Association between domains of quality of life and patients with Klinefelter syndrome: a systematic review

Objective Klinefelter syndrome (KS) is the second-most prevalent chromosomal disorder in men, though late diagnosis is very common and 50–75% of men remain undiagnosed. Evidence suggests that men with KS have impaired quality of life (QoL) but research on how the diagnosis of KS is associated with different QoL domains and what factors influence patients’ QoL is limited. This study aimed to provide a systematic review of the published evidence on factors that influence QoL in men with KS. Design Systematic review and meta-analysis with narrative synthesis. Methods Medline, Cochrane, Embase, Psychinfo, CINAHL, BASE and relevant publication reference lists were searched in January 2021. Eligible studies included randomised control trials, cohort studies, cross-sectional studies and epidemiology studies on KS and its effect on QoL and all domains of World Health Organisation (WHO) Quality of Life 100 (WHOQOL-100). Clinical studies with no date restriction published in English were included. Results Thematic analysis was completed on 13 studies, with a meta-analysis of intelligence quotient completed on 7 studies. Twelve out of the 13 studies suggested that KS negatively affected the QoL outcomes and KS was associated with impairments in physical, psychological, level independence and social relationship domains of WHOQOL-100. Meta-analysis suggested that men with KS have significantly lower full-scale Intelligence Quotient vs controls (P < 0.00001). Conclusions This is the first evidence synthesis of QoL in men with KS. Current evidence suggests that combined physical and psychological impairments affect men with KS who also experience impairments in relationships and independence in society. Further research is needed to identify factors that influence the QoL in men with KS.

The extent of mosaicism in KS causes an array of cognitive, psychosocial, and physical symptoms which can affect men at varied degrees of severity. These include hypogonadism, gynecomastia, tall stature, small phallus, reduced level of intelligence, depression, autism traits, schizotypal traits and social anxiety which lead to impaired quality of life (QoL) (2,6,7,8,9). Milder phenotype and lack of distinct dysmorphic features present a real challenge for early diagnosis (3). Testosterone replacement therapy is recommended for patients with KS once serum gonadotrophins begin to rise in early puberty or when serum testosterone levels become hypogonadal (3,10,11).
Evidence suggests that patients with KS have more impaired QoL compared to healthy controls; however, research on how the diagnosis of KS affects a patient's QoL is limited (2,12,13).
There is limited knowledge of the various symptoms, outcomes and patient experiences, which may result in health and social inequalities for patients with KS. A greater understanding of the associations between KS and the domains of QoL can better support clinical decisionmaking and meet the condition-specific needs of patients with KS.

Objectives
The objective of this study was to conduct a systematic review with meta-analysis to provide new insights and further understanding of QoL in patients with KS and to answer the following research question: • What is the association between Klinefelter syndrome and the WHOQOL-100 domains/facets of QoL?

Underpinning framework WHOQOL-100
Due to the many factors influencing QoL in patients with KS, the World Health Organisation (WHO) Quality of Life 100 (WHOQOL-100) was adopted as the overarching framework to underpin this systematic review. The WHOQOL-100 is a validated psychometric scale which can be used to measure QoL as an overall construct and across its six QoL domains: 'overall QoL', 'physical health', 'psychological health', 'level of independence', 'social relations' and 'environment' (14). The subsections were developed by the WHO, by incorporating the important aspects of QoL defined by a range of patients and health professionals from various diseases, specialisms and cultural backgrounds. The WHOQOL-100 is a universal Patient-Reported Outcome Measure (PROM) that can measure individual QoL domains to provide evidence of unmet needs and impaired aspects of QoL.

Protocol and registration
The systematic review followed the Preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines for quantitative systematic reviews (15), and the study protocol was registered with PROSPERO (CRD42020173435). The Systematic Review Without Meta-analysis (SWiM) guidelines (16) were adopted for narrative synthesis which was guided by the overarching WHOQOL-100 and its six QoL domains (17). Metaanalysis was conducted where possible by grouping studies measuring overall QoL or outcome measures relating to the domains and facets of the WHOQOL-100 structure. Review manager 5.4 (18) was utilised for this analysis.

Eligibility criteria and participants
All empirical studies involving male children and/or adults diagnosed with KS and measuring a quantifiable factor of QoL that could be defined within the WHOQOL-100 were reviewed for eligibility. No publication date restrictions were imposed (see Supplementary material, see section on supplementary materials given at the end of this article).

Search
The search was completed on 21 January 2020 using the following databases: MEDLINE (1946

Study selection
The eligibility assessment was performed in a blind independent review by two authors (BM and SL); all disagreements were resolved by consensus and did not require a third reviewer. The blind review for abstracts and full text articles was done in Rayyan QCRI systematic review manager (https://rayyan.ai/) using pre-specified inclusion/exclusion criteria ( Table 1). The PRISMA flow diagram in Fig. 1 annotates the study selection process. A detailed inclusion/exclusion review of all full text articles is included in Supplementary material.

Data items
Information was extracted from each study on the following: (i) number of participants; (ii) study settings, study country and study design; (iii) outcome measures related to the WHOQOL-100 framework (domains/ facets) of QoL such as 'physical health', 'social relations', 'psychological', 'environment', 'level of independence' or 'religion/personal beliefs/spirituality'; (iv) comparison groups where possible ( Table 1).

Risk of bias of individual studies
The Joanna Briggs quality appraisal tool for cross-sectional studies/cohort studies (19) was used on all 13 studies included in the systematic review to ensure validity and to examine the reliability with a quantifiable score on each included study. Each question was dichotomised to either YES (1 point) or NO (0 points) producing a scale ranging from 0 (poor quality) to 8 or 10 or 11 (high quality) depending on the appraisal tool. Studies were given an appraisal score depending on how many categories of the appraisal they met: 'inclusion criteria', 'study settings and subjects', 'exposure', 'confounding factors', 'outcomes' and 'statistics' (see Supplementary material).

Narrative synthesis
The SWiM reporting items protocol (16) was adopted for the narrative synthesis. Using the WHOQOL-100 framework, subgroup analyses were conducted on each of the six subgroups of the WHOQOL-100 and reported in tables including study, effect size (Cohen's d) and main findings. A meta-analysis was not possible for all studies due to the large amount of differing and heterogeneous outcome measures and scoring systems for QoL.
Therefore, the percentage of significant findings and the strength of the effect sizes are considered within each QoL sub-section. Data were too heterogeneous for meta-analysis due to the vastly different outcome measures included; therefore, to aid comparability, Cohen's d was calculated by extracting the mean difference and s.d. from KS and control groups where reported. Accepted categories for Cohen's d effect sizes as small (0.2), medium (0.5) and large (0.8) were applied (20). Where studies did not report P values, these were calculated using Fisher's exact tests to show any significant differences (P < 0.05) between patients with KS and controls.

Study characteristics
A total of 665 records were identified from the initial search of which 13 studies, 12 cross-sectional and 1 cohort, met the inclusion criteria and 7 studies were suitable for additional meta-analysis (Fig. 1). The total number of participants across the 13 studies was 829; study sample sizes ranged from 14 to 219 participants. Studies had a mixture of patient-reported, parentreported, or physician-answered questionnaires. Table 1 presents a full summary of the characteristics extracted from each study.

Quality of studies
The quality of the included studies has an impact on the confidence of findings within the review. First, when assessed many studies did not discuss or include strategies to deal with confounders. Secondly, normative data and population averages were used for controls in three studies which lowers the comparative domain score for those studies (21,22,23), Nielsen and Pelsen (24) used hypogonadal males as controls which may also reduce confidence in this study as hypogonadal males QoL outcomes could be reduced due to the symptoms of hypogonadism. Eligible studies were assessed for methodological quality using the Joanna Briggs quality appraisal tools (Table 1).

Physical health
Three studies measured outcomes related to physical health against controls (

Level of independence
There were limited measures on the level of independence in patients with KS. Work capacity was measured by Nielsen and Pelsen, but no significant differences were found regarding skilled/unskilled labour and unemployment between patients with KS and healthy controls (24).

Psychological
A significant difference (P < 0.001) was identified between controls and patients with KS in each of these outcomes: autism spectrum quotient (AQ), gender identity/ dysphoria, neocriticism, extraversion, conscientiousness, attention switching, imagination, communication, global severity index (GSI), mini mental state examination (MMSE), positive symptom distress index (PSDI), social skills including social behaviour and negative assertion (6,30,32,34). A significant difference (P < 0.001) between patients with KS and reference population was also reported by Herlihy et al. regarding the psychological measures of body-self relations, self-esteem, sexual identity and psychological distress (21) ( Table 4). Fisher et al. and van Rijn et al. reported significantly greater prevalence of autism symptoms (P < 0.001) as measured by Autism spectrum quotient; both studies and a total of (n = 77) participants were included (26,32). Furthermore, both studies had large effect sizes (>0.8) suggesting there was an association between KS and autism symptoms in these studies.
A meta-analysis was possible for Intelligence Quotient (IQ) in seven studies, six cross-sectional and one cohort longitudinal (Fig. 2). This included a total of 490 participants across all ages: 248 patients with KS and 242 controls. To measure full-scale IQ, two studies used the Wechsler Adult Intelligence Scale (WAIS) (39), two studies used Wechsler Adult Intelligence Scale -Revised (WAIS-R) (40), one study used the Wechsler Intelligence Scale for Children-III (41), and two studies reported full-scale IQ scores, participants, SD and control data however the (IQ) test used was not listed. For the meta-analysis the study CI and the overall interval was set at 95%. The meta-analysis suggests an association between lower full-scale IQ and a KS diagnosis. There was a strong significant difference between patients with KS and control suggesting a negative association between full-scale IQ for patients with KS when compared to controls. The I2 result (I2 = 54%) showed moderately high heterogeneity (42), which could be due to the varied ages of participants and differing measure of full-scale IQ.

Social relations
Eight studies included measures relating to social behaviour, sexual function, sexual satisfaction, and sexuality (6,7,26,28,31,32,35) within the social relations subsection of the WHOQOL-100 (Table 5). With the exception of Turiff et al. (6), all studies measuring social relations found that patients with KS have lower scores than their controls.
Four studies (21,26,32,34) compared patients with KS against controls which allowed effect size to be calculated (Table 5).
Two studies found that patients with KS had an increased risk in developing negative social traits of anxiety, social responsiveness, and social awareness (28,29). Tartaglia et al. found that more than 25% (n = 42) of patients with KS scored mild, moderate or severe on all domains of the Social Responsiveness Scale (SRS) except for the social awareness domain (28). Furthermore, Van Rijn et al. identified a strong effect size (d = 2.016) when measuring social responsiveness using SRS in patients with KS when compared to controls (7).

Environment
There were limited measures of 'environment', WHOQOL-100 lists the subgroups of 'environment' as; financial resources, freedom, physical safety and security, health, and social care: accessibility and quality, home environment, opportunities for acquiring new information and skills, participation in and opportunities for recreation/leisure, physical environment, and transport. Only four studies measured education (21,24,25,33) and two measured financial resources (21,33).  There were no significant differences between controls and KS regarding mental illness. However, at the initial examination 41% of KS participants had a mental illness and which was significantly higher than controls (P < 0.0021).

(Continued)
European Journal of Endocrinology Acquiring new information and skills can be linked to school attainment and completing education. Herlilhy et al. reported that 34% of patients with KS (n = 87) did not complete high school, 55% completed high school and 10% studied further than high school (21). Results showed significant differences in five categories: lack of interest in schoolwork (P < 0.05), concentration difficulties (P < 0.005), speech difficulties (P < 0.05), lack of self-confidence (P < 0.05), particularly dependent on parents (P < 0.025) (21).
Skakkebaek et al. found that patients with KS (n = 132) were significantly less likely than controls to complete high school (P < 0.001) and at least 1 year of higher education (P < 0.01) (33). School performance was also significantly worse (P < 0.001) for patients with KS when compared to healthy controls (24). Turriff et al. looked at the highest education level obtained by patients with KS (n = 310) and found that 13.6% completed post-graduate education, 23.9% college, 22.6% part of college education, 13.6% technical school, 22.2% high school and 4.1% completed elementary or junior school (6). Herlihy et al. found that 36% of patients with KS (n =87) earned less than AUS$30,000, 27% earned between AUS$30,000 and 69.999 and 30% earned more than AUS$70,000 (21). Similarly, Skakkebaek et al. found that patients with KS (n = 126) had significantly lower (P < 0.001) household income compared to healthy controls (33).

Discussion
This is the first systematic review to provide an in-depth analysis of the associations between KS and QoL. The WHOQOL-100 provided a framework which allowed sufficient synthesis for many parameters and domains of QoL, which highlighted a disparity in the QoL between patients with KS and controls. Furthermore, the metaanalysis from the included studies indicates a lower fullscale IQ is associated with KS diagnosis.
Almost all patients (95.9% of n = 829) across 12 studies included in this systematic review reported that KS had negatively affected the QoL outcome measures. When calculated between patients with KS and controls, a significant effect size (Cohen's d) was present in most outcomes measured (91.8% of 49). The effect size within this review further quantifies the difference between KS and controls in outcome measures associated with QoL providing hence evidence of the negative impact of the KS diagnosis on patients' QoL.  (62) (17), Q-LES-Q (38) and SF-36 (37), showed poorer QoL scores for patients with KS compared to controls. This is consistent with previous research which supports overall impaired QoL in patients with KS (13,43).
Psychological outcomes were the most measured subgroup of QoL and 37 of the 45 outcome measures showed a statistical significance in scores indicating that KS diagnosis is increasing the risk for patients to develop a psychological disorder including cognitive impairment. The results of this systematic review support previous research which found that patients with KS had an increase in psychiatric comorbidities including autism, attention deficit hyperactivity disorder (ADHD), psychosis, personality disorders and developmental disorders (44,45,46,47). Furthermore, our meta-analysis showed that men with KS have a significantly lower IQ than healthy controls (24,26,27,32,34,35) (Fig. 2). Kennedy et al. noted the importance of IQ as a predictor of future success (48). Previous research also shows that lower IQ is associated with negative outcomes such as increased prosocial skill deficits, criminal behaviour, post-traumatic stress disorder and lower academic achievements (48,49,50,51,52,53). As these outcomes form subgroups of the QoL construct, it is essential to understand the effect that the diagnosis of KS has on the patient's IQ, in order to provide the necessary care and support at an early-stage post-diagnosis. Further research is necessary to investigate the effect that lower IQ may have on the QoL outcomes for patients with KS.
This systematic review suggests that men with KS are at higher risk than healthy controls to develop psychiatric disorders associated with autism spectrum symptoms, but  (54), while two studies suggested improved health-related QoL outcomes in people with less severe autism symptoms (55,56). Like KS, autism has a broad phenotype with a variety of symptoms ranging from disruptive language to socio-emotional traits. However, unlike KS, the awareness and research conducted on autism are far greater which has led to earlier diagnosis and relevant support for patients diagnosed with autism, with improved QoL outcomes and wider social understanding of autism. Evidence supports the presence of autism symptoms especially social behaviours in patients with KS, yet many patients do not receive appropriate investigations nor are diagnosed with autism spectrum disorders which can have a detrimental impact on their QoL outcomes (26,32,34). School attainment and behaviour were measured by five studies which found that boys with KS had significantly lower (P < 0.05) achievement and worse behaviours than healthy controls at all levels of education (6,25,27,33,57). Recent research in education and psychology shows that behaviours and attitudes in school have a direct correlation with work status, income earnings and social status later in later life (58). Although there is limited evidence to support that boys with KS have poor school attainment and behaviours, the consequences of this may have severe lifelong implications. Therefore, further research is needed to investigate this area to develop relevant supportive mechanisms at school for young boys with KS.
This systematic review found that the diagnosis of KS has a significant negative impact on the patients' erectile function and sexual satisfaction, which is most likely secondary to testosterone deficiency and psychological disorders associated with KS (31). Further research is required to address this problem. Similarly, our review suggests that patients with KS have more increased social anxiety and impaired social skills compared to controls (7, 26,28,35). This is supported by two earlier studies which provide evidence of the negative effect that low testosterone has on social anxiety (59,60).
In conclusion, our systematic review with narrative synthesis and meta-analysis, guided by the WHOQOL-100 as an overarching framework, provides evidence that patients with KS have impaired QoL compared to healthy males. Although evidence for overall QoL outcomes was limited, subgroup analysis helped to provide a greater understanding of the WHOQOL-100 subgroups, and the extent to which each of these are affected by patients with KS. Further research is needed to understand the impact the diagnosis of KS has on patients' QoL. A significant finding from this systematic review was the lack of a conditionspecific PROM for patients with KS. Development and validation of a KS-specific PROM that would encompass all domains of QoL for this patient group and would provide a quantifiable and validated measure for QoL is therefore essential.