2022 Impact Factor
2022 Impact Factor
,1 Seung Won Lee,2 Yoowon Kwon,3 Eun Kyo Ha,4 Jaewoo An,5 Hye Ryeong Cha,6 Su Jin Jeong,5* and Man Yong Han5*
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Infantile colic is a common disease in the infant period, affecting between 4% and 28% of all infants.1 Its most prominent feature is excessive crying, which peaks at 6 weeks and disappears at approximately 4 to 5 months of age.2 Although the pathogenesis of infantile colic is not well understood, it has been generally viewed as a temporary disorder with a benign prognosis. However, it has been recently reported that infantile colic is associated with a low grade systemic inflammation caused by a pathogenic microbiota composition.3 Therefore, infantile colic may be considered as a potential early manifestation of later onset disorders including the irritable bowel syndrome (IBS). IBS is a common and chronic functional gastrointestinal disorder (FGID) at childhood that affects the quality of life, with a prevalence of 2% to 24% worldwide.4 A multinational group of experts at the Rome IV Conference in December 2014 established the most recent definition of IBS.5 A characteristic of IBS is abdominal pain or discomfort with changes in bowel habits, including the frequency of defecation and stool consistency, in the absence of an organic disease.6
The pathophysiology of IBS is not understood completely.7 Various central and peripheral mechanisms have been described in the pathophysiology of IBS, including the dysregulation of the brain-gut axis,8 altered gastrointestinal motility, visceral hypersensitivity,9 alterations in the intestinal microbiota,10 low-grade immune activation, and intestinal inflammation,7 which may be associated with increased intestinal permeability to induce exposure to antigens. Gut dysbiosis is thought to be the most convincing factor, as some factors may be derived directly from the gut dysbiosis.11 Some points of pathogenesis show an association between infantile colic and IBS.
Few previous studies have reported an association between infantile colic and the subsequent development of IBS,12,13 but there is a paucity of large cohort studies reporting the association between infantile colic in infants and IBS. Therefore, our purpose is to evaluate the association of infantile colic diagnosis at 5 weeks to 4 months of age with the onset of IBS after 4 years of age, using an analysis of an administrative nationwide cohort database in South Korea.
This study was conducted using data from the National Investigation of Birth Cohort in Korea study 2008 (NICKs-2008) dataset, which comprised 917 707 children born between 2008 and 2009 in Korea.14,15 The National Health Screening Program for Infants and Children (NHSPIC) is a population surveillance program that provides seven rounds of screening services to all health insurance subscribers from 4 months to 72 months of age. The use of de-identified individual data for research purposes was authorized under the current National Health Insurance Act. The study protocol was reviewed and approved by the Institutional Review Board of the Korea National Institute for Bioethics Policy (P01-201603-21-005). The study followed recommended guidelines for observational studies that use routinely collected health data (Supplementary Table 1).
Patient demographic characteristics, health care utilization (type of visit, medical cost, drug classification code, and disease information [International Classification of Diseases 10th revision {ICD-10 codes}] were obtained from the database of the NICKs-2008 cohort. The patients’ health care utilization data were collected during the process of claiming for healthcare services. The residential status at birth was divided into Seoul, metropolitan (Busan, Daegu, Incheon, Gwangju, Daejeon, and Ulsan), urban, and rural. The socio-economic status was classified according to quintiles and determined by the amount of the insurance co-payment. Additional information on the databases, variable definitions, and administrative codes are shown in Supplementary Table 2. The cohort was traced to the time when participants lost their eligibility for health care services due to death or immigration, or until 2017.
This study was based on the National Health Claims Database (National Health Insurance Service [NHIS]-2019-1-560) established by the NHIS of the Republic of Korea. Applications for using NHIS data are reviewed by the Inquiry Committee of Research Support; if the application is approved, raw data is provided to the applicant for a fee. We cannot provide access to the data, analytic methods, and research materials to other researchers because of the intellectual property rights of this database that is owned by the National Health Insurance Corporation. However, investigators who wish to reproduce our results or replicate the procedure can be used in the database, which is open for research purposes (https://nhiss.nhis.or.kr/).
The cohort included children who underwent the first round of the NHSPIC at 4 months to 6 months of age and with properly recorded birth weight information during the first round of the NHSPIC. In addition, children who were born before 37 weeks of gestational age, born with a birth weight of less than 2.5 kg or more than 4.5 kg, who had been admitted to the intensive care unit within the first 1 month of birth, or who were diagnosed with congenital malformations, deformations, and chromosomal abnormalities, were excluded. A total of 368 556 children met the inclusion and exclusion criteria. Among these children, 359 769 who had not been diagnosed with infantile colic from the age of 5 weeks to 4 months were allocated to the control group and 8787 who had been diagnosed with infantile colic between the ages of 5 weeks and 4 months were allocated to an infantile colic group (Fig. 1).
Figure 1. Flow diagram of the cohort. NICKs, National Investigation of Birth Cohort in Korea study; NHSPIC, National Health Screening Program of Infants and Children; GA, gestational age; ICU, intensive care unit; ICD, International classification of diseases; IPTW, inverse probability of treatment weighting.
The exposure of interest was a diagnosis of infantile colic from the age of 5 weeks to 4 months, because this time of life is known to be the most prevalent period of infantile colic.3,4 The diagnosis of infantile colic was defined as at least 1 diagnosis of ICD-10 code R10.4 (other and unspecified abdominal pain) or R68.1 (nonspecific symptoms peculiar to infancy) from the age of 5 weeks to 4 months (Supplementary Table 3). Validation data for this definition of infantile colic are presented in the Supplementary material, and the positive predictive value (PPV) was 88% (95% CI, 83-92%).
All outcomes were pre-specified. The outcome was the risk of IBS after 48 months of age. IBS was defined as a diagnosis of ICD-10 codes K58.X (IBS), more than twice, after 4 years of age.16
To assess the robustness of the results, we analyzed the association of IBS with infantile colic using the different definitions of IBS. The stricter IBS was defined as at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 of ICD-10 code K59.1 (functional diarrhea), K52.2 (allergic and dietetic gastroenteritis and colitis), K52.8 (other specific noninfective gastroenteritis), or K59.0 (constipation) after the age of 4 years. A previous study showed that this definition had a PPV of 83% (95% CI, 75-91%).17 In addition, we defined the IBS sub-type by further considering the symptomatic diagnostic ICD-10 codes. Diarrhea-predominant IBS (IBS-D) was defined as having at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 of the ICD-10 codes K59.1 (functional diarrhea) after the age of 4 years. Constipation-predominant IBS (IBS-C) was defined as having at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 of ICD-10 code K59.0 (constipation) after the age of 4 years. To apply stricter diagnostic criteria for the IBS sub-types, the strict IBS-D was defined as using least 1 prescription of drug classification 237 (antidiarrheal drug) as well as the IBS-D definition, and the strict IBS-C was defined as using at least 1 prescription of drug classification 238 (laxative drug) as well as IBS-C definition, which showed PPV 76% (95% CI, 67-85%) in a previous study.17
Propensity score (PS) matching was performed using the 97 covariates shown in Supplementary Table 2 in order to create a balance between the control group and the colic group. The children’s demographic variables included age, sex, income quintile (by insurance premium), and residence at birth. Nutritional assessments, including the status of breastfeeding within 4 months of age and the additional complementary feeding status were acquired from the first round of the NHSPIC questionnaire (4 to 6 months of age) given to children’s parents. In addition, the results of general physical examinations from head to extremity by physicians in the first round of the NHSPIC were assessed. The perinatal conditions were evaluated using the ICD-10 codes. To evaluate their clinical conditions within the first 6 months of age, the most prevalent 43 diseases, including respiratory infections, gastrointestinal infections, and dermatitis, were identified using the ICD-10 codes. In addition, 12 types of drugs, including the drugs used commonly in infants, were defined using the drug classification codes. Hospitalization records, emergency room visits, and visits to pediatricians within the first 6 months of age were recorded.
The inverse probability of treatment weighting (IPTW) on the PS was used to balance the comparison groups on the 97 covariates that were chosen previously (as defined in the Supplementary Table 2). The PS was estimated using a multivariable logistic regression with the covariates chosen a priori. Participants in the reference group were weighted (PS/[1-PS]),18,19 which produced a weighted pseudosample of patients in the reference group with the same distribution of measured covariates as the exposed group.19 The between-group differences in baseline characteristics were compared using standardized differences in both the unweighted and weighted samples (differences > 10% were considered meaningful).20 A Cox proportional hazards model was used to assess the relationship between the infantile colic at 5 weeks to 4 months of age and the development of IBS. Hazard ratios (HRs) and 95% CIs were calculated for each outcome of interest.
To minimize the type 1 error by the inflated number from IPTW, we additionally analyzed using stabilized IPTW and weight trimming. In addition, a 4:1 PS matching was performed instead of IPTW, based on the Mahalanobis algorithm with a caliper of 0.01 and multivariable regression with covariates.
In addition, prespecified subgroup analyses by type of feeding (grouped into 2 categories: breast fed, or no breast fed), birth weight (grouped into 2 categories: lower half (< 3.26 kg), or upper half (≥ 3.26 kg), sex (male or female), and income status (low or high). Furthermore, the number of episodes of infantile colic (grouped into 2 categories: once or more than once) was assessed to determine whether the higher the number of infantile colic episodes led to a higher the risk of IBS.
Two-tailed
The basic sociodemographic characteristics of the children are presented in Table 1. In the weighted data, both the control and infantile colic groups were balanced for all variables. In addition to the results that comprised the results of the physical examinations by physicians and questionnaires, including visual and auditory senses at 4-6 months of age, there was no imbalance between the 2 groups (Supplementary Table 4). The basic clinical characteristics of the children in both groups are shown in Table 2. Although there were imbalances between the 2 groups in the number of visits to pediatricians and emergency rooms (standardized differences = 22.5% and 32.3%, respectively) before weighting, all data on the hospital utilization was balanced between the 2 groups after weighting. With regard to the perinatal conditions, the prevalence of diagnosed diseases within the first 6 months of age, and the use of major drugs, and all standardized differences between the 2 groups, were less than 10% after weighting. The full set of clinical characteristics is shown in Supplementary Table 5.
Table 1 . Basic Sociodemographic Characteristics of the Participantsa
| Sociodemographic characteristics | Observed data (n = 368 556) | Weighted data (n = 723 370)b | ||||
|---|---|---|---|---|---|---|
| n (%)c | Standardized difference (%)f | n (%)c | Standardized difference (%)f | |||
| Control groupd (n = 359 769) | Infantile colic groupe (n =8787) | Control groupd (n = 363 528) | Infantile colic groupe (n = 359 842) | |||
| Sex | ||||||
| Female | 176 967 (49.2) | 4331 (49.3) | 0.2 | 178 831 (49.2) | 175 873 (48.9) | 0.3 |
| Male | 182 802 (50.8) | 4456 (50.7) | 184 696 (50.8) | 183 969 (51.1) | ||
| Residence at birthg | ||||||
| Seoul | 89 501 (24.9) | 1977 (22.5) | 5.6 | 90 259 (24.8) | 98 793 (27.5) | 0.1 |
| Metropolitan | 83 324 (23.2) | 1656 (18.8) | 10.6 | 83 809 (23.1) | 76 547 (21.3) | 0.1 |
| Urban | 142 540 (39.6) | 4031 (45.9) | 12.7 | 144 522 (39.8) | 142 079 (39.5) | 0.0 |
| Rural | 41 184 (11.4) | 1021 (11.6) | 0.5 | 41 658 (11.5) | 39 266 (10.9) | 0.1 |
| Birth weighth (mean [SD], kg) | 3.26 (0.36) | 3.24 (0.36) | 4.6 | 3.26 (0.36) | 3.24 (0.34) | 0.1 |
| Type of feedingi | ||||||
| Only breastfeeding | 163 567 (45.5) | 3745 (42.6) | 5.7 | 165 601 (45.6) | 165 291 (45.9) | 0.2 |
| Only formula milk | 124 583 (34.6) | 3344 (38.1) | 7.1 | 126 706 (34.9) | 124 049 (34.5) | 0.2 |
| Mixed | 68 775 (19.1) | 1593 (18.1) | 2.5 | 69 773 (19.2) | 68 877 (19.1) | 0.1 |
| Special milk | 1338 (0.4) | 72 (0.8) | 5.8 | 1401 (0.4) | 1581 (0.4) | 0.1 |
| Income quintilej | ||||||
| 1 (Lowest) | 27 617 (7.7) | 683 (7.8) | 0.4 | 27 903 (7.7) | 26 791 (7.4) | 0.2 |
| 2 | 52 574 (14.6) | 1334 (15.2) | 1.6 | 53 143 (14.6) | 54 083 (15.0) | 0.1 |
| 3 (Middle) | 95 514 (26.5) | 2354 (26.8) | 0.5 | 96 534 (26.6) | 96 592 (26.8) | 0.0 |
| 4 | 113 679 (31.6) | 2792 (31.8) | 0.4 | 114 922 (31.6) | 113 306 (31.5) | 0.1 |
| 5 (Highest) | 58 181 (16.2) | 1336 (15.2) | 0.2 | 58 688 (16.1) | 56 961 (15.8) | 0.1 |
aUnless otherwise specified, baseline characteristics were assessed on the birth date of the patient.
bWeighted using inverse probability of exposure weighting based on the propensity score. The propensity score was estimated using multivariable logistic regression with 97 previously covariates, as defined in Supplementary Table 2. Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudo sample of participants in the reference group with the same distribution of measured covariates as the exposure group.
cResults are reported as n (%) unless otherwise indicated.
dAs the reference group, the control group consisted of children who had not been diagnosed with infantile colic at 5 weeks to 4 months of age.
eThe infantile colic group consists of children who have been diagnosed with infantile colic at least once between 5 weeks and 4 months of age.
fDifferences greater than 10% were interpreted as meaningful differences. All standardized differences in the cohort values were < 0.05.
gMetropolitan areas were defined as 6 metropolitan cities (Busan, Incheon, Gwangju, Daejeon, Daegu, and Ulsan), urban areas as cities, and rural areas as non-city areas. Missing data of observed data: control group = 3220, infantile colic group = 102; missing data of weighted data: control group = 3281, infantile group = 3157.
hObtained from the first National Health Screening Program of Infants and Children (NHSPIC) at 4-6 months of birth.
iObtained from the first NHSPIC at 4-6 months of birth. Missing data of all data: control group = 1506, infantile colic group = 33.
jIncome status was categorized into the quintile of insurance premium at birth. Missing data of all data: control group = 12 204, infantile colic group = 288; missing data of weighted data: control group = 12 338, infantile colic group = 12 109.
Table 2 . Basic Clinical Caracteristics of the Participantsa
| Clinical characteristics | Observed data (n = 368 556) | Weighted data (n = 723 370)b | ||||
|---|---|---|---|---|---|---|
| n (%)c | Standardized difference (%)f | n (%)c | Standardized difference (%)f | |||
| Control groupd (n = 359 769) | Infantile colic groupe (n =8787) | Control groupd (n = 363 528) | Infantile colic groupe (n = 359 842) | |||
| Hospital utilization within 6 months of age | ||||||
| Number of visits to pediatricians | 341 551 (94.9) | 8688 (98.9) | 22.5 | 345 435 (95.0) | 342 290 (95.1) | 2.2 |
| Number of visiting ER | 15 903 (4.4) | 1174 (13.4) | 32.3 | 16 843 (4.6) | 18 281 (5.1) | 1.2 |
| Certain conditions (ICD-10 codes) originating in the perinatal period | ||||||
| Respiratory and cardiovascular disorders specific to the perinatal period | 14 605 (4.1) | 427 (4.9) | 4.0 | 14 845 (4.1) | 14 480 (4.0) | 0.1 |
| Infections specific to the perinatal period | 49 707 (13.8) | 1373 (15.6) | 5.3 | 50 342 (13.8) | 51 833 (14.4) | 0.2 |
| Hemorrhagic and hematological disorders of fetus and newborn | 112 149 (31.2) | 2755 (31.4) | 0.2 | 113 338 (31.2) | 112 868 (31.4) | 0.1 |
| Digestive system disorders of fetus and newborn | 10 079 (2.8) | 331 (3.8) | 5.4 | 10 283 (2.8) | 10 410 (2.9) | 0.0 |
| Prevalent diseases (ICD-10 codes) diagnosed within 6 months of age | ||||||
| Otitis media, unspecified | 7989 (2.2) | 204 (2.3) | 0.4 | 8058 (2.2) | 7869 (2.2) | 0.2 |
| Acute nasopharyngitis | 104 381 (29.0) | 3056 (34.8) | 12.5 | 105 974 (29.2) | 107 273 (29.8) | 0.1 |
| Acute pharyngitis | 37 920 (10.5) | 1103 (12.6) | 6.4 | 38 483 (10.6) | 39 901 (11.1) | 0.6 |
| Acute upper respiratory infection, unspecified | 81 665 (22.7) | 2198 (25.0) | 5.1 | 82 711 (22.8) | 84 377 (23.4) | 0.5 |
| Pneumonia, unspecified | 9496 (2.6) | 225 (2.6) | 0.1 | 9569 (2.6) | 10 044 (2.8) | 0.6 |
| Acute bronchitis | 66 304 (18.4) | 1634 (18.6) | 0.6 | 66 995 (18.4) | 66 559 (18.5) | 0.7 |
| Acute bronchiolitis | 48 798 (13.6) | 1129 (12.8) | 2.2 | 49 176 (13.5) | 50 601 (14.1) | 0.7 |
| Viral intestinal infection, unspecified | 9741 (2.7) | 459 (5.2) | 12.5 | 10 064 (2.8) | 11 772 (3.3) | 0.6 |
| Infectious gastroenteritis and colitis of infectious origin | 5416 (1.5) | 192 (2.2) | 5.2 | 5524 (1.5) | 5995 (1.7) | 0.6 |
| Gastroesophageal reflux disease without esophagitis | 8416 (2.3) | 333 (3.8) | 8.6 | 8617 (2.4) | 8590 (2.4) | 0.4 |
| Functional dyspepsia | 7635 (2.1) | 315 (3.6) | 8.7 | 7834 (2.2) | 8289 (2.3) | 0.2 |
| Noninfective gastroenteritis and colitis, unspecified | 16 372 (4.6) | 561 (6.4) | 8.2 | 16 709 (4.6) | 18 306 (5.1) | 0.8 |
| Constipation | 15 365 (4.3) | 574 (6.5) | 9.9 | 15 699 (4.3) | 16 272 (4.5) | 0.7 |
| Functional diarrhea | 5223 (1.5) | 205 (2.3) | 6.1 | 5360 (1.5) | 5375 (1.5) | 0.0 |
| Functional intestinal disorder, unspecified | 10 060 (2.8) | 357 (4.1) | 7.3 | 10 248 (2.8) | 10 274 (2.9) | 0.1 |
| Dermatitis | 15 342 (4.3) | 480 (5.5) | 5.6 | 15 630 (4.3) | 15 785 (4.4) | 0.1 |
| Pyogenic granuloma | 4750 (1.3) | 131 (1.5) | 1.1 | 4820 (1.3) | 4498 (1.3) | 0.0 |
| Urinary tract infection, site not specified | 7731 (2.1) | 231 (2.6) | 3.1 | 7861 (2.2) | 8302 (2.3) | 0.1 |
| Nausea and vomiting | 10 334 (2.9) | 495 (5.6) | 13.6 | 10 699 (2.9) | 11 869 (3.3) | 0.5 |
| Fever, unspecified | 18 237 (5.1) | 524 (6.0) | 4.1 | 18 507 (5.1) | 21 194 (5.9) | 0.1 |
| Drug (drug classification code) used within 6 months of age | ||||||
| Sedative-hypnotics | 10 825 (3.0) | 524 (6.0) | 14.3 | 11 172 (3.1) | 11 444 (3.2) | 0.5 |
| Antipyretics | 126 399 (35.1) | 3410 (38.8) | 7.5 | 127 985 (35.2) | 134 985 (37.5) | 4.4 |
| Antispasmodics | 8222 (2.3) | 638 (7.3) | 23.1 | 8741 (2.4) | 9205 (2.6) | 0.5 |
| Antihistamines | 168 646 (46.9) | 4231 (48.2) | 2.4 | 179 449 (49.4) | 168 537 (46.8) | 0.7 |
| Respiratory system drugs | 195 021 (54.2) | 4905 (55.8) | 3.3 | 197 134 (54.2) | 198 425 (55.1) | 2.8 |
| Digestive system drugs | 188 764 (52.5) | 5899 (67.1) | 30.3 | 191 926 (52.8) | 191 892 (53.3) | 0.8 |
| Hormone drugs | 17 352 (4.8) | 432 (4.9) | 0.5 | 17 530 (4.8) | 16 764 (4.7) | 0.6 |
aUnless otherwise specified, all baseline characteristics were assessed in the first 6 months of age of the participants.
bWeighted using inverse probability of exposure weighting based on the propensity score. The propensity score was estimated using multivariable logistic regression with 97 previously covariates, as defined in Supplementary Table 2 Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudo sample of participants in the reference group with the same distribution of measured covariates as the exposure group.
cResults are reported as n (%) unless otherwise indicated.
dAs the reference group, the control group consisted of children who had not been diagnosed with infantile colic at 5 weeks to 4 months of age.
eThe infantile colic group consists of children who have been diagnosed with infantile colic at least once between 5 weeks and 4 months of age.
fDifferences greater than 10% were interpreted as meaningful differences. All standardized differences in the cohort values were < 0.05.
ICD, International Classification of Diseases; ER, emergency room.
After each application of the stabilized and trimmed IPTW and the 4:1 PS matching, all standardized differences between the infantile colic and control group were less than 10% (data not shown).
Table 3 shows the associations of infantile colic with IBS. In the weighted data, 61 759 (17.2%) children with IBS were in the infantile colic group and 56 547 (15.6%) in the control group (HR [95% CI], 1.12 [1.10 to 1.13]). In the stabilized and trimmed weighted data and the 4:1 matched data, the associations between infantile colic and IBS were still statistically significant and consistent with the results in the weighted data.
Table 3 . Risk of Irritable Bowel Syndrome With Infantile Colica
| Unweighted data | |||
|---|---|---|---|
| Control groupb (n = 359 769) | Infantile colic groupc (n = 8787) | HR (95% CI)d | |
| IBSe (n [%]) | 55 887 (15.5) | 1533 (17.5) | 1.14 (1.08 to 1.20) |
| Weighted dataf | |||
|---|---|---|---|
| Control group (n = 363 528) | Infantile colic group (n = 359 842) | HR (95% CI) | |
| IBSe (n [%]) | 56 547 (15.6) | 61 759 (17.2) | 1.12 (1.10 to 1.13) |
| Sensitivity analysis | |||
|---|---|---|---|
| Stabilized and trimmed weighted data | |||
| Control group (n = 363 528) | Infantile colic group (n = 85 854) | HR (95% CI) | |
| IBSe (n [%]) | 56 547 (15.5) | 14 929 (17.4) | 1.13 (1.11 to 1.15) |
| 4:1 Matched data | |||
|---|---|---|---|
| Control group (n = 34 408) | Infantile colic group (n = 8637) | HR (95% CI) | |
| IBSe (n [%]) | 5561 (16.2) | 1504 (17.4) | 1.09 (1.03 to 1.15) |
aThe cohort consisted of the infantile colic group who had experienced infantile colic from 5 weeks to 4 months of age and the control group without infantile colic histories during the same time period.
bAs the reference group, the control group comprises children who had not been diagnosed with infantile colic at 5 weeks to 4 months of age.
cInfantile colic group consists of children who have been diagnosed with infantile colic at least once between 5 weeks and 4 months of age.
dThe hazard ratios (HRs) were assessed using a Cox proportional hazards model to examine the relationship between infantile colic histories and the risk of irritable bowel syndrome (IBS) development in the cohort.
eIBS was defined as having the diagnosis of International Classification of Diseases 10th version (ICD-10) codes K58.X (IBS) more than twice, after 4 years of age.
fWeighted using inverse probability of exposure weighting based on the propensity score. The propensity score was estimated using multivariable logistic regression with 97 previously covariates, as defined in Supplementary Table 2. Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudo sample of participants in the reference group with the same distribution of measured covariates as the exposure group.
In addition, we evaluated whether the risk of IBS increased according to the number of infantile colic episodes (Table 4). Children with both a single diagnosis and more than 1 diagnosis of infantile colic had a significant risk of IBS (HR [95% CI], 1.11 [1.10 to 1.13] and 1.11 [1.07 to 1.15], respectively).
Table 4 . Risk of the Irritable Bowel Syndrome According to the Number of Infantile Colica
| The number of infantile colic history | Unweighted data (n = 368 556) | Weighted data (n = 723 370)b | ||||
|---|---|---|---|---|---|---|
| Subjects (n) | Events of IBS (n [%]) | Subjects (n) | Events of IBS (n [%]) | HR (95% CI)c | ||
| IBS | ||||||
| Referent | Control groupd | 359 769 | 55 887 (15.5) | 363 528 | 56 547 (15.6) | Ref |
| Exposure | 1e | 7756 | 1349 (17.4) | 319 475 | 54 833 (17.2) | 1.11 (1.10 to 1.13) |
| Exposure | ≥ 2e | 1031 | 184 (17.9) | 40 367 | 6925 (17.2) | 1.11 (1.07 to 1.15) |
aThe cohort consists of the infantile colic group who had experienced infantile colic from 5 weeks to 4 months of age and the control group without infantile colic history during the same time period.
bWeighted using inverse probability of exposure weighting based on the propensity score. The propensity score was estimated using multivariable logistic regression with 97 previously covariates, as defined in Supplementary Table 2. Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudo sample of participants in the reference group with the same distribution of measured covariates as the exposure group.
cHazard ratios (HRs) were assessed using a Cox proportional hazards model to examine the relationship between infantile colic history and the risk of irritable bowel syndrome (IBS) in the cohort.
dAs the reference group, the control group comprised children who had not been diagnosed with infantile colic at 5 weeks to 4 months of age.
eInfantile colic group consists of children who have been diagnosed with infantile colic at least once between 5 weeks and 4 months of age.
The sensitivity analyses using various definitions of IBS are shown in Table 5. The risk for the stricter IBS was significant (HR [95% CI], 1.25 [1.23 to 1.28]) in the infantile colic group. For the IBS subtypes for symptoms of diarrhea and constipation, the risks of IBS-D and IBS-C in the infantile colic group were significantly high (HR [95% CI], 1.20 [1.17 to 1.23] and 1.22 [1.1.19 to 1.26], respectively). The risk remained statistically significant for the strict IBS-D and the strict IBS-C in the infantile colic group (HR [95% CI], 1.20 [1.17 to 1.23], and 1.17 [1.07 to 1.28], respectively), which were more stringent defined by their use of drug prescriptions.
Table 5 . Risk of Irritable Bowel Syndrome With Several Definitions of Infantile Colica
| Unweighted data (n = 368 556) | Weighted data (n = 723 370)b | HR (95% CI)e | |||
|---|---|---|---|---|---|
| n (%) | n (%) | ||||
| Control groupc (n = 359 769) | Infantile colic groupd (n = 8787) | Control groupc (n = 363 528) | Infantile colic groupd (n = 359 842) | ||
| Strict IBSf | 34 476 (9.6) | 1099 (12.5) | 34 926 (9.6) | 42 246 (11.7) | 1.25 (1.23 to 1.28) |
| IBS-Dg | 21 203 (5.9) | 699 (7.6) | 21 492 (5.9) | 25 256 (7.0) | 1.20 (1.17 to 1.23) |
| IBS-Ch | 19 184 (5.3) | 625 (7.1) | 19 438 (5.4) | 23 205 (6.5) | 1.22 (1.19 to 1.26) |
| Strict IBS-Di | 21 176 (5.9) | 668 (7.6) | 21 465 (5.9) | 25 207 (7.0) | 1.20 (1.17 to 1.23) |
| Strict IBS-Cj | 1689 (0.5) | 58 (0.7) | 1715 (0.5) | 1958 (0.5) | 1.17 (1.07 to 1.28) |
aThe cohort consisted of the infantile colic group who had experienced infantile colic from 5 weeks to 4 months of age and the control group without infantile colic histories during the same time period.
bWeighted using inverse probability of exposure weighting based on the propensity score. The propensity score was estimated using multivariable logistic regression with 97 previously covariates, as defined in Supplementary Table 2. Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudo sample of participants in the reference group with the same distribution of measured covariates as the exposure group.
cAs the reference group, the control group comprises children who had not been diagnosed with infantile colic at 5 weeks to 4 months of age.
dInfantile colic group consists of children who have been diagnosed with infantile colic at least once between 5 weeks and 4 months of age.
eThe hazard ratios (HRs) were assessed using a Cox proportional hazards model to examine the relationship between infantile colic histories and the risk of irritable bowel syndrome (IBS) development in the cohort.
fDefined with at least 1 diagnosis of International Classification of Diseases 10th version (ICD-10) code K58.X (IBS) and R10.X (abdominal pain) and at least 1 of ICD-10 code K59.1 (functional diarrhea), K52.2 (allergic and dietetic gastroenteritis and colitis), K52.8 (other specific noninfective gastroenteritis), or K59.0 (constipation) after the age of 4 years.
gDefined with at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 diagnosis of ICD code K59.1 (functional diarrhea) after 4 years of age.
hDefined with at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 diagnosis of ICD code K59.0 (constipation) after 4 years of age.
iDefined with at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 diagnosis of ICD code K59.1 (functional diarrhea) and at least once of drug classification 237 (antidiarrheal drug) after 4 years of age.
jDefined with at least 1 diagnosis of ICD-10 code K58.X (IBS) and R10.X (abdominal pain) and at least 1 diagnosis of ICD code K59.0 (constipation) and at least once of drug classification 238 (laxative drug) after 4 years of age.
IBS-D, diarrhea-predominant IBS; IBS-C, constipation-predominant IBS.
The subgroup sensitivity analyses were conducted to assess the robustness of the results (Fig. 2 and Supplementary Table 6). First, when divided into subgroups by the type of feeding during the 4 to 6 months age range, children with former infantile colic had a higher risk of IBS regardless of the type of feeding (HR [95% CI], 1.15 [1.12 to 1.17] in the breastfeeding group and 1.09 [1.06 to 1.11] in the no breastfeeding group), than children without infantile colic history, which was consistent with the main results. Second, all children with birth weights under 3.26 kg (median birth weight) and not less than 3.26 kg in the infantile colic group had a significant risk of IBS (HR [95% CI], 1.14 [1.11 to 1.16] and 1.09 [1.01 to 1.11], respectively). Third, the risk of IBS remained consistently significant in the infantile colic group even after dividing the subgroup according to sex (HR [95% CI], 1.15 [1.12 to 1.17] in male and 1.08 [1.05 to 1.10] in female), and economic status (HR [95% CI], 1.08 [1.06 to 1.11] in the low economic group and 1.16 [1.13 to 1.19] in the high economic group).
Figure 2. Subgroup analysis for the risk of irritable bowel syndrome (IBS) with former infantile colic. IBS was defined as more than twice the diagnosis of International Classification of Diseases 10th version codes K58.X (IBS) after 4 years of age. Data on whether children breastfed or not were acquired from the first round of the national health screening program for infants and children, which was performed at the age of 4 to 6 months. The median birthweight of participants was 3.26 kg. Economic status was defined as the insurance premium acquired from the National Health Insurance Service. Participants in the reference group were weighted as (propensity score/[1-propensity score]). This method produces a weighted pseudosample of participants in the reference group with the same distribution of measured covariates as the exposure group. Filled squares indicate hazard ratios, and straight lines indicate 95% CI. The hazard ratios were assessed using a Cox proportional hazards model to examine the relationship between infantile colic history and the risk of IBS development in the cohort.
In addition, we performed subgroup analyses using the strict IBS definition (Supplementary Table 7). The results regarding the risk of strict IBS with infantile colic were consistently significant in all subgroups.
Our large national administrative cohort showed that infantile colic in young infants may be an early clinical manifestation of functional gastrointestinal disease. Children with a history of infantile colic at the age of 4 weeks to 5 months had a statistically significant risk for the development of IBS after 4 years of age. In the subgroup analyses according to the type of feeding, birth weight, sex, and economic status, the association of infantile colic with IBS, remained consistent. Furthermore, in the sensitivity analyses using the different definitions of IBS, the associations between infantile colic and IBS remain significant.
A few studies have reported the association of infantile colic with the onset of IBS. In a randomized, double-blind, prospective study in Finland (n = 74), children with former colic-type crying during the seventh week of life had more developed functional gastrointestinal diseases by the age of 13 years, than children without former colic-type crying (28.0% versus 5.6%, respectively,
Although our results do not prove that the 2 diseases have common pathogeneses, several studies support our conclusions. Aberrant gut microbiota and low-grade inflammation may explain the link between infantile colic and IBS. Intestinal microbiota of infants with infantile colic were shown to consist of a lower abundance of Bifidobacteria and a higher abundance of Proteobacteria.2,23 Children with IBS had a 1.5-fold higher concentration of
Psychosocial factors play a role in the development of IBS and infantile colic. Children’s personalities, with difficult temperaments such as irritability and hypersensitivities and who experience parental anxieties or depression, have been thought to be risk factors for development of infantile colic.33 In addition, poor emotion regulation and parental concerns about the child were associated with a steeper increase in the prevalence of abdominal pain.34 We suggest that children with these risk personalities may develop clinical manifestations of infantile colic in young infants and IBS in children.
However, there were some limitations to our study. First, we relied on the definitions of infantile colic and IBS according to the ICD-10 codes and the drug classification codes from the administrative data. Therefore, we attempted to use the definition of IBS that validated the diagnostic yield according to previous studies16,17 and we performed additional subgroup and sensitivity analyses. Second, this was an administrative cohort that may have had unmeasured confounders. Although IPTW was used for the demographic/clinical characteristics as well as the results of the physical examinations to match the 2 groups to baseline health status and reduce potential confounding factors, the inflated numbers in the weighted data could increase the type 1 error, and our results cannot infer any causal relationship. Therefore, we tried to perform the stabilized and trimmed IPTW and 4:1 PS matching to confirm the consistency of results and minimize the type 1 error.
An important strength of this study was that it enrolled all children born between 2008 and 2009, and the data in this study had representative characteristics and the results can be generalized. Furthermore, to reduce the various potential confounding factors between the 2 groups, including demographic characteristics and clinical baseline conditions, 97 covariates were matched using PS matching. In addition, we accessed the results of the physical examinations by the physicians and the nutritional status to balance the clinical baseline conditions between the 2 groups. Finally, to verify our results, several subgroup and sensitivity analyses were performed, and all results remained consistently significant.
In conclusion, we found that children with infantile colic histories in the early infant period had a significant risk for later IBS development after 4 years of age. In particular, IBS is a representative form of a functional gastrointestinal disease that extends into the adult period. Efforts to uncover and solve the pathogenesis of infantile colic in the neonatal period may reduce the prevalence and severity of FGID from childhood to adolescence to adulthood.
Note: To access the supplementary material and tables mentioned in this article, visit the online version of
This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NRF2020R1F1A1076452). The funders had no role in the study design, data collection, data analysis, data interpretation, or writing of the report.
None.
Ju Hee Kim, Seung Won Lee, Su Jin Jeong, and Man Yong Han conceptualized and designed this study; Hye Ryeong Cha, Seung Won Lee, Jaewoo An, and Man Yong Han were involved in data collection, data analysis, and interpretation; Ju Hee Kim, Eun Kyo Ha, and Yoowon Kwon were involved in manuscript writing and revision; and Ju Hee Kim, Yoowon Kwon, Eun Kyo Ha, and Su Jin Jeong supervised the data interpretation and critically reviewed the manuscript for important intellectual content. All authors read and approved the final manuscript.
Author ORCID Information
Ju Hee Kim Seung Won Lee Su Jin Jeong Man Yong Han |
 |
