Relationships Between Family and Parent Characteristics and Functional Abilities in Children with Recurrent Pain Syndromes: An Investigation of Moderating Effects on the Pathway from Pain to Disability

Abstract

Objective To identify family characteristics associated with children’s ability to function with recurrent pain. Methods Seventy-eight children ages 7–17 years with recurrent pain syndromes [migraine headache or recurrent abdominal pain (RAP)] were recruited from clinic settings. Children completed pain diaries and the Functional Disability Inventory (FDI). Mothers and fathers completed self-report measures of psychological distress, and mothers reported on family environment. Results Controlling for the influence of pain intensity, family environment and parental distress jointly predicted children’s ability to function with pain. Among children with migraine, family environment moderated the relationship between pain and functional disability; in this group, greater pain associated with more functional disability in children from disruptive family environments, but not in children from more adaptive family environments. Conclusions For some pediatric recurrent pain sufferers, family characteristics associate with the extent of pain-related disability and may help identify children likely to experience more impaired functioning in response to recurrent pain.

Children and adolescents with chronic or recurrent pain syndromes range broadly in the degree of functional disability that results from their pain. Significant impairment due to pain can develop into “pain-related disability syndrome” (Bursch, Walco, & Zeltzer, 1998) and can affect many areas of the child’s life, including school attendance and performance, social functioning, sleep and other daily habits, and family relationships (Palermo, 2000). The association between pain and functional disability is not always linear. Rather, numerous factors inherent to the child and his/her context appear to influence the extent of disability.

Why some children with recurrent pain are able to function adaptively in the face of their pain while others become severely disabled is a question researchers have begun to address in recent empirical studies. Claar, Walker, and Smith (1999) found that perceived academic, social, and athletic competence moderated the relationship between pain symptoms and degree of functional disability in a group of adolescents and young adults with irritable bowel syndrome. Pain and disability were more strongly related in participants with lower perceived competence, suggesting that pain-related disability may have a reinforcing effect among youth who might benefit from avoiding activities at which they feel unsuccessful. Children’s coping styles and psychological distress also appear to play a role in functional disability (Reid, Lang, & McGrath, 1997; Walker, Smith, Garber, & Van Slyke, 1997). More depressed and anxious children tend to experience more functional impairment compared to children with similar chronic or recurrent pain problems but less emotional distress (Kashikar-Zuck, Goldschneider, Powers, Vaught, & Hershey, 2001; Palermo, 2000).

Beyond individual child factors, parent responses to children’s pain may unintentionally reinforce functional disability. Dunn-Geier et al. (1986) found that maternal responses in a pain-oriented situation differed between children with pain who experienced frequent school absences and those who attended school well despite their pain. More recently, Chambers, Craig, and Bennett (2002) demonstrated that mothers’ pain-promoting or pain-reducing responses influenced girls’ pain behaviors in an experimental condition.

Parents’ own distress, whether physical or emotional, may affect children’s responses to pain. Parental modeling of pain has been identified as a potential influence on how children respond to pain (Schanberg et al., 2001). Although the evidence is mixed, some studies have found significant parental psychopathology in families of youth with chronic or recurrent pain. For example, research on parents of children with recurrent abdominal pain (RAP) reveals elevated rates of anxiety and depression compared with parents of healthy controls (Garber, Zeman, & Walker, 1990; Liakopoulou-Kairis et al., 2002). Regarding functional disability outcomes, the impact of parental psychological distress has not yet been clearly established.

Taken as a whole, the existing research supports the theory that parent characteristics and parental responses to pain influence how children learn to manage their pain symptoms. Yet neither child behavior nor parent behavior alone seems sufficient to explain the complex relationship between pain and functional outcomes. Thus, it may be useful to adopt a broader family systems framework (Minuchin et al., 1975) to understand children’s responses to chronic or recurrent pain. Covelman and colleagues suggest that a child’s pain symptoms become an integral part of family transactions (Covelman, Scott, Buchanan, & Rossman, 1990). Pain-related disability may assume its own role in this transactional context—if this occurs, functional disability may become more closely linked to family environment factors than to the pain itself.

Across many pain conditions, including RAP, headache, back pain, and musculoskeletal pain, there is some evidence that subgroups of individuals who suffer from these symptoms also experience significant disruption in family relationships (Scharff et al., in press). The most consistent patterns identified include high family conflict, strong enmeshment and interdependence among family members, disorganization, and rigid family control. The patterns generally mirror characteristics highlighted in Minuchin et al.’s (1975) descriptions of psychosomatic families. Such patterns were first proposed in early studies in the area of family functioning and chronic pain (e.g., Liebman, Honig, & Berger, 1976; Norfleet, Hammett, Lichte, Lukensmeyer, & Payne, 1982), but more recent empirical studies still cite these characteristics in families of pain patients (e.g., Kopp et al., 1995; Naidoo & Pillay, 1994; Sherry & Weisman, 1988). These studies suggest that the familial context in which pain occurs can affect the experience of the pain, including the resulting degree of functional disability (Payne & Norfleet, 1986).

Although parental psychological distress and family environment characteristics have been explored in descriptive studies of children and adolescents with chronic or recurrent pain, few studies have attempted to fit these systemic influences into a framework to account for the nonlinear connection between pain symptoms and functional disability. One exception is Chambers (2003), who illustrates how family factors can affect all aspects of the child’s or adolescent’s experience of chronic pain. In response to the recent call for the development of system-based theoretical models of pain-related disability in children and adolescents (Palermo, 2000), it is important to explore the potential role that family characteristics may have in explaining individual variation in functional outcomes of pediatric pain syndromes. A variety of family factors could relate to pain-related disability, but we limit our focus to those that have been most widely identified as characteristic of families of individuals with chronic or recurrent pain. Thus, we are most interested in the roles of family conflict, rigidity or tight control in the family, a chaotic or disorganized family environment, and enmeshment among family members. Enmeshment in particular is a difficult construct to operationalize; based on Minuchin et al.’s (1975) description of enmeshment as a lack of psychological differentiation among family members and a general blurring of interpersonal family boundaries, we rely on the similar concept of low independence among family members to approximate enmeshment.

This study’s objective is to determine whether and how family characteristics relate to functional disability in a group children suffering from RAP or migraine headache. These two types of pain problems were selected because of their relatively high prevalence in pediatric populations (Hunfeld et al., 2001; McGrath, 1990). Previous studies have incorporated children with these two types of pain into single recurrent pain samples and have found that the two groups are highly similar (Liakopoulou-Kairis et al., 2002) and often overlapping (Anttila, Metsahonkala, Mikkelsson, Helenius, & Sillanpaa, 2001).

We hypothesized that (a) controlling for the child’s reported pain severity, mothers’ and fathers’ psychological distress would predict the extent of the child’s functional disability; (b) controlling for reported pain severity, family conflict, enmeshment (i.e., low levels of independence), rigidity (i.e., high levels of control), and disorganization would predict the extent of functional disability; and (c) these parental and family characteristics would moderate the relationship between pain severity and functional disability.

Method

Participants

Children and adolescents ages 7–17 years with diagnoses of RAP (n = 30) or migraine headache (n = 48) were recruited for participation in an assessment and biofeedback treatment study. Parents of participating children also took part in the study. Recruitment occurred in pediatric clinics, including neurology, gastroenterology, pain management, and primary care sites. Inclusion criteria required that participants have pain on at least a weekly basis or at least 5 days per month, without an organic diagnosis accounting for the pain. For those participants who complained of both abdominal pain and migraine headache (n = 14), pain group assignment was based on referral source and primary complaint.

Procedure

All measures used in this study were administered in the baseline assessment session before participation in the intervention study. Informed consent procedures, approved by the hospital’s institutional review board, were conducted with parents and children before administering the study protocol.

Measures

Pain Symptoms and Pain Ratings

To collect information on pain symptoms, physician diagnosis plus a pain symptom checklist (completed in interview format with the child and parent by a clinical psychologist or psychology intern) were used to determine whether children met International Headache Society criteria for migraine (Anonymous, 1988) or Rome II criteria as revised by the pediatric working group (Rasquin-Weber et al., 1999) for RAP. Because the IHS criteria were developed for adults, children were retained in the sample even if they did not meet full IHS criteria for migraine.

Pain diaries were mailed to participants before their assessment interview and were begun shortly before or concurrent with the completion of other study measures. Diaries were completed for 3 weeks and consisted of four daily ratings of pain intensity using a 5-point scale ranging from 0 (“no pain”) to 4 (“very severe pain; hurts so much I can’t do anything”). The behavioral component of the anchor was incorporated to assist children at the young end of our participant age range in responding validly to the scale. A Pain Index was calculated by averaging the four daily ratings over the 3-week period. Highest pain and number of pain-free days during this period also were calculated. Numerous studies have relied on similar diary methods to assess pain intensity (e.g., Sanders, Shepard, & Cleghorn, 1994; Scharff, Marcus, & Masek, 2002), and the diary used in this study was pilot-tested on groups of children with headache and abdominal pain. Previous studies indicate that individuals with RAP and migraine display similar means on this index (Scharff et al., 2002). In this study, eight participants failed to return completed pain diaries and were excluded from subsequent data analyses; these participants did not differ from the participants in the final sample on any demographic or pain characteristics. Among those comprising the final sample, 94.8% adhered perfectly to the diary completion task, with no missing information.

Functional Disability

The Functional Disability Inventory (FDI; Walker & Greene, 1991) is a self-report measure of current physical and psychosocial functioning because of physical health status. The 15 items assess function across a range of daily activities over the past 2 weeks (e.g., going out with friends, participating in gym class or sports, being in school all day). Respondents rate the level of difficulty of each item from 0 (“no trouble”) to 4 (“impossible”). A total score is derived from the sum of the items. The measure has demonstrated internal consistency and test–retest reliability and correlates with school absence rates (Walker & Greene, 1991). Currently, the FDI is the most common measure of functional disability for pediatric pain populations. For this study, internal consistency was excellent with a Cronbach’s alpha of .93.

Parental Psychological Distress

Both parents (whenever possible) completed the Revised Symptom Checklist-90 (SCL-90-R; Derogatis, 1977). The SCL-90 is a widely used measure of psychological distress, selected for this study because of the range of emotional and behavioral domains included and the availability of a global summary score. Additionally, both the SCL-90-R and its short form (the Brief Symptom Inventory) have been used in previous studies of parents of youth with chronic or recurrent pain conditions. The 90 items form nine dimensions measuring somatization, obsessions/compulsions, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, and psychoticism, which are summed to form a global severity index and a positive symptom distress index. The measure has demonstrated internal consistency and high test–retest reliability (.87–.90 for the summary indices) and correlates strongly with other well-established measures of psychological distress (Derogatis, 1977). The global severity index was used in our analyses, as we were interested in parents’ general psychological distress as opposed to more specific dimensions of psychological functioning. Internal consistency for the global severity index was .96 for the mothers and .95 for the fathers.

Family Environment

The Family Environment Scale (FES; Moos & Moos, 1986) was used to assess family environment. After reviewing several common family assessment tools, the FES was determined to be best suited to the purposes of this study because it included subscales that captured the constructs we wished to measure. Additionally, the FES is the family functioning assessment most commonly used in studies of chronic and recurrent pain. This 90-item instrument measures social–environmental dimensions of family functioning in the areas of interpersonal relationships, personal growth, and basic organizational structure of the family. The full instrument was administered, but only the Conflict, Independence, Organization, and Control subscales were used in analyses, as these are the dimensions on which notably high or low scores have characterized families of pain patients in previous studies. The FES has been found to discriminate well between psychiatrically disturbed and matched healthy families and possesses strong psychometric properties (Flor, Turk, & Rudy, 1987; Moos & Moos, 1986), with good test–retest reliability and adequate internal consistency scores for the subscales (.61–.78). The measure’s authors note that internal consistency is typically lower in homogenous samples, as was the case for some subscales in our sample (αs = .77, .44, .57, and .58 for Conflict, Independence, Organization, and Control subscales, respectively). Given our conceptual interest in these scales and recommendations by Moos (1990) not to over-rely on internal consistency values for the measure, we retained these scores for use in our data analyses. We relied on maternal report for the FES, given the need to limit the measures sent home for fathers (who generally did not accompany the child to the baseline interview) to complete and return.

Results

Seventy-eight children ages 7–17 were included in our data analyses. The overall completion rate among families who initially consented to the study was 89%. No differences were observed between dropouts and completers in terms of the child’s age, gender, or pain duration. Of the 90 participants who completed the study, 12 were excluded from final analyses because of missing data. The resulting sample had a mean age of 12.1 years (SD = 2.9) and was predominantly (67.9%) female, consistent with gender differences observed in the clinical population of pediatric recurrent pain patients. Most participants were White (80.8%), with 11.5% African American and 7.8% other ethnicities. Children in the sample came from highly educated families; mothers reported a mean of 16.1 years of education (SD = 2.8) and fathers reported a mean of 16.5 years of education (SD = 3.0). Additional information about the families of study participants is summarized in Table I.

Pain Characteristics

The mean duration of participants’ pain was 27.2 months; duration varied widely across the sample with a range of 3–96 months (SD = 22.6). Participants reported a mean pain index score (representing averages of the four daily pain reports across the 3-week period) of 1.3 (SD = 0.9) and a median index score of 1.0 on the 0–4 scale. Number of pain-free days was not normally distributed and was recoded into a dichotomous variable indicating the presence or absence of any pain-free days during the diary period. Age correlated moderately with these pain reports, with older children reporting more intense pain, r = .26, p < .05, and fewer pain-free days, ρ = –.26, p < .05. T tests revealed no significant sex differences on pain characteristics. The migraine and RAP groups reported similar pain index scores and maximum pain scores, but children with migraines reported pain of longer duration, t(74) = 2.5, p < .05, despite no significant group differences in age. Children with migraines were more likely to report pain-free days during the 3-week diary period compared with children with RAP, χ² = 6.1, p < .05.

Regarding referral source of participants, 70.1% of the sample came from specialty care clinics and 29.9% came from primary care. Participants from specialty care clinics reported greater pain, t(54.9) = 3.0, p < .01, and more functional disability, t(52.8) = 2.6, p < .05, compared with participants from primary care. The groups did not differ significantly on the duration of pain symptoms.

Correlations of Functional Disability and Parent/Family Variables

Descriptive statistics for the FDI, maternal and paternal SCL-90 scores, and FES scores are summarized in Table II. For the SCL-90 and FES outcomes, raw scores were converted to T scores (with population mean of 50 and standard deviation of 10) using guidelines provided in the manual for each measure. FDI raw scores were nonnormally distributed, thus a square-root transformation was performed to yield a normal distribution used for analyses. T tests revealed no significant differences on any of these measures by child’s sex or pain diagnosis. Table III summarizes zero-order correlations among the variables of interest. Partial correlations between FDI scores and background characteristics (including pain duration and reports of parental pain symptoms) controlling for pain intensity were nonsignificant except for a trend-level inverse relationship between FDI and maternal education, r = –.20, p < .10. Therefore, none of these background variables was included in the subsequent regression analyses.

Predicting Functional Disability with Parental and Family Characteristics

Preliminary analyses of the data indicated that they met the assumptions necessary for multiple regression analyses. No outlier cases were identified. Using hierarchical multivariate regression analyses, we examined the extent to which parental psychological distress and family environment characteristics predicted functional disability beyond what could be explained by the child’s pain intensity. For each regression equation, pain intensity was entered on the first step and the parent/family variable entered on the second step to determine the additional variance accounted for by each parent/family factor. Controlling for effects of pain intensity, statistically significant relationships emerged between functional disability and four of the six hypothesized predictors: maternal psychological distress (β = .29, r² change = .08, p < .01); family conflict (β = .43, r² change = .18, p < .001); family organization (β = –.30, r² change = .09, p < .01); and family independence (β = –.25, r² change = .06, p < .05). Paternal psychological distress (β = .11, r² change = .01) and family control (β = .16, r² change = .03) were not statistically significant predictors of functional disability.

A larger multivariate model was then tested to examine the combined effects of the parent and family characteristics that were uniquely significant predictors of functional disability in the three-variable models. Combined, these variables—maternal distress, family conflict, family organization, and family independence—explained a significant portion of the variance in disability beyond that accounted for by pain intensity, F(4, 64) = 8.3, p < .001. Controlling for the effects of pain intensity and the other parent and family factors, lower levels of family independence and higher family conflict were most closely associated with functional disability in the multivariate model (Table IV).

Tests of Moderating Effects

We hypothesized that maternal psychological distress, paternal psychological distress, and family environment would each moderate the relationship between pain intensity and functional disability. Moderator effects were explored using the process described by Baron and Kenny, 1986. To assess the possible moderating effect of family environment as a single construct, a composite score was created to represent the four family environment domains of interest. First, dummy variables were created for each of the four domains, assigning a value of 1, if the T score fell one standard deviation or more beyond the normative range in the direction of interest (i.e., elevated for the domains of conflict and control, below normal for the domains of independence and organization), and 0 if the T score fell within the normative range. These scores were then summed to indicate family environment across the four areas, with higher scores representing more maladaptive family environments. The summary score was positively skewed, with a mean of 0.70 (SD = 0.88).

In the full sample, neither parental distress nor family environment significantly moderated the relationship between pain and functional disability in hierarchical regression analyses incorporating two-way tests of interaction (i.e., pain × family functioning). To understand these findings better, we conducted further exploratory analyses stratifying the sample by diagnosis. In these exploratory analyses, family environment significantly moderated the relationship between pain and functional disability in the migraine group but not in the RAP group (Table V), based on the criteria for moderation posited by Baron and Kenney (1986). Post-hoc probing (Holmbeck, 2002) revealed that in the migraine group, greater pain was more directly associated with increased functional disability in children from more disrupted family environments. Among migraine sufferers from families with more adaptive family environments, increased pain intensity did not necessarily lead to increased functional disability. Although family environment failed to attain significance as a moderating variable in the RAP group, the general direction of the relationships is similar across the two diagnostic groups. Interactions also were tested for the individual FES subscales, but the moderating effect was obtained only with the composite family environment score.

Discussion

Context has long been recognized as a powerful determinant of the pain experience (Beecher, 1956). For children and adolescents, family is arguably the most central element of the social context and, therefore, a logical place to search for factors to account for differences in functional ability across individuals with similar complaints of pain. The open systems model of pediatric pain (Covelman et al., 1990; Minuchin et al., 1975) suggests that family transactional patterns can amplify or reduce a child’s symptomatic behavior. Using this model as a theoretical foundation, we explored whether parental and family characteristics associated with children’s pain-related disability after accounting for the effects of pain intensity.

As a group, the children and teens in our study had fairly long histories of pain, with a mean duration of over 2 years. Although we could not determine causal relationships in this cross-sectional study, we found no evidence in our data to suggest that pain is directly related to parental psychological distress or disruptions in family environment. Instead, more complex interrelationships appear to exist among pain, family characteristics, and the child’s pain-related functional disability. It is important to note that the mean scores on measures of parental distress and family environment reported in this study fell within the normative range. Furthermore, the relatively low internal consistency scores on the FES in our sample may indicate somewhat restricted variance across respondents on this scale. Significant findings regarding these variables, therefore, indicate relative differences and relationships within the context of generally normative levels of parental and family functioning.

The results of the hierarchical regression analyses support our hypotheses that accounting for pain intensity, parent distress and family characteristics relate to functional disability in this sample of pediatric recurrent pain sufferers. As a group, these factors explain 31% of the variance in functional disability after controlling for the effects of pain intensity. Within the multivariate context, levels of family conflict and family enmeshment (or low independence as measured by the FES) appeared to relate most closely with functional disability in this sample.

In clinical experience as well as in previous descriptive studies (e.g., Payne & Norfleet, 1986; Sherry & Weisman, 1988), observations of high levels of enmeshment in the families of children with chronic or recurrent pain are common. In highly enmeshed families, pain-related disability may preserve and extend parents’ care-taking roles and reduce the child’s independence. In families with high conflict, the child’s pain and disability may shift the focus away from sources of conflict, thus altering the dynamics among family members in ways that may subtly reinforce pain expression and disability. The results of this study offer empirical evidence to support the clinical observation that patterns of family conflict and enmeshment may influence the functional status of children with recurrent pain syndromes, regardless of pain levels. Caution must be used in interpreting these results, however, as self-report measures can only roughly approximate complex constructs like family environment or functional disability. The low internal consistency of the Independence subscale of the FES in our sample also calls for cautious interpretation. Future studies could improve on our methods through the use of more sophisticated assessments of family environment such as observational techniques with established, psychometrically sound processes for coding observations of family interactions.

Contrary to our hypotheses, paternal psychological distress and family rigidity (i.e., control) did not associate with pain-related functional disability. There are several possible explanations for the lack of findings related to paternal psychological distress. First, we successfully collected paternal report on only 60 of our 78 respondents. There may have been some sampling bias for fathers who were willing and able to participate in the study. (Unfortunately, we lack information on fathers who refused participation that would allow further exploration of this question.) It is also possible that because mothers assume a primary caretaking role in many families, maternal symptoms of distress might have a more direct relationship with the child’s daily functioning in comparison with paternal distress.

Regarding family rigidity, this has been observed as a dynamic in families of adult pain patients (Kopp et al., 1995) but may be less characteristic of families of children and teens with pain. Family disorganization predicted functional disability in the 2-predictor model, but after controlling for the influences of other variables, organization was not uniquely significant. In our data, it appears that disorganization loses predictive power in the context of overlapping and possibly more central family environment factors, such as conflict and enmeshment.

Our hypothesis that parent distress and family environment would moderate the relationship between pain and disability was not supported in the full sample. Interestingly, exploratory analyses revealed unexpected differences between the two pain diagnostic groups regarding the moderating effects of family environment. The data indicate that in the migraine group only, pain was less predictive of the extent of functional disability among children from family environments that reflect population norms, compared with children from more disruptive family environments. We conclude that pain is more likely to be associated with disability when impaired family dynamics exert an influence. For the migraine subsample, our findings were parallel to those of Claar and colleagues (Claar et al., 1999), who showed that pain and disability were more closely related in children with low levels of perceived competence. Along with perceived self-competence, an adaptive family environment may act as a protective factor that can buffer children against pain-related disability syndrome. We remain circumspect about the difference between diagnostic groups, as it may arise from the discrepancy in sample size or other random factors not accounted for in our analyses. Instead, we emphasize that the pattern of relationships among the family, pain, and disability variables is similar across the two diagnostic groups, despite statistical significance is attained only in the migraine group.

This study tested a moderational model. Future studies should also explore potential mediators of the pain–disability relationship. Both parental and child psychological adjustment, especially symptoms of depression and anxiety, would be important constructs for inclusion in mediational models of pain and functional disability.

The findings of this study must be considered in the context of its limitations. Foremost, our study is limited by a relatively small sample size that spans a broad age range. Given the cross-sectional nature of our data, we cannot draw conclusions about causal relationships. We do not know, for example, whether current patterns of parental distress and family functioning existed before the emergence of the child’s pain syndrome or arose in response to living with pain in the family. Reports of family environment were limited to maternal response; a multiple-informants approach would have provided a more complete picture of the family environment and reduced the potential for bias inherent in the single respondent’s view. In future studies, it would be useful to incorporate maternal, paternal, and child reports of family environment. Another measurement issue arose with the FDI. This measure includes items addressing general physical and psychosocial areas of functioning. It does not allow for the examination of domain-specific disability (e.g., school/academic impairment). Furthermore, we did not incorporate the parent-report version of this instrument in our study. More extensive and/or multimethod assessments of pain-related disability would advance this line of inquiry further and more clearly illuminate potential targets for intervention. The wording of our prompts for pain ratings introduced some possible contamination of pain with functional disability that cannot be corrected retrospectively. A notable strength of this study was the inclusion of fathers’ report on some measures. This is an important direction for future studies to take in continued investigations of connections between parent and family characteristics and the outcomes of pediatric pain syndromes.

Possible selection bias issues may limit the generalizability of our sample. Parents in our sample had above-average levels of education; thus the findings may not apply equally across the full spectrum of parental education or socioeconomic status (SES) levels. Additionally, our sample was recruited from medical clinics. Past studies (Claar et al., 1999) have suggested that individuals with pain who are present in medical clinics may have higher levels of functional disability than individuals who do not seek treatment. Finally, participants in our study had agreed to take part in a larger intervention study and may differ from children and families who would not choose to seek behavioral interventions for their pain symptoms.

This study highlights the central role that families play in determining how children manage recurrent pain. The results suggest that children with psychologically healthy parents and families that are not characterized by high levels of enmeshment or conflict may be more likely to function adaptively with recurrent pain. They also underscore the potential benefits of incorporating family-system–based techniques into psychological interventions for children and adolescents with pain problems. Helping families understand and, when needed, alter the ways in which their relational patterns may influence the child’s pain experience can be a crucial contribution to comprehensive pediatric pain management and rehabilitation.

Our primary goal in conducting the study was to contribute toward a comprehensive understanding of the wide variation in functional abilities of children with recurrent pain. This study was limited to parent and family characteristics, but we hope that in future studies this line of research can converge with those examining other potential influences on the pain–disability link, such as coping skills, perceived competence, and reinforcement and modeling effects. This could lead to the development of inclusive models of individual- and systems-level influences on functional outcomes among children with recurrent pain. Such models can then be used to inform and test intervention programs to maximize functional abilities in children and adolescents with chronic and recurrent pain syndromes.

The authors gratefully thank Paul Robins, PhD, for discussion of the ideas behind this work, and Neil Taylor, BA, for valuable research assistance. The study is based on work conducted by the second author with funding from the National Institutes of Child Health and Development (HD38647).

References

Author notes

¹Department of Psychiatry, Children’s Hospital Boston, and ²Harvard University Medical School