Functional bowel disorders comprise a large range of gastrointestinal symptoms such as irritable bowel syndrome (IBS), functional constipation, and abdominal pain. These symptoms are frequently seen in Western countries (1). The etiology of these disorders is multifactorial (1) and is a challenge for health care professionals.
Various studies have suggested that functional bowel disorders underlie a complex interaction between psychosocial and physiological factors through the hypothalamic–pituitary–adrenal (HPA) axis (1). The HPA axis regulates the synthesis and secretion of glucocorticoids, which helps to control the metabolism of energy substrates (2). The most important glucocorticoid in humans is cortisol, which is secreted by the adrenal cortex in response to adrenocorticotrophic hormone, which is itself released by the hypothalamus as an effect of the corticotrophic-releasing hormone (CRH) (3). Studies show that psychological stressors activate the HPA axis (3). This can have a direct effect on the motor function of the gastrointestinal tract (4,5). Also, chronic gastrointestinal pain can further enhance activation of the HPA axis, leading to a vicious cycle that may explain the persistence of the symptoms (5,6). Although some studies have indeed shown elevated CRH levels and cortisol response in adults with IBS (7–9), others have claimed the opposite or provided evidence that cortisol responses are blunted in adults with IBS (10,11).
Psychological stressors may affect individuals differently, however, and because the HPA axis is still developing during childhood (6), results on IBS and stress in adults cannot be extrapolated to the pediatric population with functional bowel disorders. Because studies with respect to HPA axis activity and functional bowel disorders in children are extremely scarce, we tested whether infants with functional constipation and abdominal pain have an abnormal profile of the HPA axis after awakening, throughout the day, and in response to a mental stressor.
PATIENTS AND METHODS
Participants and Study Design
This study was embedded in the Generation R Study, a prospective cohort study from early fetal life onward that has been described in detail previously (12,13). An ethnically homogeneous subgroup of Dutch infants was randomly selected from the total cohort to prevent possible confounding or effect modification by ethnicity. Infants were born between February 2003 and August 2005 and 1108 parents gave consent for postnatal follow-up of their child. The study was approved by the medical ethical review committee at Erasmus University Medical Centre, Rotterdam, the Netherlands.
Collection of Salivary Cortisol Samples
At the age of 14 months, parents visited the Generation R Research Centre. Before this visit parents were asked to collect 5 saliva samples (Salivette Sampling Devices, Sarstedt, Rommelsdorf, Germany) from their infant and to note the sampling times during a normal routine weekday at home: immediately after awakening (mean 07:50 AM; standard deviation [SD] 56 minutes), 30 minutes later (mean 08:25 AM; SD 56 minutes), between 11 AM and 12 PM (mean 11:52 AM; SD 32 minutes), between 3 and 4 PM (mean 15:49 PM; SD 39 minutes), and at bedtime (mean 19:33 PM; SD 57 minutes). Parents received detailed written instructions with pictures concerning the saliva sampling and were asked to keep the samples stored in a freezer until they visited the research center.
To assess how the infant copes with stress, the Strange Situation Procedure (SSP) was used during the visit in the Generation R Research Centre. This is a validated procedure described in detail by Ainsworth and Bell (14). Briefly, the procedure consisted of 7 episodes of 3 minutes each and was designed to evoke mild stress in the infant exposed to the unfamiliar laboratory environment, a female stranger entering the room and engaging with the infant, and the parent leaving the room twice (14). The SSP took place for all of the participants between 8:40 AM and 15:41 PM weekdays (mean 11:31 AM; SD 2 hours). The saliva samples were collected by a research assistant at 3 time points: before, directly after the SSP, and 15 minutes later. The infants were not supposed to eat or drink 30 minutes before sampling.
Missing data after the SSP were the result of technical or procedural problems. Reasons of nonresponse were lack of time and failure to obtain saliva samples because the infant was not familiar with pacifiers.
Samples were centrifuged and stored at −80°C and were sent on dry ice in a single delivery to the laboratory of the Department of Biological Psychology at the Technical University of Dresden. Subsequently, the cortisol levels were assessed by using a commercial immunoassay with chemiluminescence detection (CLIA; IBL, Hamburg, Germany). Intra- and interassay coefficients of variation were <7% and 9%.
To assess the cortisol stress reactivity, a delta was calculated between the last sample (15 minutes post-SSP) and the first sample (pre-SSP). The second assessment, just after the SSP, was not used because it was too close to the onset of stress. We adjusted for baseline cortisol values to take into account the law of the initial values, which indicates that the direction of physiological response depends to a large degree on the initial levels as a result of variance change or regression to the mean (15).
To assess the total cortisol secretion during the day and to take account of the differences between separate cortisol measurements within each child and the time of the measures from baseline, the area under the curve (AUC) was estimated by calculating the curve of the cortisol measurement in nanomoles per liter on the y-axis and the time between the measurements on the x-axis. To adjust for differences in the duration of the day of measurement, the AUC was divided by the number of hours between the first and the final saliva collections. This method has been described in detail by Pruessner et al (16) and Watamura et al (17), and has been used successfully in previous studies (18,19). The cortisol awakening response (CAR) was estimated as the difference between the cortisol concentrations at awakening and 30 minutes thereafter as described by Kunz-Ebrecht et al (20). As a measure of circadian cortisol decline, the slope was calculated by fitting a linear regression line for each child that predicted the cortisol values from time since awakening by using the first and last saliva samples and at least 1 other cortisol sample.
In the second year of life, each child's stool pattern was assessed by questionnaire. Functional constipation in the second year of life was defined according to symptoms of the Rome II criteria (21). To avoid the influence of metabolic disorders, infants were excluded in the analyses because of presence of a congenital heart condition, anemia in the past year, or growth retardation defined as height <−2 SD based on the Netherlands growth curves of infants of 12 to 24 months (22).
A binary definition was defined as the presence or absence of any abdominal pain in the previous 3 months in the second year of life. Additionally, the severity of abdominal pain was classified according to an adapted version of the Abdominal Pain Index as described previously by Walker et al (23). Parents were asked about the frequency of the pain episodes that was rated during the previous 3 months on a 5-point scale (ranging from 0 = not at all to 5-every day). The daily frequencies of the pain episodes were assessed on a 4-point scale (none (1), 1–2 times per day, 3–6 times per day, and throughout the day (4)). The duration of the pain episode was rated on a 4-point scale (a few minutes (1), about half an hour, a few hours, all day (4)). Finally, parents indicated the intensity of the abdominal pain on a 10-point scale (1 = no pain and 10 = the most pain possible). The 5 pain ratings were summed and considered as the index of abdominal pain.
Prenatal questionnaires completed by the mother included information on mother's educational level, parity, maternal body mass index (BMI), maternal smoking, and maternal alcohol consumption. Data on sex, birth weight, gestational age, and birth outcomes were available from obstetric records assessed in midwife practices and hospital registries (13). Breast-feeding duration was available from questionnaire data filled in when the child was 6 and 12 months old. The level of parental stress in the child's second year of life was assessed using the Nijmeegse Ouderlijke Stress Index–Kort (NOSIK) (24), the Dutch version of the Parenting Stress Index–Short Form, which has been shown to be reliable and valid (25). The NOSIK comprises 2 domains consisting of 25 items: parenting stress caused by parent factors and parenting stress caused by child factors. Only the items on the parent domain were available in this study (n = 15). Items were assessed on a 4-point scale, and the scores were summed and divided by the number of items that has been filled in. Higher scores indicate greater levels of parental stress.
From the 882 infants who participated in the Generation R Focus Study and visited our research center between June 2004 and November 2006, information on more than 1 home saliva samples was available in 483 infants. During the SSP procedure, 442 infants had more than 1 saliva samples available and were eligible for analysis. Nonresponse analysis showed that the prevalence of functional constipation and abdominal pain was not different between infants with and without cortisol measurement (8.3% vs 8.1% and 7.6% vs 7.6%, respectively). Mothers of infants with no cortisol measurements were slightly more often smokers during pregnancy (29% vs 18%) and slightly more often lower educated (3% vs 1% low education). Infants with no cortisol measurement were slightly more often girls (52% vs 44%) and were slightly more often breast-fed for longer than 6 months (46% vs 36%). No difference between infants with and without cortisol measurement was found on birth weight (3517 vs 3509 g), gestational age (40.1 vs 40.1 weeks), parity (61% vs 68% nulliparous), and parental stress score (0.26 vs 0.25). To prevent bias associated with attrition, missing data of the infants who had at least more than 1 saliva sample available (either from home sampling or during the SSP, n = 483) were multiple imputed (n = 5 imputations) on the basis of the correlation between each variable with missing values and the other patient characteristics as described previously by Sterne et al (26). To obtain the desired effect sizes and standard errors, data were analyzed in each dataset separately. Subsequently, the results of the 5 imputed analyses were pooled and are reported in this article.
Differences in characteristics between infants with and without functional constipation and abdominal pain were tested with the χ2 test for categorical variables and the Mann-Whitney U test for continuous variables.
To assess how diurnal cortisol rhythm and cortisol reactivity were associated with functional constipation and abdominal pain, logistic regression analyses were performed with functional constipation and abdominal pain as dependent variables. Linear regression analyses were performed with the Abdominal Pain Index as a dependent variable (normally distributed). Tests for linear trend were carried out fitting the indicators of cortisol diurnal rhythm and stress response as a continuous variable. To test for nonlinear trends, a quadratic term was added to the model that included the linear term. Because both the linear term and the quadratic term were not statistically significant (see the online-only supplementary table at http://links.lww.com/MPG/A48), analyses were performed after stratification of AUC, CAR, cortisol slope, and delta stress into tertiles.
Additional adjustment for potential confounders such as sex, maternal educational background, parity, maternal smoking, maternal alcohol consumption, maternal BMI, birth weight, gestational age, breast-feeding duration, and parental stress were on the basis of literature followed by the change in effect size (ie, ≥10% change in regression coefficient). Effect modification by sex was evaluated by adding the product-term of cortisol variables and sex (eg, AUC*sex) as an independent variable to the model.
Results were reported as odds ratios (ORs) and 95% confidence interval (95% CI) for the analyses on abdominal pain and functional constipation and as regression coefficients (β) and 95% CI for the analyses on the Abdominal Pain Index. Complete case analyses and analyses after the multiple imputation procedure were performed. Similar results were found after the complete case analyses, but 95% CIs for the effect estimates were narrower after the multiple imputation procedure. Therefore, only the data after the multiple imputation procedure are presented. P < 0.05 was considered as statistically significant. Statistical analyses were carried out by using SPSS 17.0 for Windows (SPSS Inc, Chicago, IL).
Patient characteristics are presented in Table 1. Of 483 infants with cortisol data, 13% and 17% had functional constipation and abdominal pain, respectively. Four percent of the infants had symptoms of both functional constipation and abdominal pain. The mean (SD) index for abdominal pain was 6.61 (1.61). The mean (SD) age of cortisol sampling during the SSP and at home throughout the day was 14.5 (0.87) and 14.4 (1.07) months, respectively.
Mean (SD) cortisol levels at time of awakening was 15.80 (9.70) nmol/L in the total study group. Between 11 and 12 PM this decreased to 7.26 (6.45) nmol/L, declining further to 3.42 (5.80) nmol/L at bedtime.
Similar diurnal patterns of cortisol were found in infants with functional constipation and abdominal pain (Fig. 1). Compared to infants with no functional constipation or abdominal pain, no significant difference was found in AUC, CAR, and circadian cortisol decline (Table 2). There was no significant association found between indices of diurnal cortisol rhythm and the abdominal pain index (Table 3). Additional adjustment for age of cortisol sampling, sex, maternal educational background, parity, maternal smoking, maternal alcohol consumption, maternal BMI, birth weight, gestational age, breast-feeding duration, and parental stress did not change these results (data not shown).
Before the SSP, mean (SD) levels in the study group were 6.26 (5.21) nmol/L. Directly after the SSP, mean (SD) cortisol levels remained relatively stable at 6.19 (5.21) nmol/L but increased to 7.18 (6.29) nmol/L 15 minutes after the SSP. The increase in cortisol levels after the SSP was higher in infants with functional constipation and abdominal pain (Fig. 2), but this was not statistically significant after adjustment for baseline cortisol levels (Table 2). No statistically significant association was found between stress reactivity and the abdominal pain index (Table 3). Additional adjustment for age of cortisol sampling, sex, maternal educational background, parity, maternal smoking, maternal alcohol consumption, maternal BMI, birth weight, gestational age, breast-feeding duration, and parental stress did not alter the results (data not shown).
No statistical interaction with sex was found in the analyses between cortisol stress response, diurnal rhythm, and functional constipation and abdominal pain (data not shown).
In this study, we demonstrated that the diurnal rhythm of cortisol and stress reactivity is not significantly associated with either functional constipation or abdominal pain. The effect of the HPA axis in young infants with functional bowel disorders has been studied relatively little. Even the results reported in adults are conflicting. For example, although FitzGerald et al (27) recently showed that controls had a higher cortisol response after acute stress than women with IBS did, Chang et al (28) showed that patients with IBS had higher cortisol levels than controls but that there was no association with baseline cortisol levels. Similarly, an earlier study in children with recurrent abdominal pain found that cortisol levels were blunted relative to the levels in controls (29).
A recent study of young children with IBS demonstrated that cortisol stress reactivity is related more to adverse life events than to the presence of IBS in children (30), suggesting that the association between cortisol and functional bowel disorders in previous studies may be confounded by psychological status. Several studies have suggested that the prevalence of depression and anxiety disorders is higher in adults with functional bowel and chronic life stress can contribute to functional bowel disorders (31–34). Although the amount of stress is difficult to quantify in young children, Dorn et al (35) showed that scores on social stress found in children with anxiety scores were similar to those found in children with recurrent abdominal pain.
Because blunted and increased cortisol levels have both been shown in patients with functional bowel disorders, we expected to find differences in cortisol levels in infants with functional constipation and abdominal pain compared to those without these complaints; however, HPA response in people with functional bowel disorders may vary according to their psychological condition. For instance, patients with IBS without psychiatric comorbidity are more sensitive to stress than those with severe depression (36). Similarly, because the HPA axis is still developing early in life and has a high intraindividual instability (37), the influence of the HPA axis in early childhood constipation and abdominal can be difficult to explore. Intervention studies also suggest that colon motility increases, not decreases, after administration of CRH to subjects with IBS (38). If cortisol levels were inversely associated with functional constipation in a subset of our study group but elevated cortisol levels could also occur in infants who had these symptoms, then the association may be canceled out.
Other neurological pathways such as the autonomic nervous system also have been suggested to play a role in functional bowel disorders. Not only has increased activity of the autonomic nervous system been found in adults with IBS (39,40) but also differences in autonomic activity in response to stress have been found in children with and without chronic abdominal pain (41). Because the autonomic nervous system responds to stress much faster than the HPA axis (42), this may have a more prominent role in functional bowel disorders, but this needs further elucidation.
The strength of this study is that the study population was not selected on the basis of the medical care they had received. As a result of reverse causality, a selected population may increase bias because children with abdominal pain seeking medical care may already have elevated cortisol levels because of the constant pain or symptoms.
Despite this strength of the study, different methodological considerations must be taken into account when interpreting our results. First, we used criteria from Rome II (21) to define functional constipation, and we were not able to specify this outcome according to the most recent evidence-based Rome III criteria (43). As a result, our results preclude conclusions on the role of the HPA axis in more severe functional constipation according to the Rome III criteria.
Second, studies also have shown that cortisol reactivity to stress collapses with increasing age and it has been proposed that the difference in cortisol levels in response to stress becomes smaller as a child ages (42). Because we had only 17% cases with abdominal pain, the small difference in cortisol stress response that we were not able to detect as statistically significant may be influenced by the small sample size. On the contrary, differences in cortisol stress reactivity are in accordance with results from other studies in the same age group, but these studies were much smaller than our study group (44,45). Third, the cortisol stress response and its physical change are thought to be time limited (6). The time lag between cortisol sampling and the assessment of gastrointestinal symptoms in our study may, therefore, explain our results because the association between cortisol secretion and functional bowel disorders may be applicable only when it is measured in short succession. At last, differences between our study and results from other studies may be caused by different types of cortisol measurement (salivary, urinary, or total serum cortisol). Nevertheless, it is thought that only the free (unbound) forms of cortisol are biologically active, and we used salivary cortisol measurements, which correlate with free (unbound) serum cortisol levels in healthy subjects (46).
In conclusion, these data do not support the hypothesis that cortisol plays a significant role in functional constipation and abdominal pain in infants age 24 months. Further studies should clarify whether other branches of the brain–gut axis may be involved in these conditions and whether there is any influence of adverse life events.
We acknowledge the contributions of the children and parents, general practitioners, hospitals, and midwives in Rotterdam. We thank Maartje P.C.M. Luijk of the Department of Child and Adolescent Psychiatry, Erasmus University Medical Center, and the Center for Child and Family Studies, Leiden University, Leiden, for assistance with the use and interpretation of the cortisol data. We also thank Lidia R. Arends from the Department of Biostatistics and Institute of Psychology for support with the multiple imputation procedure.
1. Ammoury RF, Pfefferkorn Mdel R, Croffie JM. Functional gastrointestinal disorders: past and present. World J Pediatr
2. Nieuwenhuizen AG, Rutters F. The hypothalamic-pituitary-adrenal-axis in the regulation of energy balance. Physiol Behav
3. Papadimitriou A, Priftis KN. Regulation of the hypothalamic-pituitary-adrenal axis. Neuroimmunomodulation
4. Tache Y, Bonaz B. Corticotropin-releasing factor receptors and stress-related alterations of gut motor function. J Clin Invest
5. Jones MP, Dilley JB, Drossman D, et al. Brain-gut connections in functional GI disorders: anatomic and physiologic relationships. Neurogastroenterol Motil
6. Charmandari E, Kino T, Souvatzoglou E, et al. Pediatric stress: hormonal mediators and human development. Horm Res
7. Elsenbruch S, Orr WC. Diarrhea- and constipation-predominant IBS patients differ in postprandial autonomic and cortisol responses. Am J Gastroenterol
8. Heitkemper M, Jarrett M, Cain KC, et al. Autonomic nervous system function in women with irritable bowel syndrome. Dig Dis Sci
9. Patacchioli FR, Angelucci L, Dellerba G, et al. Actual stress, psychopathology and salivary cortisol levels in the irritable bowel syndrome (IBS). J Endocrinol Invest
10. Ehlert U, Nater UM, Bohmelt A. High and low unstimulated salivary cortisol levels correspond to different symptoms of functional gastrointestinal disorders. J Psychosom Res
11. Huang CW, Lui CC, Chang WN, et al. Elevated basal cortisol level predicts lower hippocampal volume and cognitive decline in Alzheimer's disease. J Clin Neurosci
12. Jaddoe VW, Bakker R, van Duijn CM, et al. The Generation R Study Biobank: a resource for epidemiological studies in children and their parents. Eur J Epidemiol
13. Jaddoe VW, van Duijn CM, van der Heijden AJ, et al. The Generation R Study: design and cohort update until the age of 4 years. Eur J Epidemiol
14. Ainsworth MD, Bell SM. Attachment, exploration, and separation: illustrated by the behavior of one-year-olds in a strange situation. Child Dev
15. Jamieson J. The law of initial values: five factors or two? Int J Psychophysiol
16. Pruessner JC, Kirschbaum C, Meinlschmid G, et al. Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology
17. Watamura SE, Donzella B, Kertes DA, et al. Developmental changes in baseline cortisol activity in early childhood: relations with napping and effortful control. Dev Psychobiol
18. Luijk MP, Saridjan N, Tharner A, et al. Attachment, depression, and cortisol: deviant patterns in insecure-resistant and disorganized infants. Dev Psychobiol
19. Saridjan NS, Huizink AC, Koetsier JA, et al. Do social disadvantage and early family adversity affect the diurnal cortisol rhythm in infants? The Generation R Study. Horm Behav
20. Kunz-Ebrecht SR, Kirschbaum C, Marmot M, et al. Differences in cortisol awakening response on work days and weekends in women and men from the Whitehall II cohort. Psychoneuroendocrinology
21. Rasquin-Weber A, Hyman PE, Cucchiara S, et al. Childhood functional gastrointestinal disorders. Gut
1999; 45 (suppl 2):II60–II68.
22. Fredriks AM, van Buuren S, Burgmeijer RJ, et al. Continuing positive secular growth change in The Netherlands 1955-1997. Pediatr Res
23. Walker LS, Smith CA, Garber J, et al. Development and validation of the pain response inventory for children. Psychol Assess
24. De Brock AJLL, Vermulst AA, Gerris JRM, et al. Nijmeegse Ouderlijke Stress Index (NOSI)—Manual
. Lisse, the Netherlands: Swets en Zeitlinger; 1992.
25. Haskett ME, Ahern LS, Ward CS, et al. Factor structure and validity of the parenting stress index-short form. J Clin Child Adolesc Psychol
26. Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ
27. FitzGerald LZ, Kehoe P, Sinha K. Hypothalamic-pituitary-adrenal axis dysregulation in women with irritable bowel syndrome in response to acute physical stress. West J Nurs Res
28. Chang L, Sundaresh S, Elliott J, et al. Dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis in irritable bowel syndrome. Neurogastroenterol Motil
29. Alfven G, de la Torre B, Uvnas-Moberg K. Depressed concentrations of oxytocin and cortisol in children with recurrent abdominal pain of non-organic origin. Acta Paediatr
30. Videlock EJ, Adeyemo M, Licudine A, et al. Childhood trauma is associated with hypothalamic-pituitary-adrenal axis responsiveness in irritable bowel syndrome. Gastroenterology
31. Camilleri M, McKinzie S, Busciglio I, et al. Prospective study of motor, sensory, psychologic, and autonomic functions in patients with irritable bowel syndrome. Clin Gastroenterol Hepatol
32. Talley NJ, Boyce P, Owen BK. Psychological distress and seasonal symptom changes in irritable bowel syndrome. Am J Gastroenterol
33. Dunlop SP, Jenkins D, Neal KR, et al. Relative importance of enterochromaffin cell hyperplasia, anxiety, and depression in postinfectious IBS. Gastroenterology
34. Posserud I, Agerforz P, Ekman R, et al. Altered visceral perceptual and neuroendocrine response in patients with irritable bowel syndrome during mental stress. Gut
35. Dorn LD, Campo JC, Thato S, et al. Psychological comorbidity and stress reactivity in children and adolescents with recurrent abdominal pain and anxiety disorders. J Am Acad Child Adolesc Psychiatry
36. Ahrens T, Deuschle M, Krumm B, et al. Pituitary-adrenal and sympathetic nervous system responses to stress in women remitted from recurrent major depression. Psychosom Med
37. Tollenaar MS, Jansen J, Beijers R, et al. Cortisol in the first year of life: normative values and intra-individual variability. Early Hum Dev
38. Fukudo S. Role of corticotropin-releasing hormone in irritable bowel syndrome and intestinal inflammation. J Gastroenterol
2007; 42 (suppl 17):48–51.
39. Mazur M, Furgala A, Jablonski K, et al. Dysfunction of the autonomic nervous system activity is responsible for gastric myoelectric disturbances in the irritable bowel syndrome patients. J Physiol Pharmacol
2007; 58 (suppl 3):131–139.
40. Chelimsky G, Czinn SJ. Techniques for the evaluation of dyspepsia in children. J Clin Gastroenterol
41. Puzanovova M, Arbogast PG, Smith CA, et al. Autonomic activity and somatic symptoms in response to success vs. failure on a cognitive task: a comparison of chronic abdominal pain patients and well children. J Psychosom Res
42. Jansen J, Beijers R, Riksen-Walraven M, et al. Cortisol reactivity in young infants. Psychoneuroendocrinology
43. Rasquin A, Di Lorenzo C, Forbes D, et al. Childhood functional gastrointestinal disorders: child/adolescent. Gastroenterology
44. Fortunato CK, Dribin AE, Granger DA, et al. Salivary alpha-amylase and cortisol in toddlers: differential relations to affective behavior. Dev Psychobiol
45. Davis EP, Granger DA. Developmental differences in infant salivary alpha-amylase and cortisol responses to stress. Psychoneuroendocrinology
46. Törnhage CJ. Salivary cortisol for assessment of hypothalamic-pituitary-adrenal axis function. Neuroimmunomodulation