To assess peer relationships of adolescents with juvenile primary fibromyalgia syndrome (JPFS) compared with matched classroom comparison peers (MCCPs) without a chronic illness. JPFS is characterized by chronic musculoskeletal pain, sleep disturbance, fatigue, and difficulty with daily functioning. Adolescents with JPFS often report problems with school and participating in peer activities, placing them at risk for social isolation from their peers and psychosocial adjustment problems.
Participants were 55 adolescents with JPFS (ages 12–18 years) from a pediatric outpatient rheumatology clinic and 55 MCCPs. Data on peer reputation and peer acceptance were collected from teachers, peers, and self report in a classroom setting with no focus on JPFS.
Adolescents with JPFS were perceived (by peer and self reports) as being more isolated and withdrawn and less popular. Adolescents with JPFS were less well liked, were selected less often as a best friend, and had fewer reciprocated friendships.
Our findings suggest that adolescents with JPFS are experiencing problems with peer relationships. Given the central role that peer relationships play in psychological development of children, and because peer rejection and isolation have been associated with subsequent adjustment problems, these findings are concerning. Longitudinal studies of adolescents with JPFS are needed to ascertain whether these patients are at long-term risk and will provide a foundation for the need for early interventions. Results are discussed within the context of earlier findings for other adolescents with chronic illness and rheumatic conditions, such as juvenile idiopathic arthritis, who demonstrated no social problems.
Fibromyalgia syndrome is a debilitating pain condition that occurs in 4% of adult women and 0.5% of adult men in the US. Adults with fibromyalgia often do not improve with treatment and experience significant social, emotional, and vocational disability (1–5). Fibromyalgia likely begins at an early age, and is increasingly being diagnosed in adolescents. Similar to adults, adolescents with juvenile primary fibromyalgia syndrome (JPFS) experience generalized musculoskeletal pain, multiple tender points upon palpation, sleep difficulties, and associated symptoms such as persistent fatigue, psychological distress, and irritable bowel syndrome (6). The overall incidence is not clear; however, adolescents with JPFS represent ∼7% of new patient diagnoses in rheumatology clinics (7). JPFS occurs more commonly in girls (6), and the age at onset is typically between 13 and 15 years. JPFS is often difficult to diagnose and treat because of the lack of objective diagnostic findings on laboratory tests, subjective nature of pain reports, poor response to conventional medical treatment, and high levels of distress expressed by patients.
Adolescents diagnosed with JPFS have unfortunately received relatively little attention in the research literature. Compared with adults, patients with JPFS face unique developmental and social challenges as they cope with daily pain and difficulties with daily activities. JPFS has been shown to compromise psychosocial functioning, evidenced by poor school attendance and increased emotional distress (8–10). In one study, adolescents with JPFS were found to miss an average of 5 days of school per month, putting them at risk for academic difficulties, decreased social contact, and peer rejection (9). With the lack of specific medical findings, adolescent patients and parents anecdotally report that it is difficult for them to explain to peers and school teachers why they often have problems engaging in everyday activities, participating in sports, and attending school regularly. This might serve to exacerbate the psychosocial impact of JPFS.
In the clinical setting, school-related social difficulties are often mentioned by parents and adolescents as a significant concern. Parents often report chronic school absenteeism, social withdrawal, and few friends. Adolescents with JPFS often state that they are rejected because teachers and peers do not understand the nature of JPFS and believe they are merely seeking attention or avoiding academic obligations. Current knowledge about social difficulties in JPFS has relied primarily on anecdotal reports or research studies using self-report or parent-report measures. No study to date has entered the school environment of adolescents with JPFS and evaluated actual peer relationships and social adjustment. This is especially significant within the context of considerable work demonstrating the central importance of peer relationships for normal social, emotional, and behavioral development of children, and indicating that peers are the preferred source of information about a child's social functioning (11–14). Peer relationships in school-age children have been shown to be a remarkably stable predictor of later psychosocial adjustment, and early disruption in these relationships is a risk factor for subsequent problems. Sophisticated research methodologies have been developed to assess social functioning of adolescents in the field of developmental psychopathology, but these methodologies have not been applied to the study of social functioning of adolescents with JPFS.
Our research group has successfully developed a methodology of classroom-based peer assessment to assess the social functioning of children and adolescents with chronic illness using a case–control design (15–17); this well-established design was used in the current study of adolescents with JPFS. In this design, information about peer reputation and peer acceptance for every student in a classroom is gathered from peers, without revealing the identity of the target subject to the class or revealing the health focus of the study. The best-matched comparison peer for the target subject (based upon closest date of birth, same sex, and same race) is selected from the class to be the matched classroom comparison peer (MCCP) for the study. This strategy permits collection of data without stigmatizing the adolescent with JPFS or biasing peer responses. Next, peer reputation and peer acceptance ratings for each adolescent are based on ratings from the entire class, increasing the robustness of measurement and decreasing subjectivity associated with self report or parent report. Finally, the matching procedure for selecting comparisons ensures that target status is not confounded with multiple social, economic, or cultural risk factors, and it does not have the limitations normally associated with a matching design (18).
The primary objective of this study was to evaluate whether adolescents with JPFS have more difficulties with peer relationships than MCCPs without a chronic illness. We hypothesized 1) that adolescents with JPFS would be described by peers, teachers, and self report as being more sensitive/isolated than MCCPs, and 2) that adolescents with JPFS would have fewer friends and be less well liked than MCCPs.
PATIENTS AND METHODS
This study was the first phase of a larger ongoing research program evaluating the social functioning of youth with chronic illness (16, 17, 19). Participants in this study were 55 adolescents with JPFS recruited from a pediatric rheumatology clinic and 55 MCCPs selected from the classroom of each target adolescent with JPFS.
Adolescents with JPFS.
To be eligible for participation, adolescents with JPFS had to be 12–18 years of age and be attending regular school (no full-time special education or full-time home-bound instruction). Diagnosis of JPFS was made by a pediatric rheumatologist and verified by chart review. The criteria were generalized musculoskeletal aching at ≥3 sites for ≥3 months in the absence of other underlying conditions; laboratory tests within normal limits (hemoglobin, thyroid function tests [free T4 and thyroid-stimulating hormone]); severe pain in 5 of 11 tender point sites; and at least 3 of the following criteria: fatigue, irritable bowel, poor sleep, chronic anxiety or tension, chronic headaches, subjective soft tissue swelling, numbness, and pain modulation by weather factors or stress (6). Patients with other chronic rheumatic diseases were excluded. This study was carried out in compliance with current ethical standards for research with humans and was approved by the hospital institutional review board.
After the adolescents with JPFS and their parents signed a written consent/assent and Health Insurance Portability and Accountability Act release form, permission to conduct the study in the adolescent's classroom was sought from their school principal. Written consent was obtained from the teacher and the parents of the students in the class. A detailed description of recruitment procedures is provided in Figure 1.
Teacher and classroom data collection.
Teacher-report sections of the measures were administered by the research assistant during an individual visit with the teacher prior to the classroom visit. After consent forms were returned, peer nomination data were collected in the classroom of the adolescent with JPFS.
Selection of matched classroom comparison peers.
MCCPs without a chronic illness were chosen using the following method. A class list with a roster of students who provided consent to participate was generated. From this list, 1 MCCP was chosen from each class based on having the closest date of birth and being of the same sex and race as the adolescent with JPFS. The health status of the MCCP (i.e., absence of chronic illness) was confirmed during a subsequent telephone call to the family. (The MCCP families were called to request their participation in a second study related to family functioning, which is ongoing. If they declined to participate, or were ineligible due to the existence of a chronic medical condition, the family of the next closest matched comparison peer was called. A total of 80% of MCCPs were first-choice peers, 16.4% were second- or third-choice peers, and 3.6% were fourth- or fifth-choice peers.)
Measures were selected on the basis of 1) psychometric properties, 2) developmental appropriateness, and 3) previous use in studies of social adjustment and chronic illness.
Revised Class Play.
The Revised Class Play (RCP) is a 30-item measure that has been found to measure 4 dimensions of social reputation (What is the child like?) in a broad age range of children (20): popularity/leadership (e.g., someone who everybody likes), prosocial (e.g., someone who is polite), aggressive/disruptive (e.g., someone who is too bossy), and sensitive/isolated (e.g., someone who is often sad). Ratings were obtained from teachers and peers, and self ratings were obtained from the adolescents. Participants were instructed to imagine that they were the directors of an imaginary play and were asked to “cast” members of the class into a variety of roles. Each participant was provided a list of children in the class who were the same sex as the adolescent with JPFS (to eliminate sex-role stereotyped nominations and increase the likelihood that the adolescent with JPFS would be named). Participants were asked to nominate 1 adolescent from the list for each role. The same individual could be picked for more than 1 role. Next, they completed the RCP a second time, rating the roles that they thought would be most appropriate for themselves. The dimensions of the RCP have been shown to be both internally consistent and stable (15, 21) and have documented predictive validity (12, 13). Scoring for the RCP was conducted by assigning each adolescent an item score based upon the number of times they were nominated for each role (item). Item scores were summed for each source to create subscale scores on each of the 4 dimensions and 4 nonsocial attributes. Data on RCP scores for each classroom were transformed into Z scores (mean ± SD 0 ± 1) to adjust for unequal class sizes and participation rates.
Three Best Friends.
The Three Best Friends measure is a stable indicator of social acceptance (Is the child liked?) and mutual friendships (22) and has been used in a number of past studies in children with chronic illness (15–17, 23). Adolescents were asked to choose the 3 peers in their class whom they thought of as their best friends. Using these data, each adolescent was given 1) a social acceptance score based upon the number of times he or she was chosen as a friend by peers, and 2) a mutual friendship score indicating how many of their friendship selections were reciprocated. All scores were converted to Z scores separately for boys and girls to adjust for unequal numbers of participating adolescents and the classroom sex composition, which might be expected to affect the scores.
Like Rating Scale.
This measure of social acceptance was utilized because it provides a measure of overall popularity that is complementary to the best friend measure of social acceptance, and it has been shown to be a reliable and stable measure of likeability (24). Adolescents were asked to rate all of their classmates on a 5-point scale (where 1 = someone you do not like and 5 = someone you like a lot). An average liking score was computed for each adolescent based on the ratings they received from peers. These data were also transformed into Z scores (separately for boys and girls) to facilitate comparison with the other measures.
In this study, as in our prior research with children with chronic illness, 9 additional roles were added to the end of the RCP relating to nonsocial attributes that may be affected by chronic illness such as being sick a lot (3 roles), academic competence (2 roles), athletic ability (2 roles), and physical appearance (2 roles) (23). The items were identical in format and were scored in the same way as the other items on the RCP.
In this study, the power to detect moderate to large effect sizes (0.60) on RCP variables with a sample size of 55 matched pairs (2-tailed test, α = 0.05) was 88%.
Paired-samples t-tests were used to test the differences between the participants with JPFS and the MCCPs in peer reputation, peer acceptance, and nonsocial attributes. To control for the risk of Type I error due to multiple comparisons, Holm's correction (25) was used for each of the 4 peer reputation dimensions of the RCP based on peer, teacher, and self ratings, as well as for the 3 peer acceptance measures (total of 15 comparisons; α = 0.05–0.003). Holm's correction was not applied to comparisons for nonsocial attributes because these analyses were exploratory. Finally, regression analyses were conducted to test for differential age effects on peer reputation measures for participants with JPFS and MCCPs.
To compare our findings with prior research using identical methodology to study social functioning in another pediatric rheumatic condition (juvenile idiopathic arthritis [JIA]), descriptive information (means and SDs) from a previously published study (26) is presented in Table 1. Statistical comparisons with the current sample were not conducted because it was a different cohort. However, the results of the current study can be more clearly highlighted when contrasted with the same peer nomination measures in a different rheumatic condition.
Table 1. Z scores for patients with JPFS and matched controls on peer reputation and social acceptance (mean ± SD 0 ± 1), with JIA comparison*
JPFS (n = 55)
Controls (n = 55)
JIA (n = 57)
Values are the mean ± SD unless otherwise indicated. JPFS = juvenile primary fibromyalgia syndrome; JIA = juvenile idiopathic arthritis; RCP = Revised Class Play. JIA comparison from Reiter-Purtill et al study (26).
Significant with Holm's correction applied for multiple comparisons (25).
A total of 99 patients from the pediatric rheumatology clinic who met the diagnostic criteria for JPFS were screened. Of these, 11 were excluded because they were home schooled and 2 were eliminated due to full-time special education. Of the 86 eligible patients, we were able to contact 81 families, and 59 families consented to participate (72.8%). Of 59 schools, 55 gave permission for the study to be conducted. Data were reported for 55 adolescents with JPFS and 55 MCCPs (n = 110) from 55 classrooms (1,289 peer ratings representing >80% of class participation). The 55 adolescents with JPFS were primarily female (94.5%) with an average age of 15.04 years. A total of 49 (89%) were non-Hispanic white and 6 (11%) were African American. MCCPs were matched in terms of sex and age. All pairs were matched in terms of race except for 1 (because the target child was the only child in the classroom of that race).
Social reputation (What is the child like?).
After correcting for multiple comparisons, findings from peer and self nominations demonstrated that adolescents with JPFS received significantly more nominations on the sensitive/isolated dimension and significantly fewer nominations on the popular/leader dimension (Table 1) compared with MCCPs. Based on the standard deviation and inspection of the distribution of scores, peer reputation ratings of adolescents with JPFS were found to have a much wider spread of scores (greater variability) than MCCPs (Levene's test for homogeneity of variances F = 18.14, P < 0.000). Therefore, we calculated the number of participants who received a rating ≥1 SD above the mean on the sensitive/isolated dimension in each group (JPFS and MCCP). This technique has been used to identify subgroups of children in prior research with nomination sociometric data (27, 28). Ratings by peers, teachers, and self ratings all indicated that a higher percentage of adolescents with JPFS (30.9%, 23.6%, and 27.3%, respectively) fell above the cutoff compared with MCCPs (9.1%, 10.9%, and 9.1%, respectively). Adolescents with JPFS were more than twice as likely as their peers to be having difficulties in this area. No differential age effects for peer reputation measures were found.
Teacher ratings were inconsistent with peer and self ratings on some dimensions of the RCP. Teachers tended to nominate adolescents with JPFS for the sensitive/isolated dimension more frequently, but the difference did not reach statistical significance. Also, unlike peer and self ratings, teachers did not nominate adolescents with JPFS less frequently on popularity/leadership. Adolescents with JPFS were seen by teachers as significantly less aggressive/disruptive than MCCPs.
In comparing results from this study with standardized peer nomination data from a JIA sample (Table 1), the differences between the JPFS and JIA samples were striking. Peer, self, and teacher ratings all tended towards adolescents with JPFS being nominated more often as being sensitive/isolated. Also, peer and self reports indicated significantly lower popularity/leadership nominations for adolescents with JPFS. In the JIA sample, there were no significant differences in any of the 4 dimensions of the RCP from the perspective of peers, teachers, or self report.
Social acceptance (Is the child liked?).
Adolescents with JPFS were significantly less well liked, were selected less often as a best friend, and had fewer reciprocated friendships than MCCPs (Table 1). In contrast, the JIA sample showed no differences from MCCPs in social acceptance ratings.
Paired-samples t-tests indicated that adolescents with JPFS were more likely to be seen as being sick a lot, less athletically competent, and less physically attractive by their peers. They were not perceived as being different from MCCPs in terms of their academic competence (Table 2). Although these were exploratory findings, they are consistent with the lower overall nominations for adolescents with JPFS in terms of social acceptance. Nonsocial attributes were not reported for the JIA study, therefore comparative scores are not available.
Table 2. Z scores for patients with JPFS and matched controls on nonsocial attributes (mean ± SD 0 ± 1)*
JPFS (n = 55)
Controls (n = 55)
Values are the mean ± SD unless otherwise indicated. See Table 1 for definitions.
Significant at α ≤ 0.005 (no correction used for multiple comparisons).
This study assessed the peer relationships of adolescents with JPFS. Data were collected from teachers, peers, and self report in a classroom setting with no focus on JPFS. Our findings demonstrate that adolescents with JPFS are consistently rated as more sensitive/isolated than their healthy peers from the perspective of peers, teachers, and self nominations. They are also seen as demonstrating fewer popular/leadership behaviors. Finally, from the perspective of peers, adolescents with JPFS have significantly fewer friends, fewer reciprocated friendships, and are less well liked. Although there was some variability in perceptions about adolescents with JPFS among their peers, adolescents with JPFS were much more likely to have social problems characterized by social isolation and peer rejection. Interestingly, teachers did not perceive adolescents with JPFS as having fewer popular/leadership behaviors, but as being significantly less aggressive/disruptive. Our group has also reported less aggressive/disruptive behavior for other groups of teenagers with chronic illness (16, 29) and we have speculated that chronic illness may have a protective social effect. However, in this study, teacher ratings on the popularity/leadership and sensitive/isolated dimensions suggest that teachers do not appear to be aware of the extent to which adolescents with JPFS are having peer-related difficulties. This is not surprising given that adolescents with JPFS may be more withdrawn and less disruptive in class, and therefore less likely to come to the attention of teachers in a busy classroom setting.
The results of our study support clinical impressions of providers and reports by parents and adolescent patients with JPFS concerning the social difficulties these patients experience with their peers. The specific reasons for these social problems are not clear at this time. The comparison group in the current study consisted of MCCPs without a chronic illness. Therefore, one potential reason adolescents with JPFS were less integrated into the peer group was that they were perceived as being sick a lot and less athletically competent than MCCPs. However, prior research using the same methodology to examine social relationships of children and adolescents with JIA (who also experience pain, physical limitations, and school absences) demonstrated no problems in social functioning in children with JIA (26). Children with a number of other chronic diseases including cancer (15, 16, 19, 23, 30) and sickle cell disease (31) also have not reported significant social problems. All of these studies demonstrated that children with chronic illness appear to be remarkably similar to their comparison peers on measures of social and emotional functioning, suggesting a great deal of psychological hardiness.
Why do adolescents with JPFS manifest problems with peers when other groups of adolescents with chronic illness show a different pattern of results? It is conceivable that the greater difficulty with social functioning among adolescents with JPFS may be a result of greater incidence of psychiatric symptoms (such as increased anxiety, social withdrawal, and mood difficulties). The research literature has documented a substantially higher lifetime risk for mood and anxiety disorders in patients with fibromyalgia compared with patients with rheumatoid arthritis (32). We have found that adolescents with JPFS are more likely to have depressive symptoms than adolescents with chronic low back pain (9). We have also found that almost 40% of adolescents with JPFS have at least 1 psychiatric diagnosis (9, 33), usually a mood or anxiety disorder. It is also possible that adolescents with JPFS have more social problems because of challenges related to making a definitive JPFS diagnosis. Our prior studies were conducted with children with very specific medical diagnoses. Future work needs to explore potential linkages between social functioning, mood difficulties, and JPFS.
Unfortunately, this pattern of combined peer rejection and social isolation has been found to be moderately stable over time for children without a chronic illness, even when a new peer group gives ratings during subsequent academic years (34). Additionally, for children without a chronic illness, social isolation and peer rejection have been linked to subsequent difficulties with depression, anxiety, loneliness, negative self worth, and substance use (12–14, 35, 36). Rejection by peers in elementary school is also linked to increased risk for leaving school early, academic underachievement, involvement with antisocial peers, poor work-school engagement, and unemployment (37–39). The long-term implications of early peer relationship difficulties for adolescents with JPFS are currently unknown.
This study had several limitations. First, the adolescents with JPFS were a clinical sample of patients. The incidence of JPFS in the community is not known; therefore, the findings cannot be generalized to nonpatient samples. It is possible that non–treatment-seeking adolescents with JPFS do not show the same level of peer difficulties as those who come to the attention of the health care system. Second, peer data were not available for the 10% of adolescents with JPFS who were ineligible for the study because they were not attending regular school. These patients may have even more severe social difficulties, but we were unable to evaluate them due to the school-based methodology in this study. Finally, we report no data on the emotional well-being or social skills of adolescents with JPFS. Future work will help clarify some of these limitations.
In summary, our findings suggest that adolescents with JPFS seen in a pediatric rheumatology clinic setting are more likely to be having significant difficulties with respect to peer relationships including social rejection and isolation, unlike patients with other rheumatic conditions such as JIA. Psychosocial assessment and referral for supportive services should be recommended for patients who report such problems to health care professionals. Future studies should focus on gaining a better understanding of the linkages between JPFS and social problems, as intervention strategies will vary significantly. For example, if adolescents with JPFS report more depression, treating this mood problem could help social functioning. Alternatively, if adolescents with JPFS have social skills deficits, then social skills training might be advantageous. Recent work has demonstrated the effectiveness of school-based programs to enhance the peer relationships of children with physical disabilities (40, 41), which might be helpful for adolescents with JPFS. Finally, longitudinal studies of adolescents with JPFS are essential to assess whether these early social difficulties predict later problems with psychosocial adjustment and emotional functioning. Early identification and targeted interventions for adolescents with JPFS may lead to more comprehensive and successful treatment approaches, preventing the symptoms and dysfunction reported by many adults with fibromyalgia.
Dr. Kashikar-Zuck had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Kashikar-Zuck, Graham, Noll.
Acquisition of data. Kashikar-Zuck, Lynch, Graham, Swain, Mullen, Noll.
Analysis and interpretation of data. Kashikar-Zuck, Lynch, Mullen, Noll.