Recurrent abdominal pain in children : the concept, aetiology, diagnostics and prognosis

Abstract

Background: Recurrent abdominal pain (RAP) due to an abdominal pain-related functional gastrointestinal disorder (ap-FGID) is one of the most common disorders of childhood and it accounts for much suffering and many healthcare efforts. Both school absence and parental work absence can be considerable when young children are affected. The aetiology of ap-FGID is regarded multifactorial, where the bidirectional gut-brain axis is thought to have a central role, involving gut microflora and psychosocial events. The biopsychosocial model seems superior to standard medical care when treating children with ap-FGID. The current terminology adheres to the symptom-based Rome criteria of different ap-FGIDs, most recently updated 2016 in Rome IV. There are scarce evaluations of the accuracy of the paediatric Rome diagnostic criteria and suggested alarm symptoms in diagnosing ap-FGID.

Methods: This thesis was founded on four studies, based on two population-based birth cohorts and one clinical study. In study I, 258 children 4-17 years old seeking care for gastrointestinal complaints were evaluated. Validated Rome III based patient-administered symptom questionnaires were filled in at the initial visit, blinded to the clinician. One year after inclusion a reference diagnose was set through review of electronic records. The sensitivity and specificity of the Rome III criteria for ap-FGID combined with selected alarm symptoms and/or laboratory tests were explored in four-field tables. Study II was based on a population-based birth cohort, BAMSE (n=4089). The association between antibiotic use and RAP at 12 years was explored. Information on early childhood antibiotic use during the first two years of life was based on parental questionnaires. Antibiotic use during the three years before gastrointestinal evaluation at 12 years of age was derived from the Swedish prescribed drug register. RAP was defined as monthly abdominal pain according to questionnaire self-report. Study III was performed in a subset of the lifestyle birth cohort ALADDIN (n=470). Children growing up in families of different lifestyles were explored regarding ap-FGID at five years of age. The three different, predefined family lifestyles were anthroposophic, partly anthroposophic and non-anthroposophic and classified through parental questionnaires. The Rome III-defined ap-FGIDs irritable bowel syndrome, functional dyspepsia and functional abdominal pain were assessed at five years through questionnaires and telephone interviews. Study IV was based on the BAMSE cohort. RAP was measured on three time points: in early childhood, containing pooled data from 12 and 24 months, at 12 years and at 16 years. In early childhood RAP was defined as parental report of abdominal pain during the previous 6 and 12 months respectively. Questionnaire data at 12 and 16 years were self-reported, and RAP defined as weekly abdominal pain. At 16 years the questionnaires also allowed for definitions according to Rome III for ap-FGID, irritable bowel syndrome, functional dyspepsia and functional abdominal pain. Binominal generalised linear model with log link was used to assess the association expressed as relative risk, between RAP in early childhood or at 12 years and RAP at 12 and 16 years, and any ap-FGID and irritable bowel syndrome at 16 years.

Results: In study I we found that Rome III criteria for ap-FGID in combination with absence of any alarm symptoms had high specificity (0.90) but very low sensitivity (0.15) in diagnosing children with ap-FGID. Alarm symptoms were equally common (81 % in total cohort) regardless of the cause of gastrointestinal complaints. In study II we found no evidence that antibiotic use was associated with monthly RAP at 12 years of age. The lifestyle study (III) revealed an increased risk of ap-FGID at five years in children growing up in families with a partly anthroposophic lifestyle, adjusted Odds Ratio 2.6 (95 % confidence interval: 1.1-5.9). No specific factor of the anthroposophic lifestyle could be identified as bearing this increased risk. The prognosis study (IV) showed that most children with RAP in early childhood were not affected at 12 years of age, adjusted Relative Risk (RR) 1.7 (95 % CI: 0.9-3.0). Most children with RAP at 12 years of age were not affected at 16 years but had an increased risk of RAP at 16 years, adjusted RR 2.1 (95 % CI: 1.7-2.7), and of ap-FGID at 16 years of age, adjusted RR 2.5 (95 % CI: 1.8-3.4). In the cohort 33 % of children had RAP at least once between one and 16 years of age.

Conclusion: The Rome III criteria combined with absent alarm symptoms classified very few children with ap-FGID. When criteria were fulfilled, and alarm symptoms negative in children with gastrointestinal complaints, the risk of an organic disease was low. Antibiotic use was not a major risk factor for RAP. Children in families with distinct differences in family lifestyle had a two-fold increased risk of ap-FGID at five years, but mechanisms are unclear. We speculate in differences in childhood coping development. The prognosis in RAP was benign, with most children growing out of their symptoms. The increased risk of persistent RAP between 12 and 16 years that was found, can be useful in guiding clinicians on how to follow up children with ap-FGID.