|Home | About | Journals | Submit | Contact Us | Français|
The current study examined the prevalence and presentation of psychosocial symptoms in pediatric patients and their parents presenting for first time appointments at a gastroenterology (GI) clinic compared to healthy controls (HC). One hundred GI patients, aged 8–17 years, and their mothers were compared to 100 age- and gender-matched HC on measures of child and parental behavioral/psychosocial functioning, depression, and anxiety. Results revealed significant correlations between mother- and child-reported internalizing behaviors and psychological symptoms. Significant group differences were observed in internalizing problems, adaptive and social skills, and leadership competency, as well as parental interpersonal sensitivity, depression, phobic anxiety, and number of psychological symptoms. GI patients are at increased risk for psychosocial and social dysfunction compared to healthy peers. Psychosocial factors should be considered when assessing patients in GI clinics. Patients might benefit from treatment plans that involve adjunctive behavioral intervention to assist patients in managing their conditions.
Psychosocial comorbidity plays a significant role in pediatric health care. While the majority of children with health conditions continue to function well, a subset of patients across conditions develop psychosocial comorbidity. This is particularly important because untreated or unrecognized mental illness can negatively impact clinicians’ medical treatment decisions and treatment outcomes (Cass, Volk, & Nease, 1999; Simon et al., 1998) and can increase long-term utilization of health care resources (Katon et al., 1990). In a recent study examining mental health problems in pediatric patients presenting to the emergency department for medical, nonpsychiatric complaints, 45% of patients met criteria for a mental disorder, 23% met criteria for more than one disorder, and 21% of patients’ mothers screened positive for mental illness (Grupp-Phelan et al., 2007). Moreover, 18% of mothers of pediatric patients presenting to both emergency and primary care settings have demonstrated a positive screen for major or sub-threshold depression. Importantly, 76% of those mothers who screened positive for major depression and 17% of those who screened negative reported that their psychological symptoms negatively affected their ability to care properly for their children (Grupp-Phelan, Whitaker, & Naish, 2003). Thus, psychological comorbidity appears to be highly prevalent among pediatric patients and parents seeking general health care, and this comorbidity can present significant barriers to care in patients and their parents. It is unknown, however, to what extent patients presenting at pediatric Gastrointestinal (GI) subspecialty care clinics demonstrate psychosocial comorbidity that might compromise care and use of health care resources.
GI disorders and complaints are relatively common among children and adolescents. For example, recurrent abdominal pain (RAP) affects approximately 30% of youth (McGrath, 1990), and irritable bowel syndrome (IBS) has been identified as one of the most significant pediatric chronic diseases, affecting approximately 10% of the general population, 6% of middle school children, and 14% of high school children (Hyams, Burke, Davis, Rzepski, & Andrulonis, 1996). Although pediatric GI disorders vary with respect to treatment course and duration, they share many common features, including symptoms like abdominal pain, functional limitations (e.g., missing school), and psychosocial distress such as anxiety, depression, trauma, and environmental stress (Ballenger et al., 2001; Dufton, Dunn, & Compas, 2009; Levy et al., 2006; Walker & Greene, 1989). Psychosocial comorbidity is also common in adults with GI disorders (Ballenger et al., 2001), suggesting a potentially long-term course of comorbid GI and psychosocial symptoms. In addition, parents of children with GI disorders have demonstrated increased psychosocial symptoms including depression in mothers of children with IBD (Burke et al., 1994) and depression and anxiety in mothers of children with RAP (Walker & Greene, 1989). Importantly, the presence of psychosocial dysfunction can have a significant effect on the outcome of GI disorders (Ballenger et al., 2001). Evidence suggests that psychosocial factors can significantly impact global well-being, health-related quality of life, and health care seeking (Levy et al., 2006) as well as symptom severity and course and the overall impact of the health condition on various aspects of child development (Walker & Jones, 2005).
A primary limitation in the extant literature, however, is that prior studies have focused on specific disease groups such as RAP or IBS. Although these studies provide important data for understanding psychosocial aspects of a few specific conditions, an understanding of psychosocial functioning in a broader group of GI patients would be particularly helpful for clinicians who may see patients with a variety of presenting complaints in their practice on a given day. Such data could be used by clinicians to identify patients at risk for increased morbidity and utilization of health care resources, and to understand the relative impact of psychosocial and physical factors when developing treatment plans.
In an effort to address this methodological gap in the current literature, the purpose of the present study was to (1) examine the prevalence and demographic/developmental presentation of psychosocial symptoms in pediatric patients and their parents presenting for first time appointments at a GI specialty clinic and (2) compare the expression of symptoms to a matched healthy control group. In contrast to prior research, this study assessed patients across GI conditions to provide data on the general GI clinic population as opposed to one specific subgroup of this population. Self-report and parent reports of child behavioral functioning, anxiety, depression, psychological trait characteristics, and parent psychopathology were obtained. It was hypothesized that patients with GI complaints and disorders would exhibit higher levels of anxiety, depression, and impairments in social functioning compared to healthy controls (HC), and that a positive correlation between child and parent reported psychopathology would be observed.
Participants were 100 children and adolescents, aged 8–17 years, and their mothers, who presented consecutively for a new (first) patient appointment to four physicians in the pediatric Gastroenterology, Hepatology, and Nutrition (GI) clinic at a large Midwestern children’s hospital. Children and adolescents with pervasive developmental disorders (e.g., autism, intellectual disability), acute conditions requiring hospitalization, advanced organ disease, or with a known comorbid chronic physical illness (e.g., Crohn’s Disease, diabetes) were excluded from the study. Comorbid chronic physical illnesses were defined as any previously diagnosed condition that required ongoing treatment of symptoms. Patients with comorbid GI conditions were excluded because they would have been seen and diagnosed previously; thus they would not be representative of new patient visits. Patients with other comorbid conditions were excluded in order to eliminate any reports of psychosocial symptoms that may be attributable to another chronic condition. The GI group was compared to 100 HC who were matched on gender and age (±1 year). HC were screened prior to participation and confirmed by parent report to have no history of developmental disorders, chronic physical illness, or gastrointestinal disorders and/or complaints.
In the GI group, new patients were identified following scheduling of a new patient visit and were sent a letter describing the study 1 month in advance of their appointments. Postcards were included for the family to return if they were not interested in participating. Ten percent of contacted families (i.e., ten families) returned the postcard resulting in 100 GI participants. Within 2 weeks following this mailing, families were contacted by study staff to discuss participation and obtain informed consent. Participants in the GI group completed the measures during their clinical appointments in the GI clinic. The HC group was recruited from a hospital maintained database of children and families who expressed interest in participating in research and from a hospital-wide email announcement for participants. No families declined to participate. Participants in the HC group completed measures at scheduled study visits at the hospital. Measures utilized in this study were selected on the basis of prior use in pediatric chronic illness populations and strong psychometric properties. Families were reimbursed $10 following participation. This study was approved by the Institutional Review Board at Cincinnati Children’s Hospital Medical Center.
The MASC (March, Parker, Sullivan, Stallings, & Conners, 1997) is a 39-item self-report inventory of anxiety. It yields a standardized Total Score and scores in five domains: Physical Symptoms, Harm Avoidance, Social Anxiety, Separation/Panic, and Anxiety Disorder Index. It has excellent 3 week (mean ICC score = .93) and 3 month (mean ICC score = .79) test–retest reliability, and has adequate convergent and discriminant validity. Internal consistency for this sample was .88.
The CDI (Kovacs, 1992) is a 27-item self- report inventory of depressive symptoms. In addition to a Total Score, it provides standardized scores for the following scales: Negative Mood, Interpersonal Problems, Ineffectiveness, Anhedonia (i.e., inability to experience pleasure), and Negative Self-Esteem. The CDI has been used extensively and has demonstrated good reliability (α ranging from .71 to .89) and validity. Internal consistency for this sample was .87.
The EPI-J (Eysenck & Eysenck, 1968) is a 57-item standardized measure of three psychological traits: Neuroticism (i.e., anxiety, emotional overresponsiveness), Extraversion (i.e., outgoing, uninhibited, impulsive, social tendencies), and Psychoticism (i.e., solitary, not caring for others, not fitting in). The EPI-J has excellent test–retest reliability over 1 year (r = .82–.88) and acceptable concurrent validity. Internal consistency for this sample was .66.
Parents of children provided demographic information (e.g., marital status, education, etc.) during the study visit.
The BASC-P (Reynolds & Kamphaus, 1992) is a widely used measure of child behavioral and emotional functioning. The BASC-P yields standardized scores for broad aspects of psychopathology (Externalizing Problems, Internalizing Problems, and Behavioral Symptoms Index) and adaptive functioning (Adaptive Skills), specific clinical problem areas (Hyperactivity, Aggression, Conduct Problems, Anxiety, Depression, Somatization, Atypicality [i.e., tendency to behave in ways that are considered odd or weird], Withdrawal, and Attention Problems), and social functioning (Adaptability, Social Skills, and Leadership). This measure has excellent internal consistency (α = .80 and .90 with adolescents), test–retest reliability (α = .80s–.90s over a 1-month period), and acceptable validity. Internal consistency for this sample was .88.
The SCL-90-R (Derogartis, 1994) is a multidimensional self-report inventory designed to screen for a broad range of psychosocial symptoms in adults. It yields standardized scores for global aspects of psychological symptomatology (Global Severity Index, Positive Symptom Total, and Positive Symptom Distress Index), and specific domains (Somatization, Obsessive–Compulsive, Interpersonal Sensitivity [i.e., discomfort in interpersonal situations and negative expectations for relationships], Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism [i.e., extremely withdrawn and isolated]). In addition, we determined caseness which was defined as having a Global Severity Index score ≥63 or two or more of the specific domain scales with a score of ≥63. The SCL-90-R is a widely used measure with adequate reliability (α = .78–.90) across scales. Internal consistency for this sample was .97.
Descriptive analyses were conducted for demographic variables across GI and HC groups, and for diagnosis/presenting complaint in the GI group. Independent samples t-tests and chi-square analyses were then conducted to examine demographic differences between GI and HC groups. Additionally, bivariate correlation analyses were conducted between total and composite scores for each of the measures and age to determine strength of relationships and correlations between child and parent reports. Summary scores for each of the measures were analyzed using three-way multivariate analysis of variance (MA-NOVA) tests, with group (GI versus HC), gender (male versus female), and age (Child [≤12 years, n = 115] versus Adolescent [≥13 years, n = 85]) as factors. Subsequently, a multivariate analysis of covariance (MAN-COVA) was conducted using summary scores on outcome measures to examine differences between GI and HC groups.
Table 1 presents demographic characteristics of the sample by group. Significant differences between groups were observed for marital status and both mothers’ and fathers’ years of schooling, in which the HC group had a higher percentage of married (and lower percentage of divorced) parents, and both parents had approximately 1 more year of schooling compared to the GI group. These variables were subsequently included as covariates in analyses concerning group comparisons. Table 2 illustrates the presenting complaints and medical diagnoses of the GI group as determined at the clinical evaluation and obtained through chart review. The modal diagnosis was RAP (44%), followed by constipation (18%), and IBS (10%).
Table 3 presents bivariate correlations between child’s age and total/composite scores for each of the measures for both the GI and HC groups. Significant correlations emerged in the GI group between age and the BASC-P Behavioral Symptoms Index (r = −.20, p < .05) and between age and the EPI-J Psychoticism scale (r = .23, p < .05). A significant correlation was also observed between age and the EPI-J Psychoticism scale (r = .22, p < .05) in the HC group. Significant correlations between reports of internalizing behaviors by children and mothers were also observed in the GI group. For example, maternal reports of internalizing problems (as measured by the BASC-P Internalizing scale) were positively correlated with children’s self-reported depression on the CDI Total (r = .55, p < .01), anxiety on the MASC Total (r = .35, p < .01), MASC ADI composite score (r = .25, p < .05), and EPI-J Neuroticism scale (r = .50, p < .01). In the HC group, only the BASC-P Internalizing scale and EPI-J Neuroticism scale were significantly correlated (r = .29, p < .01). In the GI group, statistical relationships were also observed in child-reported and mother-reported symptoms as evidenced by correlations between the SCL-90-R GSI score and CDI Total (r = .32, p < .01) MASC Total (r = .20, p < .05) and EPI-J Neuroticism scale (r = .30, p < .01). In the HC group, the SCL-90-R GSI score was correlated with the CDI Total (r = .33, p < .01) and EPI-J Neuroticism scale (r = .33, p < .01). Across child-reported measures in the GI group, anxiety traits (as measured by the EPI-J Neuroticism scale) were strongly correlated with anxiety symptoms on the MASC Total (r = .47, p < .01) and depression symptoms on the CDI Total (r = .70, p < .01). In the HC group, the EPI-J Neuroticism scale was strongly correlated with the MASC Total (r = .62, p < .01) and CDI Total (r = .68, p < .01).
Summary scores for all outcome measures were included in a three-way MANCOVA with group, gender, and age as factors. Results indicated statistically significant results for group (F = 5.65, p < .01), gender (F = 7.99, p < .01), and age (F = 2.72, p < .01), but not for interactions of these factors (all p’s > .05). Because there was no interaction between group and gender or age, these modifying variables were not subjected to further analyses. A MAN-COVA, with marital status and both mothers’ and fathers’ years of schooling entered as covariates, was subsequently conducted to examine differences between GI and HC groups on scale scores for each of the outcome measures.
Table 4 shows the means, standard deviations, and omnibus and univariate F values for group comparisons on each of the measures. No differences between the GI and HC groups were observed on child-report measures. On parent-report measures, elevated somatization was observed in the GI group compared to the HC group (F = 3.81, p < .01). GI patients also demonstrated increased symptoms of internalizing problems (F = 1.86, p < .05) and poorer adaptive skills (F = 2.19, p < .01), social skills (F = 2.01, p < .05), and leadership competency (F = 2.68, p < .01) compared to the HC group. With respect to parent psychopathology, mothers in the GI group reported higher levels of depression (F = 1.63, p < .05), phobic anxiety (F = 1.99, p < .05), interpersonal sensitivity (F = 1.86, p < .05), and positive symptom totals (F = 1.63, p < .05) compared to mothers in the HC group. In terms of caseness, 19% of mothers in the GI group met this criterion in contrast to 14% in the HC group (χ2(1) = 1.34, p > .05). Because approximately half of the sample was comprised of patients diagnosed with RAP or IBS, two disorders that have been shown to have increased rates of psychosocial difficulties, an additional MANOVA was conducted to examine whether patients with either RAP or IBS differed across scale scores from patients diagnosed with other disorders. Results of these analyses revealed no significant group differences on MASC, EPI-J, or CDI scales. However, patients with RAP or IBS demonstrated better adaptive skills (F = 2.35, p < .05) and social skills (F = 3.11, p < .01) than patients with other diagnoses. Parents of children with RAP or IBS demonstrated lower interpersonal sensitivity (F = 1.94, p < .05), higher depression (F = 1.96, p < .05), and lower psychoticism (F = 1.81, p < .05) scores compared to parents of patients with other diagnoses.
This study examined prevalence and demographic/developmental presentation of psychosocial symptoms in pediatric patients and their parents presenting as new patients to a GI clinic compared to matched HC. As hypothesized, there were correlations between GI patients’ and mothers’ reports of patient internalizing problems, including mother-reported overall internalizing symptoms and child-reported depression, anxiety, and neuroticism. There were also correlations between mother and GI patient symptoms in that global parent distress was correlated with depression, anxiety and neuroticism. Also consistent with the hypotheses, GI patients demonstrated higher levels of internalizing problems and greater impairment in social skills, leadership competency, and adaptive skills compared to HC per parent report. These findings suggest that GI patients are at increased risk for developing significant deficits in psychological and social functioning, community and school adaptation, and decision making/problem solving skills compared to their peers. These issues should be taken into consideration when assessing new patients in outpatient GI clinics. Moreover, patients might benefit from treatment plans that address both medical and psychosocial factors contributing to their illness. For example, treatments might involve adjunctive behavioral intervention, problem solving skills training, social skills intervention, and/or biofeedback to assist patients in managing their conditions. Such treatments have been shown to improve functioning in other pediatric populations (Vignolo et al., 2008; Wade, Walz, Carey, & Williams, 2009; Wysocki et al., 2008).
Additional findings that emerged from this study included a significant difference in parent psychological symptoms between the GI and HC groups. Parents of GI patients reported greater interpersonal sensitivity, depression, and phobic anxiety symptoms as well as greater overall number of symptoms compared to parents in the HC group. It is plausible that psychological symptoms in parents of pediatric GI patients are related to their children’s GI symptoms; however, the strength of this relationship and whether parent psychological symptoms are antecedent or a result of their children’s GI symptoms remains unclear. Taken together, the findings in this study suggest that both GI patients and their parents are at increased risk for psychosocial dysfunction. Although the causal nature of these symptoms remains unclear, comprehensive treatment for these patients should involve assessment of patient and family stress to determine factors that might contribute to the maintenance of GI symptoms and thus, prolonged functional morbidity. Specifically, it would likely be beneficial for GI providers to incorporate broad assessments of psychosocial functioning in patients and parents into their practice, receive professional consultation from a licensed psychologist regarding the appropriateness of referral based on these data, and have a streamlined system of referring patients to mental health care providers either within or outside the medical facility at which patients receive care for their GI condition.
Patients with RAP and IBS demonstrated marginally better functioning on measures of social and adaptive skills than patients with other diagnoses, and parents of children with RAP or IBS demonstrated better functioning on interpersonal sensitivity and psychoticism scales than parents of children with other diagnoses. The exception to this was depression, in which parents of children with RAP and IBS scored higher than parents of children with other diagnoses. These findings are contrary to prior research that indicates greater psychosocial distress in patients with RAP and their parents (Walker & Greene, 1989). However, the timing of assessment in this study may account for the results in that all patients and parents were seen early in the diagnostic and treatment process and may not have experienced symptoms long enough to develop commonly observed psychosocial sequelae.
Importantly, the significant findings observed in this study were based on parent-report assessment tools, and child-report measures did not yield significant findings. This might suggest that, given the findings of parental distress in this sample, parents perceive increased psychosocial difficulties in their children as a function of their own distress, resulting in an overestimation of symptoms. Alternatively, this may indicate that children presenting for new patient visits are at the beginning stages of developing psychosocial difficulties and are not recognizing the subtle, yet salient changes in behavior that is being identified by their parents. Thus, obtaining child-reported symptomatology would still be critical to understanding patient global functioning. Further research is needed to clarify this issue.
There are several notable strengths of this study. First, GI participants were sampled from a single site at a distinct time point in their medical care (i.e., first appointment), thereby avoiding the heterogeneity of sampling approaches that characterize much of the research in this area. Second, parallel data were collected from both children and mothers. Third, measures were obtained both for psychological/behavior problems and social functioning. Fourth, GI participants were matched with HC on age and gender. Fifth, the sample of 100 participants in each group was substantially larger than what is often found in research on pediatric populations. Sixth, measurement of anxiety consisted of both trait (i.e., relatively stable characteristic) and state (i.e., transitory emotional condition) assessments.
There were also several limitations that warrant caution in interpreting findings. First, psychopathology was measured as a dimensional construct, accounting for multiple facets of psychosocial functioning, and diagnoses of anxiety and depression were not obtained. Although GI participants and mothers reported elevated symptoms relative to controls in some areas, it is unclear if there are group differences on diagnoses of internalizing conditions. Second, information was obtained from mothers only and not fathers. Unfortunately, this shortcoming is common in pediatric research, and it is possible that the perspectives of fathers might have led to alternative conclusions. Third, the magnitude of some of the group differences was modest and the majority of T-scores were within the average range. However, given that these new patients were in the initial stages of diagnosis and treatment, these differences and T-scores are still meaningful as they might represent the early phases of the development of psychosocial dysfunction. Fourth, although the GI sample was homogeneous with respect to type of clinic and time of access, there was variability in GI diagnoses. Although this was planned in order to provide an assessment of general GI patients at new patient visits, it also highlights a common methodological issue in pediatric research, in which large numbers of participants that represent all parameters of illness expression are exceedingly difficult to obtain. Fifth, the socioeconomic status of this sample may be slightly higher than others, limiting generalizability of the findings. Finally it is unknown if the HC population is representative of the population of all children age 8–17 years since they were self selected as those willing to participate in research.
The present study should be viewed as an initial step toward identifying the comorbidity and effects of psychosocial dysfunction in children with GI complaints and disorders and their families. Future research should focus on examining rates of internalizing disorders diagnoses, particularly anxiety and depressive disorders, in the GI population compared to HC. Moreover, examination of the impact of somatization on internalizing symptoms will be important to our understanding of the nature of psychosocial dysfunction in this population. Additionally, studies that examine the role of social functioning and adaptive skill problems, including etiology and functional impact on treatment outcomes in GI disorders are needed to determine the utility of assessment and treatment of these factors. Further, longitudinal studies are needed to examine the long-term outcomes associated with increased psychosocial dysfunction in this population. Ideally, such studies will utilize large sample sizes across multiple sites, examine both patient and parent psychosocial factors, and stratify on GI complaint/diagnosis. Finally, research aimed at developing and testing biopsychosocial interventions will be needed to ensure that GI patients receive comprehensive treatment and optimize health outcomes.
This manuscript was supported by NIMH grant 53703. We extend our sincere appreciation for help with patient recruitment to the GI physicians: Michael Farrell, M.D., Mitchell Cohen, M.D., and Ajay Kaul, M.D., nurses: Susan Wagner, RN and Kim Dietrich, RN, and the staff of the Gastroenterology, Hepatology and Nutrition Clinic at Cincinnati Children’s Hospital Medical Center.
Kevin A. Hommel, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039, USA. University of Cincinnati College of Medicine, Cincinnati, OH, USA.
Kelly L. McGraw, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039. University of Cincinnati College of Medicine, Cincinnati, OH, USA.
Robert T. Ammerman, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039, USA. University of Cincinnati College of Medicine, Cincinnati, OH, USA.
James E. Heubi, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039, USA. University of Cincinnati College of Medicine, Cincinnati, OH, USA.
Molly Hansen, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039, USA.
Ellen Dunlap, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave.—MLC 7039, Cincinnati, OH 45229-3039, USA.
Deborah C. Beidel, University of Central Florida, Orlando, FL, USA.