stress/anxiety, depression, joint examination of depression and anxiety, eating disorders, personality disorders and traits, and psychotic disorders (Tables 1 and 2).   Brace increases stress moderately in both groups. The two groups have significant statistical differences, solely in relation to stress levels, due to body deformation (p = 0.004), where the group treated surgically reported higher stress levels. Age at initiation of treatment increases the stress levels of patients treated with a brace (p = 0.029). Significant statistical correlation (p = 0.008) pre-operatively between the degree of translation and stress levels due to body deformation in patients treated conservatively.  The braced patients revealed more stress when investigated for their braces than for their deformity (median = 18) comparing to BSSQ Brace (median = 9). The BSSQ deformity revealed a median of 17 points in patients managed with exercises, 18 in patients managed with a brace, and 12 in patients before spinal surgery correlation, between the total score of BSSQ deformity and the following parameters: Cobb angle (r = −0.34, p < 0.05).  More scoliotic than nonscoliotic participants showing depressive symptoms (45% vs. 33% of subjects, respectively). The differences between the subjects with mild and moderate deformities were also significant (p < 0.05), with a greater tendency for depressive symptoms in subjects with milder deformities, more women than men exceeded 10 point threshold, which means severe depression.  The cohort were experiencing an elevated level of anxiety but their mood was normal. Mean anxiety score was 7.4 (SD 3.84); 14 (60.9%) fell below the cut-off and 9 (39.1%) were above. The higher preoperative scores showed significant p-values for anxiety/depression (p = 0.05) and internalizing (p = 0.05) on the YSR scale The level of pain correlated significantly with preoperative anger (p = 0.02), social problems (p = 0.01), and attention problems (p = 0.05). AIS patients were significantly lighter (p < 0.001) and had significantly lower BMI scores (p < 0.001); 25% of the series had BMI scores which were within the range considered to be an anorexic predictor in the multivariate regression model.  In the neurosis set, all the scales of anxiety, depression, and somatization were significantly increased in the AIS group (p = 0.010, 0.003 and 0.002, respectively). In the personality disorder scales showed a significantly increase in the AIS group (p = 0.001).

Results of Individual Studies
The majority of the studies found during the literature research marked the association between AIS and increased levels of stress/anxiety. Nine studies from 2001 to 2018, indicated an overall moderate to low stress amongst a large part of the adolescent population treated by brace, surgery, or physiotherapy (Tables 1 and 2). On the one hand, Misterska [25,26], as well as Motlagh [27], Kotwicki [28], and Leszczewska et al. [29], using the same inventory of BSBQ, concluded that the brace increased the level of stress in comparison to the stress induced by the deformity alone. Similarly, Glowacki et al. [22], evaluated the anxiety of braced adolescents by STAI-C inventory, showing a constant brace-related anxiety during a 12-month observation period. Two additional parameters that Mistreska et al. pointed out pertained to the statistically significant correlation between apical translation and stress level, as well as the age of initiation of the conservative treatment [24]. However, from the studies jointly examining depression and anxiety (Tables 1 and 2), Sanders et al. found that 32% of patients with AIS exhibited significant psychological distress, regardless of the type of treatment. In this study, anxiety represented the most common concern for teenagers, with a similar occurrence as in pediatric cancer and heart transplant patients [39]. Supportive evidence was provided by a recent large-database epidemiologic study from 2012 to 2016, showing that among the overall prevalence of 7% of mental health disorders in AIS patients, the highest percentages were attributed to anxiety [41].
On the other hand, adolescents with severe scoliosis requiring surgery, experienced a significant increase in anxiety levels postoperatively, which was explained by the pain intensity [21]. Rullander et al. marked a similar correlation between the level of pain and anxiety (p = 0.03) in Trauma Symptom Checklist for Children, reporting that the studied adolescents had higher stress symptoms before, than after, surgery [40]. The higher symptoms of anxiety in AIS surgery patients compared to healthy controls was also demonstrated by the study of Duramaz et al.; although, indicating a notable decrease in such symptoms after the spinal correction [35]. The stressful experience of scoliosis surgery was also indicated by the scores of HADS anxiety inventory in another study, concluding that operative treatment resulted in elevated levels of anxiety [38]. From a similar biopsychosocial perspective, Wong et al. demonstrated a remarkable prevalence of back pain amongst AIS patients, regardless of a surgical treatment. The high prevalence of chronic back pain increased the odds for anxiety, depression, insomnia, and daytime sleepiness [36]. The morbidities for anxiety and depression were also investigated by Wang et al., in patients with AIS and their parents, showing that only 7.3% had scored within the cutoff point of 10 that denoted probable major depressive disorder and 3.2% for severe anxiety. However, this cross-sectional study showed that the morbidities of parental depression and anxiety were 14.1%, significantly higher than those in the control group, proposing a causal relationship between parents' mental disorders and adolescents' general distress [34]. Table 2 summarizes the characteristics of the studies focusing on the association between AIS and depression. The depressive symptoms, when evaluated by Lin et al., showed an overall medium extent, which was significantly higher amongst female patients. The greater presence of such symptoms in AIS than in juvenile IS patients reflected the vulnerability of adolescence to mental health issues. The parameters positively associated with the severity of depression were the duration of brace treatment and the severity of the main curvature, as was already reported above within the studies examining the anxiety levels in braced patients [31]. Indeed, depression was proved to be more prevalent amongst scoliotic patients during a 5-year follow up; specifically, Chang et al. found higher hazard ratios of depressive disorders in patients with scoliosis and mostly amongst young adults and the middle-age [32]. In parallel, the risk of depression was also demonstrated by Baird et al., who found that 18% of AIS patients had scores worthy of further assessment for a potential diagnosis of depression [37]. Finally, the only study in the current literature that evaluated depression in young women with AIS using the Beck Depression Inventory marked a greater tendency for depression in female patients with mild and moderate deformities [33].
Regarding the matter of eating disorders, variable results have been published (Table 2). Alborghetti et al., using the Structured Clinical Interview for DSM-1 Disorders (SCID-1), showed a significantly higher prevalence of eating disorders in the scoliosis group when compared to an Italian female population (9.2% for anorexia nervosa, 7.7% for bulimia nervosa, and 5.3% for EDs). Interestingly, it was also highlighted that only girls with AN had a significantly higher degree of deviation of the vertebral column than adolescents without AN, F (1, 76) = 3.87, p < 0.05, indicating an association between the severity of scoliosis and the presence of AN [42]. Similarly, Smith et al. demonstrated the relationship between a diagnosis of AIS and anorexia nervosa, indicating that more than a half of the studied patients had BMI scores within the range considered to be anorexic [43]. On the other hand, Zaina et al., using the Eat-26 questionnaire, showed a low prevalence of eating disorders in AIS female patients. The significantly lower BMI remained present, but it was considered typical of scoliosis patients [44]. From a common perspective, Smith et al. showed that, notwithstanding the typically lower body mass index (BMI) of the studied group, the prevalence of eating disorders remained very low and similar to the general population [45].
Regarding personality pathology, results from the included studies suggested a pattern of neurotic personality traits ( Table 2). These findings were demonstrated by the studies of Oh and Misterska et al. Both authors, using personality inventories, indicated higher scores in the neurotic scales amongst AIS patients compared to healthy controls [47,50]. Oh et al. [50], assessing males during the end of puberty obtained more accurate indices of personality disorders, using a conventional type of the Minnesota multiphasic personality inventory test, which is one of the most widely used psychological assessment tools [50]. Similar scores were demonstrated by Misterska et al., reporting higher scores in the neurotic scale for the group treated surgically, whereas the AIS group under brace treatment had higher scores in the manic scale [47]. Matsunaga, on the basis of a clinical trial, confirmed the trend towards neuroticism, investigating the impact of brace therapy in the personality functioning and emotional responses of the studied adolescents. Although these traits seemed to be modifiable, it is important to notice that just one month after the start of therapy, the MPI assessment showed that 25% of the studied population encountering neurotic characteristics, compared to 2% before bracing [48]. Additionally, Matsunaga [48], along with the studies of Leak [46] and D'Agata et al. [49], have jointly unveiled a passiveintroverted tendency among AIS patients, connected in all cases with brace-therapy and the challenging task of compliance.
Finally, two studies evaluated an association between psychosis and scoliosis ( Table 2). The study of Oh et al. is twice referred to, but alternatively interpreted, since MMPI is identified, as well as symptoms associated with psychosis. Significantly higher scores in the psychopathy set including schizophrenia were reported among the 213 examinees (p = 0.010) [50]. Likewise, an increased risk of schizophrenia in patients with idiopathic scoliosis was detected by Malmqvist et al., over a median follow-up time of 9.5 years after the AIS diagnosis. The odds ratio for schizophrenia was significantly higher in patients with IS (HR, 1.52; 95% CI, 1.03-2.23), suggesting a possible link between these two conditions [51].

Discussion
The present literature review is an attempt to summarize all the available studies designed to investigate the possible relationship between AIS and mental health disorders. The existing studies are mainly cross-sectional, (15 of 30) and based on questionnaire assessments, with small consecutive samples. In terms of stress/anxiety evaluation, BSSQ was the most widely used tool; only one study in the current bibliography used a GAD inventory [34], and two used the state-trait anxiety inventory [23,35]. The overall stress in AIS population was low to moderate; a large-database epidemiological design showed a positive association between AIS and mental health conditions, especially anxiety [41].
Accordingly, an assessment of youthful scoliotics with BASC-2 inventory indicated similar results of clinically significant distress, regardless of the type of treatment prescribed [39]. An additional finding related to surgically-treated patients that needs further investigation pertains to the positive correlation between the pains these patients experience and preoperative anger, social problems, and attention problems [40].
The current studies and reviews underscore the need to attend to the psychological assessment of AIS patients, and provide much supportive evidence of the stressful and demanding experience of AIS. Physical complaints, lower self-esteem, higher depression scores, feelings of shame and body inferiority, and suicidal thoughts have been detected and raise concern [52,53]. Accordingly, in our review the association between AIS and depression was highlighted [32,33].
Body dissatisfaction, which has been bibliographically referred to as a major psychological effect of AIS [54], is also connected to eating disorders [55]. In the present review, an Italian study demonstrated a higher prevalence of EDs in scoliotic women [42]. The EAT-26 [44] and eating disorder examination questionnaire (EDE-Q) [45] evaluation, however, provided different results, showing a lower incidence of EDs in female scoliosis patients than in the general population. In any case, the scoliosis group weighed less and had lower BMI scores, and the possible eating pathology amongst this population has to be further investigated on the basis of larger sample sizes and clinical interviews.
As for personality pathology, previous findings have outlined that the threat to body image can lead some female patients to develop the defense mechanism of denial of their condition [55], with feelings of shame and difficulties in social integration [56]. In parallel, brace-treatment and the need for compliance may evoke neurotic behaviors, due to the insecurity and fear for the treatment result, and reverse the patterns of relating and interacting between family members [57]. Finally, a possible association between idiopathic scoliosis and schizophrenia was supported by two studies [50,51]. The high psychometric properties of the Minnesota Multiphasic Personality Inventory (MMPI) for detecting psychopathic traits [50] and the large-database retrospective design have to be considered as strengths of these two designs. However, limitations exist concerning the study population, where only males from the military conscription examinations were included [50]. In any case, there is much space for further investigation to detect personality and psychotic disorders in cases of AIS; conducting large cohort studies enrolling women and men.
Future studies should be distinctly directed to the mental health of adolescents and adults with scoliosis, in terms of large-data based epidemiological studies and systematic reviews. In clinical research, adolescents should be addressed with specific screening tools along with psychosocial interviews, such as HEEADSSS, which allows the clinicians to comprehensively evaluate mental health in puberty and indicate risk and protective factors [58]. Randomized clinical trials should also be conducted, in order to evaluate the effectiveness of psychoeducational and psychotherapeutic interventions on adolescents with psychological distress, due to their deformity. Consequently, it turns out that psychological support should constitute an integral part of the overall therapeutic protocol of AIS, in order to address the growing need for psychosocial support of scoliotic patients and to reduce the burden of mental health issues in the future. This review has some limitations; studies on overall quality of life and psychological well-being were not included. Moreover, especially in case of depression and psychosis, we could have included studies with an admixture of AIS cases with other types of scoliosis. However, no such data were clearly defined and, due to the high prevalence of AIS, we could assume that the largest part of the studied population was diagnosed with AIS with a continuation in adulthood.

Conclusions
The findings of the present review suggest associations between AIS, depression, anxiety, and neuroticism. Future research should be promptly redirected to a screening perspective for mental health in AIS subjects, with the design of multidisciplinary interventions.