Evaluating the role of the Bristol stool chart in diagnosing functional constipation in pediatric populations

Abstract

BACKGROUND

Functional constipation (FC) is a common condition among children worldwide. Although the Rome IV criteria are diagnostic standard, evaluating infants and toddlers can be challenging. The Bristol stool chart (BSC) is a simple tool for describing stool consistency, but its diagnostic accuracy in young children is uncertain.

AIM

To compare the diagnostic performance of the BSC and Rome IV criteria in children under four years.

METHODS

A prospective cross-sectional study was conducted from June 2025 to October 2025 among children aged 6 months to 4 years attending well-child visits at the Hospital. Parents completed validated Thai versions of the Rome IV questionnaire and the BSC. FC was diagnosed using the Rome IV criteria. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the BSC were calculated.

RESULTS

Among 275 children enrolled, 43 (15.6%) met the Rome IV criteria for FC. The FC group had a significantly higher mean age than the non-FC group (30.1 ± 11.9 months vs 21.2 ± 12.2 months, P < 0.001). A family history of constipation was associated with increased FC prevalence (P = 0.017). On the BSC, stool types 1-2 indicated constipation. Compared with the Rome IV criteria, the BSC demonstrated sensitivity of 48.8%, specificity of 98.3%, PPV of 84.0%, and NPV of 91.2%.

CONCLUSION

The BSC shows high specificity but low sensitivity in young children; stool types 1-2 support FC and are best used to confirm diagnosis alongside the Rome IV criteria.

Key Words: Children; Functional constipation; Bristol stool chart; Pediatric; Prevalence

Core Tip: Stool assessment tools are essential for evaluating pediatric constipation. The Bristol stool chart (BSC) is widely used but has limitations when applied to young children. BSC use should be integrated with clinical judgment and other diagnostic criteria, such as the Rome IV criteria. The BSC exhibits the highest accuracy when classifying stools at the extremes. BSC types 1-2 may serve as supportive indicators for confirming functional constipation.

INTRODUCTION

Constipation is a common global health problem that significantly diminishes quality of life and imposes a considerable burden on healthcare systems[1]. Children with constipation frequently present to general practitioners and pediatricians, contributing substantially to healthcare costs[2]. Functional constipation (FC) reported prevalence rates show substantial geographic variation. In North and South America, the prevalence has been estimated to range from approximately 10% to 23% across pediatric populations, including infants and adolescents[3,4]. In contrast, pooled prevalence estimates in Europe are lower, reported at 6.9% and 8.17% among children aged 0-4 years old[5].

In Asia, prevalence estimates differ widely, from 0.7% to 29.6%[6-10]. Recent studies in Thailand show that the prevalence is 2.4% using the Rome III criteria[11] and 8.1% in teenagers using the Rome IV criteria[12]. Globally, pooled research shows that FC affects around 9.5% of children[10], with prevalence ranging from 0.5% to 32.2%[6], depending on the age range investigated. The most common age range is 2 years to 4 years, which commonly coincides with the era of toilet training. However, constipation may develop as early as infancy, with reported rates in this age group ranging between 17% and 40%[13]. FC is diagnosed in children using the Rome criteria, which are regarded as the diagnostic gold standard[14]. A systematic analysis found that the median prevalence of childhood constipation is 12%, with neonatal incidence ranging from 0.05% to 39.3%. Given the significant impact of FC on healthcare utilization, the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition collaborated in 2014 to publish evidence-based guidelines for the evaluation and treatment of childhood constipation[1]. Since then, diagnostic procedures have advanced, particularly with the establishment of the Rome IV criteria in 2016, which have since been used worldwide[1,14,15]. In Thailand, validation studies have demonstrated the reliability of the Thai versions of the Rome III and IV questionnaires for diagnosing functional gastrointestinal disorders[12,16-18], including FC, across various pediatric age groups. However, obtaining an accurate clinical history in infants and toddlers remains challenging. Therefore, parental observation of stool characteristics is critical for diagnosis. To assist in stool assessment, several tools have been developed, including the modified Bristol Stool Scale, the Amsterdam Infant Stool Scale (AISS), the Brussels Infant and Toddler Stool Scale (BITSS), and the 3-D stool assessment instrument[16]. Despite these advances, the Bristol stool chart (BSC) remains the most widely used tool worldwide due to its simplicity and accessibility. It categorizes stools into seven sorts: Types 1-2 indicate constipation, types 3-5 are normal stools, and types 6-7 indicate diarrhea. Nonetheless, information on the BSC diagnostic accuracy in young children is sparse. While most studies use the BSC to define “hard stool” for comparison with the Rome IV criteria, its sensitivity and specificity specifically in this population remain unclear. As a result, the purpose of this study was to compare the BSC’s diagnostic accuracy to the Rome IV criteria for children under the age of four. We hypothesized that the BSC may serve as a useful adjunctive tool for diagnosing FC in early childhood.

MATERIALS AND METHODS

Study design and participants

A prospective, descriptive cross-sectional study was undertaken at the Hospital’s outpatient well-child clinic between June 2025 and October 2025. Eligible participants were children aged 6 months to 4 years who attended routine well-child visits during the study period. Children with known structural gastrointestinal abnormalities, neuromuscular disorders, thyroid disease, or developmental delay were excluded from the study. This study used a fixed sample derived from all eligible children attending the well-child clinic during the study period. Therefore, no sample size calculation was performed (Figure 1). The study protocol was approved by the Institutional Review Board of Naresuan University Hospital (IRB No. P3-0034/2568). The study was explained in detail to all parents. Before enrollment, with the completion of the questionnaire, written informed consent was obtained.

Figure 1 Flowchart of study enrollment. PPV: Positive predictive value; NPV: Negative predictive value; LR: Likelihood-ratio; CNS: Central nervous system; GI: Gastrointestinal.

Questionnaires: Parents completed two instruments: (1) The validated Thai version of the Rome IV diagnostic questionnaire for children under the age of four; and (2) The BSC. The Thai Rome IV questionnaire, translated and validated by the Rome Foundation, has good content validity (index of item-objective congruence = 0.74), internal consistency (Cronbach’s α = 0.750), and test-retest reliability [intraclass correlation coefficient = 0.807, 95% confidence interval (CI): 0.694-0.881][17]. FC was identified using the Rome IV criteria (Table 1). The BSC divides stools into seven categories: Types 1-2 indicate constipation, types 3-5 indicate normal stool, and types 6-7 indicate diarrhea (Figure 2)[19,20]. The BSC was created by Lewis and Heaton[20].

Figure 2 Bristol stool chart. Type 1 separate hard lumps, like nuts, type 2 sausage-shaped but lumpy, type 3 like a sausage but with cracks on the surface, type 4 like a sausage or snake, smooth and soft, type 5 soft blobs with clear-cut edges, type 6 fluffy pieces with ragged edges, a mushy stool, type 7 watery, no solid pieces.

Table 1 Rome IV criteria.

Rome IV criteria for pediatric functional constipation
From infants to 4-year-old children
A child with FC must have 1 month of at least 2 of the following or 2 or fewer stooling a week and a history of
Excessive stool retention
Painful or hard stooling
Large-diameter stools
Large fecal mass in the rectum
In children trained in the toilet, additional points are added to the criteria
At least one episode per week of stool incontinence after being trained in toileting
History of large-diameter stools that may obstruct the toilet

Rome IV criteria; fifth edition; Rome foundation. FC: Functional constipation.

Data collection

Demographic and clinical information, including age, sex, caregiver, family history, socioeconomic background, and bowel habits, was collected using the questionnaires. Stool form was simultaneously assessed using the BSC (Supplementary Figure 1).

Statistical analysis

The data were analyzed using STATA version 18, including a two-sample comparison of proportions with a power of 0.8, an alpha level of 0.05, and a group ratio of 2:1 (resulting in a total of at least 21 children). However, this study used a fixed sample comprising all eligible children attending the well-child clinic during the study period. Therefore, the power calculation was not intended to determine the sample size a priori but is reported for transparency in the statistical approach. Descriptive statistics were used to analyze demographic factors. The BSC sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were determined using the Rome IV questionnaire as a reference standard. 95%CI were supplied. Fisher’s exact test was employed for categorical variables and independent t-tests for continuous variables for comparing proportions. P values < 0.05 were considered statistically significant.

RESULTS

Participant characteristics

The study included 275 youngsters. Of them, 43 (15.6%) satisfied the Rome IV criteria for FC, while 232 (84.4%) were classed as non-FC. The demographic and clinical data are summarized in Table 2. The mean age of the FC group was significantly higher than the non-FC group (30.1 ± 11.9 months vs 21.2 ± 12.2 months; P < 0.001). There were no significant differences between the groups in terms of gender, body weight, family structure, parental education, or household income. However, a positive family history of constipation among first-degree relatives was significantly higher in the FC group (27.9%; P = 0.017). In subgroup analyses, 131 children aged 6-18 months (12 with FC) and 144 children aged 19-48 months (31 with FC) were analyzed. In both age groups, children with FC were older than those non-FC (6-18 months: 15.0 ± 2.7 months vs 11.3 ± 3.6 months, P = 0.0008; 19-48 months: 36.0 ± 8.2 months vs 31.7 ± 8.9 months, P = 0.0152). A first-degree family history of constipation was associated with FC in the 6-18 months group (P = 0.009) but not in the 19-48 months group (P = 0.413).

Table 2 Demographic data and family status, n (%).

Data	Functional constipation (n = 43)	Non-functional constipation (n = 232)	P valve
Age, month (mean ± SD)	30.1 ± 11.9	21.2 ± 12.2	< 0.001a
Male	19 (44.2)	102 (44.0)	1.000
Body weight, kg (mean ± SD)	12.8 ± 3.8	11.3 ± 3.2	0.009a
Single family	17 (39.6)	77 (33.2)	0.484
Main caregiver: Mother	30 (69.8)	153 (66.0)	0.507
Main caregiver educational level			0.248
Mandatory	14 (32.56)	95 (40.96)
Bechalor	26 (60.47)	108 (46.55)
> Bechalor	3 (6.98)	29 (12.50)
Family status			0.794
Together	38 (88.4)	207 (89.2)
Separated	5 (11.6)	25 (10.8)
Family income			0.186
Low income (< 30000 baht/month)	16 (37.2)	113 (48.7)
First-degree relative with a history of constipation	12 (27.9)	29 (12.5)	0.017a

^aP < 0.05.

Parental stool reports

Among children diagnosed with FC, parental reports of stool consistency were as follows: 21 (48.8%) reported hard stools, 12 (27.9%) reported normal stools, 2 (4.7%) reported soft or mucous/liquid stools, and 8 (18.6%) reported alternating mucous and liquid. When assessed using the BSC, 21 (48.8%) were classified as types 1-2, 14 (32.6%) as types 3-5, and 8 (18.6%) as types 6-7 (Table 3).

Table 3 Results from the questionnaire on bowel habits in the Rome IV criteria, n (%).

Description	Total (n = 275)	Functional constipation (n = 43)	Non-functional constipation (n = 232)
Bristol stool chart	275 (100)
1-2 (hard)	25 (9.1)	21 (48.8)	4 (1.7)
3-5 (normal)	215 (78.2)	14 (32.6)	201 (86.6)
6-7 (loose/liquid)	35 (12.7)	8 (18.6)	27 (11.6)
Defecation frequency	275 (100)
< 3 × /week	7 (2.5)	5 (11.6)	2 (0.9)
3-6 × /week	75 (27.3)	24 (55.8)	51 (22)
Daily	118 (43)	10 (23.3)	108 (46.6)
2-3 × /day	71 (25.8)	4 (9.3)	67 (28.9)
> 3 × /day	4 (1.5)	0	4 (1.7)
Stool consistency	275 (100)
Hard	25 (9.1)	21 (48.8)	4 (1.7)
Normal	146 (53.1)	12 (27.9)	134 (57.8)
Soft	65 (23.6)	2 (4.7)	63 (27.2)
Mucous	4 (1.5)	0	4 (1.7)
Liquid	0	0	0
Alternating mucous and liquid	35 (12.7)	8 (18.6)	27 (11.6)
Painful defecation	266 (100)
Yes	45 (16.9)	32 (76.2)	13 (5.8)
No	221 (83)	10 (23.8)	211 (94.2)
Large diameter stool	275 (100)
Yes	35 (12.7)	30 (70)	5 (2.2)
No	240 (87.3)	13 (30.2)	227 (97.8)
Withholding behavior	263 (100)
Never	220 (83.7)	12 (32.4)	208 (92)
1-3 times/month	24 (9.1)	13 (35.1)	11 (4.9)
> 1 time/week	19 (7.2)	12 (32.4)	7 (3.1)
Fecal incontinence	91 (100)
Never	66 (72.5)	11 (50)	55 (79.7)
< 1 time/month	16 (17.6)	5 (22.7)	11 (16.0)
1-3 times/month	8 (8.8)	5 (22.7)	3 (4.4)
> 1 time/week	1 (1.1)	1 (4.6)	0
Stools obstruct the toilet	91 (100)
Yes	3 (3.3)	2 (9.1)	1 (6.5)
No	88 (96.7)	20 (90.9)	68 (98.6)

Diagnostic accuracy of the BSC

Using the Rome IV questionnaire as the reference standard, the BSC demonstrated sensitivity of 48.8% (95%CI: 33.3-64.5) and specificity of 98.3% (95%CI: 95.6-99.5). The PPV was 84.0% (95%CI: 63.9-95.5), and the NPV was 91.2% (95%CI: 87.0-94.4) (Supplementary Figure 1). This indicates that when the BSC yields a positive result suggestive of FC, there is an 84.0% PPV that the child truly meets the Rome IV diagnostic criteria for FC. The full screening characteristics, including a positive likelihood ratio of 28.3 and a negative likelihood ratio of 0.5, are provided in Table 4. A positive likelihood ratio of 28.3 suggests that children with FC are 28 times more likely to have a positive BSC result than children without the condition, indicating a strong rule-in value when the BSC is positive because a positive likelihood ratio greater than 10 is generally considered to provide strong evidence to rule in a diagnosis. The area under the curve for BSC in the diagnosis of FC using the Rome IV criteria as the reference standard was 0.74 (95%CI: 0.66-0.81) (Figure 3).

Figure 3 Diagnosis of functional constipation using the Rome IV criteria as the reference standard, with calculation of the area under the curve for Bristol stool chart. AuROC: Area under the receiver operating characteristic; CI: Confidence interval.

Table 4 Screening characteristics of the Bristol stool chart compared to Rome IV functional constipation.

Parameter	Result	95%CI
Sensitivity (%)	48.8	33.3-64.5
Specificity (%)	98.3	95.6-99.5
Positive predictive value (%)	84	63.9-95.5
Negative predictive value (%)	91.2	87.0-94.4
Positive likelihood ratio	28.3	10.2-78.4
Negative likelihood ratio	0.5	0.4-0.7

CI: Confidence interval.

DISCUSSION

In this study, the prevalence of FC among children under 4 years of age was 15.6%, which aligns with previous studies in Asia, including a reported prevalence of 10.6% in China[8]. These results suggest that FC is a significant and representative health problem in this pediatric age range. The mean age of children with FC in our cohort was 30.1 months, indicating that early screening for constipation may be particularly beneficial beginning around 18.2 months of age. Furthermore, our findings indicate that a positive family history of constipation is associated with an increased risk of FC, supporting a role of familial factors, potentially including genetic susceptibility, in its pathogenesis. The significant association observed in the 6-18 months age subgroup may reflect inherited vulnerability manifesting early in life. A previous study in preschool-aged children reported a 3.788-fold higher risk of FC among those with parental histories of the condition[21]. Prior studies have also demonstrated familial aggregation of childhood FC, suggesting a heritable component, possibly mediated by inherited differences in colonic motility and enteric nervous system function[22]. Genetic influences may be more prominent in early childhood, whereas behavioral and environmental factors may become increasingly relevant with age[21,22]. The BSC is widely used as a practical tool for evaluating stool consistency due to its simplicity, clarity, and ease of interpretation by both parents and caregivers. In this study, the BSC demonstrated very high specificity (98.3%) and a strong positive likelihood ratio (28.33), supporting its utility as a confirmatory tool for FC diagnosis. However, the sensitivity was relatively low (48.8%), which limits its effectiveness as a screening instrument. In practical terms, BSC types 1-2 can be considered reliable indicators of FC, whereas types 3-7 cannot exclude the condition and therefore warrant further evaluation using the Rome IV criteria. Previous studies have compared the BSC with various alternative stool assessment instruments. The AISS, for instance, incorporates stool color and volume but is considered less practical for toilet-trained children[23]. The BITSS[19], which was specifically developed for diaper-wearing children, has demonstrated greater sensitivity in detecting FC than the BSC. Additionally, modified versions of the BSC have been tested, with some evidence suggesting higher sensitivity, particularly in older children[14,24,25]. Nevertheless, despite the existence of these alternatives, the BSC remains the most widely applied and accessible tool in clinical practice.

Our findings reinforce the clinical value of the BSC as a supportive diagnostic tool rather than a stand-alone screening in pediatric populations. While its high specificity enhances its diagnostic confidence in identifying FC, its low sensitivity underscores the need for additional confirmatory assessments, such as the Rome IV questionnaire. Moreover, diagnostic accuracy could potentially improve if the BSC were supplemented with descriptive criteria or adapted for infants and toddlers, where stool characteristics are highly variable and often reported subjectively by caregivers.

CONCLUSION

Although several stool assessment instruments are available for pediatric use, the BSC remains the most widely utilized due to its inherent simplicity and accessibility. This study demonstrated that, when compared with the Rome IV criteria, the BSC exhibits very high specificity but limited sensitivity for diagnosing FC in children under four years of age. Specifically, BSC types 1-2 serve as supportive indicators of FC, while types 3-7 cannot reliably exclude the diagnosis. Therefore, the BSC should not be applied as a stand-alone screening tool but may function effectively as a confirmatory instrument, particularly in cases with strong clinical suspicion. Combining the BSC with the Rome IV criteria enhances diagnostic accuracy and strengthens clinical decision-making in pediatric constipation assessment. Future research should focus on modifying stool assessment scales tailored for infants and toddlers to improve diagnostic sensitivity while maintaining practicality in both clinical and community settings.

ACKNOWLEDGEMENTS

The authors would like to thank Mr. Sagoontee Inkate for assistance with the statistical analysis and Ms. Daisy Jimenez Gonzales for English language review. We also extend our gratitude to the children and nursing staff at the Pediatric Outpatient Department of Naresuan University Hospital, Naresuan University, for their support of this study.