1. Introduction

Thirty-eight percent of school-children report abdominal pain weekly,¹but only 2-4% of children who report abdominal pain seek medical consultation.¹,² Despite this, abdominal pain is the most common cause of consultation to the pediatric gastroenterologist. Out of all children (> 4 years) who consulted pediatric gastroenterology in a tertiary care center in the US, 52% of them consulted for an abdominal pain associated-functional gastrointestinal disorder (AP-FGID).³The cost of care of children with AP-FGIDs is substantial. In the Netherlands, the costs of care for a child with AP-FGIDs exceeds €2500.⁴ The overall cost of care for AP-FGIDs in the US grew by 300% in a 12 year period⁵ and evaluation of a single child for an AP-FGIDs in the US is estimated > $6000.⁶ The cost of a single consultation for abdominal pain in Uruguay accounted for 3.8% of the per capita health care spending of one year.⁷Although, the economic impact of AP-FGIDs has not been studied in most countries, based on the high worldwide prevalence of FGIDs it is likely to be a global economic problem.

The magnitude of the problem calls for preventative measures and plans of care adapted to local needs. However, in order to establish what the local needs are, it is imperative to first obtain reliable data on the local epidemiology. Due to the low ratio of consultation for abdominal pain, studies performed at the doctor's office are prone to selection bias. Thus, only community studies provide the information needed. Conducting epidemiological studies in private and public schools allows obtaining data from children with different socio-economic statuses. There have been several school studies investigating the prevalence of FGIDs, however none of them were conducted in private and public schools in multiple cities throughout different regions of one country.

In addition to establishing the regional epidemiology, nationwide studies provide a unique opportunity to achieve a better understanding of the pathogenesis of FGIDs that is currently elusive. This information is key to establish prevention strategies and effective treatments. The most accepted model to explain FGIDs propose that these disorders result from the interaction of the child's genetic background, early life experiences, intestinal microbiota, psychological, and social influences. As these factors vary by region, establishing comparisons among regions could help advance the knowledge of the pathogenesis of FGIDs.

We conducted a large school-based study of FGIDs in multiple cities in Colombia using standardized methods, validated questionnaires adapted to the terms used by children, and the Rome III diagnostic criteria. The aims of this study were to evaluate prevalence of pediatric FGIDs throughout Colombia, determine if regional differences exist, and identify associated factors.

2. Methods

This is the largest study conducted by the Functional International Digestive Epidemiological Research Survey (FINDERS) group and the broadest national study in the Western Hemisphere on FGIDs in children. This cross-sectional study was carried out at 12 schools in 10 cities of Colombia. Colombia is the second most populated country in South-America with a population of approximately 50 million⁸ inhabitants. Most of the population is concentrated along the Caribbean coast and the Andean regions. In Colombia, 85% of primary care education students and 60% of the secondary-school level students attend public schools.⁹ To enhance racial, ethnic, and socioeconomic diversity and maximize the external validity of the data, we selected public and private schools as study sites. The study was purposefully designed to enroll a higher proportion of children from public than private schools to resemble the education system of the country. To better represent the population distribution of Colombia, we selected more cities with high population density than low. The cities of the study are distributed through the four main regions of Colombia, the Caribbean region (adjacent to the Caribbean Sea), Andean region (by the mountains), the Amazon region (tropical rainforest), and the Pacific region (neighboring the Pacific Ocean). Together these four regions hold 97.3% of the population in Colombia.

To ensure consistency and to allow comparison within multi-national cross-cultural studies, the study was conducted in a systematic manner following the same methods used in previous studies by the international consortium FINDERS.¹,¹⁰-¹²The methods used by FINDERS were based on two school studies performed previously and a prior small study by FINDERS in Colombia. The first study was piloted in a single school in Pittsburgh¹³ followed by a larger study in Chicago to streamline methods.² Planning of the current study in Colombia started by translating the English version of the QPGS-III questionnaire into Spanish and local language according to established standards. The self-report child version of the QPGS-III questionnaire (a validated instrument created to assess the full spectrum of FGIDs according to the Rome III criteria) was first translated into Spanish by a bilingual pediatric gastroenterologist and two bilingual medical students. This was followed by the reverse translation of the QPGS-III by another member of the research team. The resulting English version was then compared with the original English version of the QPGS-III by a native English speaker who was not part of the research team to ascertain fidelity. The finalized Spanish version of the QPGS-III was then presented to two focus groups of children in two different cities in Colombia (Cali and Pasto) to ensure terms were adequately understood. Once the Spanish version of the QPGS-III was tested in the focus groups, the methods and the Spanish version of the questionnaire were again tested in the first study by FINDERS in Pasto, Colombia.¹The data obtained from this study using the Spanish version of the QPGS III were analyzed statistically for internal consistency using Cronbach's alpha analysis.¹⁴The study showed high internal consistency. In short, the methods used in the current study were the following: parents of children 8-18 years of age received a package with information on the study at home, a questionnaire on their child's past medical history and a consent form. Children who assented (and whose parents consented) to participate in the study underwent an informative session in class explaining the study and questionnaires. The participants then completed the data forms in class with a member of the research team available for questions. The data was entered into an electronic database. In order to ensure accuracy of data transfer, 10% of the questionnaires were reviewed and compared with the paper forms collected at the school level.

The study was approved by the principal of each of the schools and by the Institutional Review Board and Human Subjects Committee of Universidad del Valle of Cali, Colombia

Statistical analysis included calculation of odds ratios, 95% confidence intervals, and p values to describe the associations of interest (Stata 10 software, College Station, Texas). Fisher's exact test was used when appropriate. Two-sided p values <0.05 were considered statistically significant.

3. Results

A total of 5062 school-children's families were sent the information package and invited to participate in the study. 4394/5062 (86.8%) dyads (children/caretaker) assented/consented to participate. Participants ranged from 8 to 18 years of age (mean = 11.9, SD = 2.3). There were 2279 males (51.9%) and 2115 (48.1%) females; 2702 (61.5%) were children (8 to 12 years) and 1692 (38.5%) adolescents (13 to 18 years). Most children who participated in the study were from public schools 3546 (80.7%) and 848 (19.3%) were from private schools.

4. Discussion

This is the largest epidemiological study conducted in the Western Hemisphere aimed at describing FGIDs in children using the Rome III criteria. The study included 4394 children from 12 public and private schools in 10 geographically dispersed cities in Colombia, providing a broad-based survey throughout a Latin American nation. The nationwide distribution of our sample and the high acceptance rate to participate in the study (86.8%) indicates the external validity of our data. We found that almost one in four children (23.7%) had a FGID according to the Rome III criteria. The prevalence of FGIDs found in this study was similar to that found in Quito (22.8%),¹²the capital of Ecuador which borders Colombia to the south. Similar to the results of this study (11.6%), functional constipation was the most common FGID in Ecuador (11.8%).

We found that 10.4% of children in Colombia had AP-FGIDs. This is again within close range to the prevalence of AP-FGIDs found in Ecuador (9.4%) using the same methods.¹² A review of AP-FGIDs in children found that worldwide pooled prevalence of AP-FGIDs was higher than in our sample 13.5% (95% CI 11.8-15.3).¹⁵The study identified IBS as the most common AP-FGID, but at higher rates (8.8%, 95% CI 6.2-11.9) than in our study (5.1%). The study found a high range of AP-FGID prevalence across countries ranging from 1.6% to 41.2%. Prevalence of adult IBS is estimated to be 21-26% in Peru¹⁶ and 19.9% in Colombia, which are higher than in our study in children. The reasons for the differences in prevalence of AP-FGIDs between children and adults and among geographic regions remains unknown and should be studied with large international comparative studies using standardized methodology and subjects of different age groups.

FGIDs were more common in children from private than public schools and in children whose parents were separated. Another Latin-American study also found that children from divorced/separated families had a higher prevalence of FGIDs¹¹ than children from intact families. Although the design of this study does not allow determination of causality, the results are in line with the biopsychosocial model that proposes psychological and social factors play a role in the pathogenesis of FGIDs. According to the biopsychosocial model, AP-FGID symptoms result from the interaction of gastrointestinal and central nervous system factors.¹⁷Stress has pathophysiological effects in the gastrointestinal tract. Higher stress levels can trigger or exacerbate abdominal pain through visceral hypersensitivity and changes in motility. Children with depressive or anxious symptoms are more likely to suffer from AP-FGIDs¹⁸ and studies in school-children found an association between anxiety and depression with abdominal pain.¹⁹Divorce is a major stressor²⁰ and parental separation and divorce are associated with higher rates of depression.²¹Studies comparing early adolescents from divorced and intact families found that adolescents from divorced families had higher rate of anxiety and more depressed mood.²²A higher prevalence of FGIDs in children from private schools was also found in other studies conducted in Latin-American countries such as El Salvador¹¹ and Ecuador.¹²A study that interviewed more than 500 adolescents in Colombia found higher levels of anxiety among children of public schools.²³However, the study was performed in a single town near Bogota and it is also known that the interrelation between the psychological factors involved in the pathogenesis of FGIDs is complex.²⁴Moreover, other non-psychological factors could explain these findings. Future studies by FINDERS will specifically investigate these issues.

Despite previous studies in the United States showing that children of parents with IBS reported more gastrointestinal symptoms,²⁵,²⁶we did not find an association between FGIDs and family history. A family history of FGIDs also did not associate with higher prevalence of FGIDs in Salvadoran children.¹¹The reason for this difference is unclear. Cultural differences in the way a FGID is conceptualized and differences in parenting and coping styles between the US and Latin-American societies may explain these differences.

We did not find gender predominance in FGIDs. These results are similar to those of pediatric studies in Mexico,²⁷ Ecuador,¹² Panama,¹⁰ and El Salvador.¹¹ A meta-analysis by Lovell and Ford¹⁶ found a female predominance in adults with IBS. However, this difference was not found when South-American studies were analyzed as a whole.

Limitations of our study include the completion of questionnaires exclusively by children and not by parents. It can be argued that children know their symptoms better than their parents, and may be reluctant to share information regarding bowel movements with their parents, while having comfort reporting on an anonymous questionnaire. We cannot rule out the possibility that although our study included multiple cities in various regions, the results may not be generalized to all cities in Colombia. Additionally, concurrent medical diagnoses may be present in surveyed children with or without FGIDs and were not assessed in the study.

In conclusion, this study provides relevant information for public health planning, finding a high prevalence of FGIDs in children of Colombia. Children from private schools and from families whose parents have separated are at higher risk of FGIDs, and nausea is associated with increased rates of FGIDs in children.