Functional constipation in infancy and early childhood: epidemiology, risk factors, and healthcare consultation

Childhood functional constipation (FC) is a significant health problem [1]. Even in young children the disease negatively affects health related quality of life and leads to considerable healthcare costs [2‐4]. It was noted that children under the age of one year had the highest rate of emergency department visits for constipation and related symptoms in the USA and the costs of care had risen by 121% from 2006 to 2011 [5].

Childhood constipation often has its roots in infancy and early childhood. A retrospective chart review of children with constipation revealed that the median age of onset was 2.3 years, with the 25th percentile to 75th percentile range being 0.8 to 4.8 years respectively [6]. An Italian birth cohort study has shown that the prevalence/onset of constipation at 3, 6, and 12 months was 11.6, 13.7 and 10.7%, respectively [7]. A representative community study in the USA noted that 4.7% of infants and 9.4% of toddlers were suffering from functional constipation (FC) [3].

During infancy, the transition from breastfeeding to formula feeding or the introduction of solid foods, is sometimes a trigger for the onset of FC [8]. Furthermore, an association has been suggested between cow’s milk protein allergy and FC. An improvement after a cow’s milk–free diet in young children with FC has been described ranging from 28 to 78% [9]. The latter is supported by evidence of histological changes in the mucosa of the colon indicating inflammation [9].

Poor toilet training in the toddler period is another important risk factor for the development of constipation [10]. When not properly toilet trained, these young children often exhibit stool withholding behavior, which leads to a vicious cycle of stool withholding, pain while passing stools and infrequent bowel motions. Therefore, studying risk factors related to constipation in early childhood may reveal important clues for its etiology and perhaps be helpful in formulating preventive strategies.

Psychological stress is a well-known risk factor to develop FC in older children. Several studies have shown an association between family and school related psychological stressors and constipation in older children [11]. In addition, teenagers who experienced any form of major child maltreatment are also known to develop FC [12]. However, the role of psychological stress and exposure to child maltreatment in developing FC in younger children is not known.

FC is a significant problem in the developing countries as well. A study from Sri Lanka noted that 15.4% of school children had FC [13]. However, there is a dearth of data from developing countries, particularly in Asia, on constipation in infancy and early childhood. Therefore, this study was undertaken with the objectives of a) studying the prevalence of FC in infancy and early childhood, b) identifying risk factors for developing constipation in early life, and c) studying the healthcare consultation patterns of infants and toddlers with FC.

This cross-sectional study provides the first epidemiological data about the prevalence of FC amongst infants and toddlers in a developing country. The prevalence rate of FC in infants (8.3%) was higher than that of toddlers (6.6%). Children aged between 37 and 48 months showed the highest prevalence (13%) of FC. Living in an urban area, being underweight and being subjected to physical violence were significantly associated with FC. The majority of children with FC were treated with dietary manipulations and non-pharmacological interventions, while almost 40 % of the children received laxatives.

FC is a common problem in childhood across the world. A similar prevalence, compared to our results, of 8.5% was found in a cross-sectional Korean study. Although, their included age category differed from our study (25–84 months vs 6.5–48 months) [21]. The most recent systematic review on the epidemiology of FC in children has shown that 0.7 to 28.8% are suffering from FC [1]. However that review did not report separate prevalence rates for young children. Recently, van Tilburg et al. reported a prevalence of FC of 4.7 and 9.4% in respectively for infants and toddlers living in the USA [3]. A similar study from Thailand including children of 4 months to 5 years of age found a much lower prevalence (1.6%) [22]. Higher prevalence rates in children aged between 3 and 5 years are reported Hong Kong (28.8%) [23] and The Netherlands (12%) [24].

In this cohort of Sri Lankan infants there was a higher prevalence rate of constipation (8.3%) compared to infants from the USA (4.7%) and those from Thailand (1.6%) [3, 22]. The peak age of developing constipation in Sri Lankan children was 3–4 years while in Thailand this was 2–3 years. The highest prevalence of FC in children aged between 3 and 5 years was reported in Hong Kong (28.8%) [23]. Our prevalence rate of 6.9% around 2–3 years is lower than the data from Thailand (7.2%), USA (9.4%) and The Netherlands (12%). All studies (except the study from Hong Kong which used Rome II criteria) have used the Rome III criteria for the diagnosis of constipation and collected data from parents of young children. The wide variation in prevalence could be attributed to variation in dietary patterns, cultural differences in toilet training, differences in child rearing, and perhaps other social determinants unknown to us. International collaborative studies using the same methods and defined age groups are needed to generate a clearer global picture of the epidemiology of FC in young children.

In this study we found a large percentage of young children with constipation with a history of painful or hard bowel motions (94%) and a history of large diameter stools (66%). This is in accordance with a previous study conducted in Sri Lanka in older children in which 71% of children had a history of painful or hard bowel motions and 66% had large diameter stools [13]. Studying a large cohort of children similar to our sample, Loening-Baucke reported painful defecation as the most frequently reported symptom (67%) [25]. Another study reported painful defecation and hard stools in 43 and 92% infants and young children respectively [26]. Surprisingly, none of the children in this study suffered of fecal incontinence. Generally, faecal incontinence is a feature of severe constipation. Lack of faecal incontinence suggests that most of these children may have suffered from mild constipation.

We studied a number of potential socio-demographic risk factors that could be associated with FC. Living in an urban area of the district was the only significant factor associated with FC in this cohort of Sri Lankan children. This finding is in accordance with our previous findings, where older children living in urban areas of Sri Lanka had a higher tendency to develop FC than their rural counterparts [13]. Ludvigsson has made a similar observation in children living in Sweden [27]. Although previous studies in older children have noted an association between gender and development of constipation [13], our data did not support this. It has been suggested that the differential prevalence of FGDs are related to differences in sex hormones in adults [28]. The fact that these hormonal profiles are not well established in infants and young children would explain the lack of difference between girls and boys.

Associations have been described between constipation and other FGDs in children and obesity and being overweight [29]. A hospital-based study on children with morbid obesity reported a delay in colonic transit confirming constipation in these children [30]. Moreover, in young children attending daycare centers in Korea, constipation was significantly associated with 2 h or less of outdoor play activities per day, and three or fewer servings of vegetables and fruits per day [21].

Contrary to these findings, our data showed that overweight or obesity, was not a risk factor for FC. A recent study on school children in Colombia confirms our findings [31]. However, this may be due to the small number of children with obesity/overweight in our sample. For the first time, we noted that children with underweight have a higher tendency to develop FC. It is perhaps necessary to explore the possibility of abnormalities in transit and anorectal function in underweight children with constipation as well.

In a previous study, we noted home-related stress and abuse to predispose children to develop FC [12]. We found that children subject to violence developed significantly more FC, but this correlation was not confirmed by logistic regression analysis. Studies among adults have also found that facing adverse life events as young children are a risk factor to develop IBS in adulthood [32]. Therefore, we hypothesized stressful life events could predispose young children to develop FC. However, contrary to our hypothesis, home related stresses were not associated with FC in infants and young children. Similar to our findings, studying young children (7–48 months) with constipation living in the West bank, Gaza strip and Jordan, Froon-Torenstra and co-workers noted that stressful life events had not contributed to the development of constipation [33]. Economic crises in the family and change of residence were also not associated with FC. Previous studies in adults and young children suggest that the brain-gut-axis plays an important role in developing FGDs after facing stresses and abuse [34]. Our finding of a lack of significant association between stress and abuse and development of FC in young children may be due to several reasons. First, the brain-gut-axis of young children may not be fully mature to appreciate the stresses, so that the alterations that lead to the development of FGDs are minimal. Second, a time lag may be necessary to develop FGDs after exposure to adverse life events. Over-emphasis of these events by adults with severe FGDs and possible recall bias in those retrospective studies are also possible reasons. In addition, we had to rely on mothers to collect information regarding home related stresses. Whether these factors truly lead to stress that can alter the brain-gut axis, leading to the development of constipation in young children, need further studies, including long-term follow-up of children faced with adverse life events and ill-treatment.

The majority of infants and young children with FC have received treatment for their symptoms. However, only 24% of young children were seen by a medical doctor. These consultation rates are better than a previous study in Sri Lanka on older children, which reported that only 3% of the affected children and adolescents sought medical advice for their symptoms [35]. It is possible that parents of young children are more conscious of abnormal bowel habits of their children than parents of older children who are toilet trained and independent. Adolescents are also reluctant to discuss their bowel habits with their parents. This is a possible reason for the lower healthcare consultation in the previous study.

We were able to categorize the treatment modalities into 3 main groups. They include: dietary advice, non-pharmacological modification and laxatives. Recent guidelines have clearly indicated the lack of therapeutic efficacy of increasing dietary fiber and consumption of water in the management of FC in children [36, 37]. Despite these facts, nearly half of the children received dietary interventions as primary treatment. Poor toilet training is recognized as a potential risk factor for FC especially in young children [10]. However, only 32% of children in this study received advice on toilet training. Oral laxatives are the currently recommended first line treatment for FC by both National Institute of Clinical Excellence (NICE), UK and combined European and North American Societies of Pediatric Gastroenterology, Hepatology and Nutrition Guidelines. Surprisingly, only one third has received oral laxative as therapy. Therefore, it is imperative to recognize that a nationwide educational program is essential to educate medical practitioners as well as the lay public about effective treatment options and possible consequences of poorly treated constipation.

There are several strengths in our study. We used a large sample of infants and young children in this study. Therefore, our findings could be generalized to the entire country. Since our study was based on clinics conducted by trained medical personnel, organic causes could be excluded by reviewing the child health development records. We used widely accepted Rome III criteria to define FC in infants and young children.

However, there are limitations to our study as well. There was no suitable questionnaire to assess symptoms of FGDs in infants and young children at the time we conducted the study. Therefore, we created the questionnaire by using the bowel habits questions from the QPGS questionnaire of children and adolescents. The QPGS questionnaire is a widely used and accepted way of assessing bowel habits of older children and its validated version is available in native language of Sri Lanka (Sinhala). Our group has used it in previous studies successfully [38‐40]. We relied on mothers to obtain details of bowel habits of their children and used a self-administered questionnaire, which could have led to exaggeration of symptoms, especially when mothers are suffering from FGDs themselves. In accordance with other studies reporting on prevalence of FGDs in young and older children a rectal examination was not performed, mainly because of ethical reasons [3, 13]. This might underestimate the prevalence of FC in children.

Our study has several implications. Firstly, the data show that FC is an important health problem in younger children as well. The lack of association of stressful life events, which are currently considered as important risk factors, and FC in this age group, compared to the older children possibly indicates that there is potentially a lag period between exposure to stress and development of FC. This window period could be used to manipulate the brain-gut-axis to reduce the risk of developing FGDs in children with early interventions. However, this concept needs further exploration. We also found that at least one third of children with FC are treated with ineffective therapeutic modalities. This demands the institution of a sustainable awareness program to educate medical practitioners as well as the general public.

FC is a common clinical problem in infants and young children in Sri Lanka. Living in urban areas and being underweight for the age are significantly and independently associated with FC. Contrary to previous findings in older children, stressful life events do not significantly predispose young children to develop FC. One quarter of children received treatment by a medical doctor for their symptoms and a large proportion of infants and young children received ineffective therapeutic interventions according to currently accepted guidelines.