Secondary Logo

Journal Logo

Original Articles: Gastroenterology

Socioenvironmental Factors Associated With Constipation in Hong Kong Children and Rome III Criteria

Tam, Yuk H.*; Li, Albert M.; So, Hung K.; Shit, Kam Y.*; Pang, Kristine K.*; Wong, Yuen S.*; Tsui, Siu Y.*; Mou, Jennifer W.*; Chan, Kin W.*; Lee, Kim H.*

Author Information
Journal of Pediatric Gastroenterology and Nutrition: July 2012 - Volume 55 - Issue 1 - p 56-61
doi: 10.1097/MPG.0b013e31824741ce
  • Free


Childhood constipation is a common gastrointestinal disorder worldwide and has become a major public health issue because of the heavy financial burden it incurs (1). Although constipation in children can be secondary to organic causes or sexual abuse, they account for <5% of cases, and the vast majority of the patients are considered to have functional constipation (2). For many years, research into the epidemiology of childhood constipation has been hampered by the lack of uniform criteria to define constipation (3,4). The recent development of the Rome III criteria for functional constipation has been an important milestone for research in childhood constipation because the Rome III criteria take into account the different aspects regarding constipation and are not as restrictive as the Rome II criteria (5).

The understanding of the pathophysiology of functional constipation in children is still primitive. Genetic predisposition (6,7), psychological distress (8,9), and inadequate fiber intake (7,10) have been proposed to be the causative factors. The most important mechanism that has been widely investigated and described in the pediatric population is stool withholding or volitional stool retention behavior (2,4,11). The former is usually described in toddlers with typical retentive posturing, whereas the term volitional stool retention is used more appropriately in older children (5). Such behavior features primarily voluntary postponement of defecation by children, leading to a vicious cycle that eventually develops to chronic constipation.

A systematic review found that only 50% of children with functional constipation were successfully treated and taken off laxatives after 12 months of intensive treatment in either general pediatric or pediatric gastroenterology settings (12). The development of functional constipation in a child is a process that may be triggered by the presence and interaction of multiple factors and end up in the final common pathway of volitional stool retention if the situation is allowed to perpetuate or progress. It is a reasonable belief that many chronic difficult cases managed in tertiary referral centers could have been prevented if factors that perpetuate volitional stool retention were corrected at an earlier stage. Such factors may be present in the environment surrounding a child's day to day living and are likely to vary with geographical and cultural differences. Identifying such factors helps to educate the public and provide guidance for preventive counseling at the primary care level to reduce the number of chronic and severe cases that inevitably require specialist or tertiary-level care.

We hypothesized in the present study that socioenvironmental factors in the family, school, and the lifestyle of a child had bearing on the prevalence of constipation among Hong Kong Chinese children through their subtle influence in encouraging postponement of defecation and hindering the development of regular bowel habits. We aimed to investigate such factors in our society and to estimate the prevalence of constipation among Hong Kong Chinese children in the present epidemiological study, which to the best of our knowledge was the largest-scale study ever conducted in Chinese children using pediatric Rome criteria.


Sample Size Estimation and Study Sample Selection

This was a cross-sectional questionnaire survey conducted in May and June 2011 in Hong Kong elementary schools serving local Chinese students. We assumed that the prevalence rate of constipation in our elementary school children was 12%, the median prevalence in the latest systematic review (13), with an accepted error of ±1.5%. According to the latest figure provided by the government, there are approximately 310,000 elementary school students in our community. The sample size required for the study would be approximately 1900. Giving an expected response rate of 65% to 75%, a target of 3000 children was calculated.

Hong Kong is composed of 3 major areas, namely Hong Kong Island, Kowloon, and New Territories. Government figures show the New Territories have >50% of the total population. Based on the population ratio, elementary schools serving local Chinese children were randomly selected from the 3 areas. The school management was contacted and invited to participate in the study. Finally, 10 schools were selected and agreed to participate, giving a potential study population of 3215 children. The geographical distribution of the 10 schools was 6 in New Territories, 2 in Kowloon, and 2 in Hong Kong Island.

Questionnaire and Data Collection

An informational letter was sent to the parents or legal guardians of each potential study subject to explain the study and invite them to participate by filling out an anonymous questionnaire. The completed questionnaire was sealed in an accompanying envelope and returned to schools within 2 weeks for collection.

The 2-sheet questionnaire contained 39 items and was written in the Chinese language, which should be easily comprehended by an adult with an elementary to a middle school education level. An adult, either a parent or the legal guardian if neither parent were living with the child at the time of the study, was asked to respond to the questions with the assistance of the child after consenting to participate. The questionnaire covered 4 main areas: sociodemographic information of the child and family, school toilet environment, lifestyle of the child, and questions related to bowel movements and abdominal pain.

Definition of Constipation

The questions related to bowel movements and abdominal pains were developed from the parent report form of QPGS-Rome III version (14). Participants were defined to have constipation in the present study if they fulfilled the Rome III diagnostic criteria by having 2 or more of the 6 conditions for at least 2 months while not meeting the criteria for irritable bowel syndrome: ≤2 defecations per week; ≥1 episodes of fecal incontinence per week; retentive posturing or excessive volitional stool retention; history of painful or hard bowel movements; passage of large stools that clog the toilet; and detection of large fecal mass in the rectum by doctors or nurses (5).

Specially-designed Scoring Systems

Participants were asked to rate the cleanliness and the facilities of the school toilets. Based on the participants’ responses to specific questions, scores of 1 to 5 for cleanliness and 0 to 10 for facilities were calculated for analysis in the study. The highest score would be given for the facilities if the participant thought there were always enough toilet chambers for bowel movements; toilet paper was always available; and toilet seats, door locks, and the flushing system always functioned properly.

A fiber-intake score and fast food–consumption score, both on a scale of 0 to 6 calculated according to the participants’ responses to specific questions, were designed for analysis as crude assessment of dietary fiber intake and frequency of fast food consumption. Fast food varieties included Western type (fried chicken/hamburgers) and traditional local street food (a variety of high-energy and fried food). The higher the fast food-consumption score, the more frequent was the visit to fast food outlets. Questions related to fiber intake asked about daily consumption of fruits and vegetables. Based on the previous findings in our Hong Kong children that the mean daily fiber intake was only 45% of that recommended in Western countries and combined intake of fruits and vegetables accounted for 60% of daily fiber intake, a fiber intake score of 0 and 6 in the present study represented an estimated intake of, respectively, 1 standard deviation below and 1 standard deviation above the mean daily fiber intake expected from children of our study age group (10,15).

Statistical Analysis

The prevalence of constipation in our study subjects was estimated with the 95% confidence interval. Constipated and nonconstipated children were compared in a univariate analysis using the χ2 and t tests for categorical and continuous data, respectively. Logistic regression was used in a multivariate analysis to identify independent risk factors associated with constipation and expressed in odds ratio with 95% confidence intervals. A P value of <0.05 was considered statistically significant. The study protocol was approved by the joint ethical committees of the university and the hospital.


From all of the returned questionnaires, 2318 (72% response rate) had sufficient entry to either define or exclude constipation and thus were qualified for analysis. The participants’ ages were from 6 to 15 years with a mean age of 9 ± 1.9 years. Ninety-eight percent of the participants were from 6 to 12 years old. Fifty-one percent of the participants were boys. Two hundred eighty-two children (12.2%, 95% confidence interval [CI] 10.9%–13.5%) were defined by Rome III criteria to have constipation. There was no difference in the prevalence rates between boys and girls (11.6% vs 12.3%; P > 0.05). The prevalence rates of the age groups 6 to 7, 8 to 9, 10 to 11, and 12 to 15 years were 16.8% (95% CI 13.8%–19.8%), 13% (95% CI 10.6%–15.5%), 8.5% (95% CI 6.4%–10.6%), and 9.8% (95% CI 6%–13.7%), respectively (P < 0.01) (Fig. 1).

Prevalence rates of constipation according to age groups.

Body mass index (BMI) z score was calculated for each participant according to the age, sex, and self-reported body weight and body height. Participants with BMI z scores at or above the 95th percentile of our Hong Kong Chinese children were defined in the study to be obese (16,17). A total of 16.7% and 16.1% of constipated and nonconstipated study subjects were found to be obese in the study. There was no difference in the prevalence of constipation between the obese and nonobese children (11.5% vs 11.1%; P > 0.05).

Constipation was found in 27.5% of children who lived with neither parent compared with 13.6% in those who lived with either one/both parents (P < 0.05). Children who ate dinner without the presence of parents >50% of the time experienced more constipation than others who had either one/both parents eat dinner with them at home ≥50% of the time (15.9% vs 11.4%; P < 0.05). No difference was detected between constipated and nonconstipated groups in the parents’ professions, number of siblings, and the average number of family members sharing 1 bathroom at home.

Overall, 36.3% of participants refused to pass their bowel movements at school toilets when they experienced the urge. Another 16.7% were neutral, whereas 47% expressed their willingness. Of those children who refused to use school toilets for bowel movements, 16.3% had constipation compared with 10.1% and 8.7% of those who were willing and who were neutral (P < 0.01). No difference was detected between constipated and nonconstipated children in their scoring for the cleanliness and the facilities of their school toilets.

Constipation was found to be more prevalent in children who slept <7 hours per night (17.4% vs 11.8%; P < 0.05), and who spent ≥2 hours per day finishing homework (14.6% vs 10.9%; P < 0.01) during regular school days. Constipated children had lower fiber-intake score (3.5 vs 3.9; P < 0.01; 95% CI mean difference 0.18%–0.61%), and higher fast food–consumption score (2.8 vs 2.4; P < 0.01; 95% CI mean difference 0.16%–0.57%) than nonconstipated children (Table 1).

Univariate analysis comparing the socioenvironmental characteristics of constipated and nonconstipated subjects

The overall characteristics noted among the constipated children in the present study included male sex (49.2%), age 6 to 9 years (70.5%), obesity (16.7%), living with neither parent (4%), eating dinner with parents <50% of the time (21.1%), refusal to have bowel movements in school toilets (48.9%), spending ≥2 hours per day doing homework (42.7%), and nighttime sleep duration <7 hours (9.7%).

Multivariate analysis after adjusting age and sex found eating dinners with one/both parents <50% of the time (OR 1.52; 95% CI 1.01%–2.31%), refusal to pass bowel movements in school toilets (OR 1.97; 95% CI 1.42%–2.74%), nighttime sleep <7 hours (OR 1.87; 95% CI 1.04%–3.33%), and frequent consumption of fast food (OR 1.14; 95% CI 1.03%–1.26%) to be independent risk factors for constipation in our study children (Table 2).

Independent risk factors for childhood constipation by multivariate analysis


The pediatric literature has found that the prevalence of constipation ranged from 0.7% to 29.6% (3,13). Diverse cultural, dietary, genetic, and environmental conditions and the lack of a uniform definition of constipation, particularly in early studies, have been attributed to the wide variance of prevalence rates. The development of the Rome criteria for pediatric functional gastrointestinal disorders and the evolution to the latest Rome III version have substantially facilitated medical research in childhood constipation worldwide and have laid the foundation for sound comparisons to be made between epidemiological studies in populations with racial and geographical differences. To date, data from studies using Rome III criteria to investigate pediatric constipation prevalence remain limited. Our finding of 12.2% of children having constipation is comparable with the prevalence rates of 10% and 15.4% reported in 2 studies conducted in the United States (18) and Sri Lanka (19), both of which used pediatric Rome III criteria to define constipation. Other investigators in China found that 24.9% of adolescents had constipation as defined by adult Rome III criteria (20). A systematic review found that Asian countries have a lower reported prevalence of constipation (median 10.8%) compared with Europe (median 19.2%) and North America (median 16%) in combined pediatric and adult populations (13). Although pediatric data from Asia remain limited, our findings, together with others, suggest that Asian children appear to be affected by constipation not to a lesser extent than their counterparts in Western societies (10,19,21,22).

We found the highest prevalence of constipation in our youngest participants aged 6 to 7 years. Others have reported a higher prevalence at 5 to 6 years of age compared with older schoolage children (23,24). An increased prevalence observed at the age of 2 to 3 years in preschool children has been attributed to the time of toilet training (25,26). The peak prevalence at around 6 years of age noted in schoolage children may simply reflect the beginning of the gradual decline in prevalence from the peak at the preschool age of 2 to 3 years (26). It is uncertain whether any significant life event such as advancing to the unaccustomed environment of elementary education has any bearing on the observed increase in constipation in children at approximately 6 years of age. Anxiety and stressful life events experienced at schools have been shown to be associated with constipation in adolescents (8).

We did not find any sex difference in the prevalence of constipation in our study children. Female predominance has been frequently noted in adults with constipation and a median female/male reported ratio of 1.5 has been found in a systematic review (13). There are conflicting data in the pediatric literature regarding the sex ratio in the occurrence of constipation. The predominance of boys in outcome studies, which included significant number of patients with fecal incontinence, may be explained by the higher prevalence of fecal incontinence among constipated boys (3). A similar prevalence for boys and girls (7,19,26,27) and higher prevalence for girls (10,22,28) has both been reported in previous pediatric epidemiological studies. To date, there is no convincing evidence to suggest any significant sex dominance in the prevalence of childhood constipation.

Case-control studies in tertiary settings have reported more constipation in obese children (29,30) or a higher prevalence of obesity in constipated children (31). Disordered eating pattern and less physical activity have been postulated to contribute to the association between obesity and constipation (32). We did not find any difference in the prevalence of constipation between obese and nonobese participants in the present study. The limitation of self-reported body weight and height has raised uncertainty about the accuracy of the calculation of BMI z scores of the participants. Differences in the study design and the study sample selection may also explain the difference in the results. Further population- or school-based studies with onsite measurement and symptoms assessment by investigators are needed to provide better answers about the association of obesity with childhood constipation in the general population.

The association of socioenvironmental factors with childhood constipation has not been widely investigated. Previous studies focused on the socioeconomic status of the family and the stressful events that happened in the family, the school, or the living surroundings (7–9,19,23). Stressful events such as family members having severe illness, parents’ separation, parents’ job loss, failure in school examination, separation from best friend, and living in an area affected by war have been reported to be associated with childhood constipation (8,9,19). Stress may modulate the brain–gut axis and affect gut motility. Adults of lower social, economic, and education levels were found to have more constipation (33). Previous pediatric studies, however, did not find any association between the socioeconomic status of family and constipation in children (3). Similarly, in our study we did not find any association between our constipated children with parents’ professions and size of family. The difference in the pathophysiology of constipation between adults and children may explain the difference in the risk factors. Because the major mechanism in the pathophysiology of childhood constipation is the vicious cycle resulting from volitional stool retention, we postulated that environmental factors that potentially encourage postponement of defecation may have a bearing on the development of constipation in children. Our finding that children living with neither parent and children eating dinners with parents <50% of the time had more constipation reflected the effect of complete absence or infrequent presence of parents in the family life of children. We believe that the inadequate company of parents at home may have a negative effect on the development of regular bowel habits in children. It is a common phenomenon in our society that many parents need to work long hours and children stay at home, most of the time under the care of domestic helpers. Parents may understandably become neglectful of their children's bowel habits when the time they can spend with their children in family is limited. Lack of parental prompting and support may encourage chronic postponement of defecation in a child who mentally is still developing and may not appreciate the importance of good bowel habits, particularly when other risk factors are present in the surrounding environment. Proper counseling in the primary care setting may identify childhood constipation at its early stage by raising the parents’ alertness to habitual postponement of defecation in their children and strengthen the parents’ role in helping children develop regular bowel habits.

In Hong Kong, the majority of elementary schools operate whole-day programs. Including the time for transport, elementary school students spend not fewer than 8 hours in schooling. This is equivalent to half, if not more, of the awake time of a child after excluding nighttime sleep. Our results identified refusal to use school toilets despite the urge for bowel movements to be an independent risk factor for constipation in our study children. A similar finding was reported by others (7). Because of the long hours spent in schools, we believe that refusal to pass bowel movements at school toilets may increase the risk of developing constipation by its effect of perpetuating volitional stool retention. We did not find any difference in the ratings of cleanliness and facility of school toilets between the nonconstipated and constipated children. A child's own sentiment about the use of school toilets for bowel movements appears to be more important than the cleanliness and facilities of school toilets in determining childhood constipation. Teachers’ attitude, child's personality, and peers’ reaction may contribute to how accepted a child feels about using school toilets for bowel movements. Public education to promote a positive attitude toward the use of school toilets may prove to be an effective tool to reduce the number of severe and chronic cases of constipation in children.

Similar to previous studies, we found inadequate fiber intake to be associated with constipation (10). The risk factor of frequent consumption of fast food for childhood constipation may better be explained by a certain dietary habit rather than a specific type of food, which increases the risk of having constipation. Children who habitually consume fast food may more likely be those who eat an unbalanced diet with high lipids and low fiber. A study conducted in Chinese adolescents also found frequent consumption of fried food to be associated with constipation (20).

Our findings suggested that a busy afterschool life with homework and inadequate sleep was associated with constipation. To the best of our knowledge, the present study is the first to investigate such an association. A heavy burden with homework after school may be a hindrance to the development of regular bowel habits in the evening at home and may encourage the postponement of defecation. Our results confirmed the social belief that Hong Kong children sleep less than their Western counterparts. The mean nighttime sleep durations for Western children ages 6 and 12 years are 10.9 ± 0.7 and 9.3 ± 0.7 hours (34). Overall, only 40% of our participants reported sleeping >8 hours per night. Nighttime sleep duration <7 hours was found to be an independent risk factor for constipation among our study subjects. Inadequate sleep, particularly if associated with heavy academic activities, can be a source of psychological distress to children. Fatigue as a result of inadequate sleep may also discourage children from using the toilet and favor defecation postponement. It is uncertain whether fatigue may modulate gut motility through its role in the brain–gut axis because fatigue has been reported to be associated with irritable bowel syndrome in adolescents (20). Further studies are needed to investigate the interaction of heavy homework or other academic-related activities, inadequate sleep, and the resultant emotions in the etiology of constipation in schoolage children.

Our study was limited by its cross-sectional nature, and the fact that the questionnaires were filled out without supervision of the investigators may create misinterpretation of the questions. Nevertheless, our finding that 12.2% of elementary school students had constipation still suggests that childhood constipation is prevalent in Asian societies. Socioenvironmental factors play an important role in determining the prevalence of childhood constipation by their influence in contributing to volitional stool retention behavior. Such risk factors likely vary with cultural and geographical differences. Increasing public alertness to these risk factors and correcting them may help prevent childhood constipation or stop its progression in earlier stages.


1. Liem O, Harman J, Benninga M, et al. Health utilization and cost impact of childhood constipation in the United States. J Pediatr 2009; 154:258–262.
2. Tabbers MM, Boluyt N, Berger MY, et al. Diagnosis and treatment of functional constipation. Eur J Pediatr 2011; 170:955–963.
3. van den Berg MM, Benninga MA, Di Lorenzo C. Epidemiology of childhood constipation: a systematic review. Am J Gastroenterol 2006; 101:2401–2409.
4. Benninga MA, Voskuijl WP, Taminiau JA. Childhood constipation: is there new light in the tunnel? J Pediatr Gastroenterol Nutr 2004; 39:448–464.
5. Rasquin A, Di Lorenzo C, Forbes D, et al. Chidlhood functional gastrointestinal disorders: child/adolescent. Gastroenterology 2006; 130:1527–1537.
6. Roma E, Adamidis D, Gianniou M, et al. Epidemiology of chronic constipation in Greek children. Hell J Gastroenterol 1999; 12:58–62.
7. Inan M, Aydiner CY, Tokuc B, et al. Factors associated with childhood constipation. J Paediatr Child Health 2007; 43:700–706.
8. Devanarayana NM, Rajindrajith S. Association between constipation and stressful life events in a cohort of Sri Lankan children and adolescents. J Trop Pediatr 2010; 56:144–148.
9. Ozokutan BH, Zoroglu S, Ceylan H, et al. Psychological evaluation of children with idiopathic constipation and their parents. Pediatr Int 2005; 47:311–315.
10. Ip KS, Lee WT, Chan JS, et al. A community-based study of the prevalence of constipation in young children and the role of dietary fibre. Hong Kong Med J 2005; 11:431–436.
11. Rajindrajith S, Devanarayana NM. Constipation in children: novel insight into epidemiology, pathophysiology and management. J Neurogastroenterol Motil 2011; 17:35–47.
12. Pijpers MA, Bongers ME, Benninga MA, et al. Functional constipation in children: a systematic review on prognosis and predictive factors. J Pediatr Gastroenterol Nutr 2010; 50:256–268.
13. Mugie SM, Benninga MA, Di Lorenzo C. Epidemiology of constipation in children and adults: a systematic review. Best Pract Res Clin Gastroenterol 2011; 25:3–18.
14. Walker LS, Caplan A, Rasquin A. Questionnaire on pediatric gastrointestinal symptoms, Rome III Version (QPGS-RIII). In: Drossman DA, Corazziari E, Delvaux M, et al, eds. Rome III: The Functional Gastrointestinal Disorders. 3rd ed. McLean, VA: Degnon Associates; 2006: 963–90.
15. Williams CL, Bollella M, Wynder EL. A new recommendation for dietary fiber in childhood. Pediatrics 1995; 96:985–988.
16. Leung SS, Cole TJ, Tse LY, et al. Body mass index reference curve for Chinese children. Ann Hum Biol 1998; 25:169–174.
17. So HK, Nelson EA, Li AM, et al. Secular changes in height, weight and body mass index in Hong Kong children. BMC Public Health 2008; 8:320.
18. Wald ER, Di Lorenzo C, Cipriani L, et al. Bowel habits and toilet training in a diverse population of children. J Pediatr Gastroenterol Nutr 2009; 48:294–298.
19. Rajindrajith S, Devanarayana NM, Adhikari C, et al. Constipation in children: an epidemiological study in Sri Lanka using Rome III criteria. Arch Dis Child 2010;97:43–5.
20. Zhou H, Yao M, Cheng GY, et al. Prevalence and associated factors of functional gastrointestinal disorders and bowel habits in Chinese adolescents: a school-based study. J Pediatr Gastroenterol Nutr 2011; 53:168–173.
21. Kajiwara M, Inoue K, Usui A, et al. The micturition habits and prevalence of daytime urinary incontinence in Japanese primary school children. J Urol 2004; 171:403–407.
22. Chung JM, Lee SD, Kang DI, et al. An epidemiologic study of voiding and bowel habits in Korean children: a nationwide multicenter study. Urology 2010; 76:215–219.
23. Uguralp S, Karaoglu L, Karaman A, et al. Frequency of enuresis, constipation and enuresis association with constipation in a group of school children aged 5–9 years in Malatya, Turkey. Turk J Med Sci 2003; 33:315–320.
24. Yong D, Beattie RM. Normal bowel habit and prevalence of constipation in primary-school children. Ambul Child Health 1998; 4:277–282.
25. Loening-Baucke V. Prevalence, symptoms and outcome of constipation in infants and toddlers. J Pediatr 2005; 146:359–363.
26. Loening-Baucke V. Prevalence rates for constipation and faecal and urinary incontinence. Arch Dis Child 2007; 92:486–489.
27. Iacono G, Merolla R, D’Amico D, et al. Gastrointestinal symptoms in infancy: a population-based prospective study. Dig Liver Dis 2005; 37:432–438.
28. Uc A, Hyman PE, Walker LS. Functional gastrointestinal disorders in African American children in primary care. J Pediatr Gastroenterol Nutr 2006; 42:270–274.
29. Misra S, Lee A, Gensel K. Chronic constipation in overweight children. J Parenter Enteral Nutr 2006; 30:81–84.
30. Fishman L, Lenders C, Fortunato C, et al. Increased prevalence of constipation and fecal soiling in a population of obese children. J Pediatr 2004; 145:253–254.
31. Pashankar DS, Loening-Baucke V. Increased prevalence of obesity in children with functional constipation evaluated in an academic medical center. Pediatrics 2005; 116:e337–e380.
32. Pourhoseingholi MA, Kaboli SA, Pourhoseingholi A, et al. Obesity and functional constipation: a community-based study in Iran. J Gastrointestinal Liver Dis 2009; 18:151–155.
33. Bytzer P, Howell S, Leemon M, et al. Low socioeconomic class is a risk factor for upper and lower gastrointestinal symptoms: a population-based study in 15000 Australian adults. Gut 2001; 49:66–72.
34. Iglowstein I, Jenni OG, Molinari L, et al. Sleep duration from infancy to adolescence: reference values and generational trend. Pediatrics 2003; 111:302–307.

children; constipation; prevalence; Rome III criteria

Copyright 2012 by ESPGHAN and NASPGHAN