Measuring health-related quality of life and well-being in adults with fecal incontinence: a comparative psychometric evaluation Authors: Sylvester R Groen, Daniel Keszthelyi, Zsa Zsa R M Weerts, Sadé L Assmann, Stéphanie O Breukink, Brigitte A B Essers Published: 2026 ABSTRACT AbstractBackgroundWhile traditional disease-management regarding fecal incontinence (FI) primarily focuses on symptom-control and health-related quality of life (HrQoL) outcomes, patients often prioritize broader aspects of health, like well-being. This study evaluates the psychometric performance of ICEpop CAPability measure for Adults (ICECAP-A), a capability-based well-being instrument, alongside two established HrQoL-instruments, EQ-5D-5L and the FI quality of life (FIQL) questionnaire, in an outpatient-population with FI.MethodsData were obtained from patients with FI recruited from a randomized-controlled trial evaluating the efficacy of anal plugs for FI ((The CONFIDEnCE trial, NCT04657588). Psychometric evaluation of ICECAP-A involved the assessment of the convergent validity, known-group validity (examined across predefined subgroups), acceptability, and floor and ceiling effects.ResultsAmong 72 participants (mean age 67.3years; 83.3% female), the mean ICECAP-A capability score (0.86, scale 0–1) exceeded the mean EQ-5D-5L utility (0.71, scale − 0.446–1) and the FIQL score (2.34, scale 1–5). ICECAP-A showed good convergent validity (76.6% predefined hypotheses confirmed). Known-group validity was comparable across instruments; ICECAP-A demonstrating significant differences by EQ-VAS, anxiety, and education level, but not outperforming EQ-5D-5L or FIQL. EQ-5D-5L showed stronger effect sizes for anxiety (p < 0.001), depression and symptom severity (non-significant), while FIQL was stronger for education level (p = 0.002). Questionnaire completion was 84.7% overall, reflecting trial dropout only (100% among completers) with no substantial floor or ceiling effects.ConclusionsIn patients with FI, ICECAP-A showed psychometric performance comparable to EQ-5D-5L and FIQL, despite its broader conceptual scope. Well-being appeared less influenced by symptom severity, possibly reflecting coping or adaptation. Further research is warranted evaluating ICECAP-A’s psychometric advantage over traditional HrQoL-instruments in this clinical context.Trial registrationNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12955-026-02566-0.Keywords: Fecal incontinence, Well-being, Quality of life, ICECAP-A, Euroqol, FIQL Introduction Fecal incontinence (FI) is a chronic anorectal disorder, defined by the Rome IV criteria as the involuntary loss of fecal material for a duration of at least three months [1]. With an estimated global prevalence of 8% among community-dwelling adults, FI is likely under-reported due to patient embarrassment and limited clinical inquiry [2, 3]. Although it may result from a single major anorectal insult, FI often arises from multifactorial causes, associated with ageing, obstetric trauma, menopause, colorectal inflammation, neurological disorders, obesity, and chronic comorbidities [4, 5]. The pathophysiology involves a complex interaction between anatomical and functional integrity of the anal sphincter, anorectal sensation, neuromuscular coordination, and stool consistency [6, 7]. Beyond physical symptoms, FI exerts a profound psychosocial burden, often leading to embarrassment, anxiety, diminished self-esteem, activity avoidance, and social isolation. Its impact on health-related quality of life (HrQoL) and overall well-being is substantial, with many patients experiencing significant psychological distress [8, 9]. A wide range of treatment options exists for fecal incontinence, yet complete symptom resolution is achieved in only a minority of patients [10]. Treatment success is closely linked to individual experiences, specific symptom burdens, and HrQoL. While clinical management traditionally emphasizes symptom reduction and physical outcomes, emerging evidence suggests that patients often prioritize restoration of normalcy, independence, emotional well-being, and a sense of security [11]. Consequently, effective FI management should extend beyond physiological parameters to incorporate psychosocial dimensions such as fear, activity avoidance, and overall life enjoyment [12, 13]. Although HrQoL measures aim to capture these aspects, commonly used tools such as the validated generic 5 level EQ-5D questionnaire introduced by the Euroquol group (EQ-5D-5L) and the condition-specific FI quality of life (FIQL) questionnaire [14–16] tend to focus primarily on physical health, potentially overlooking broader domains relevant to mental, social, and public health outcomes. A shift toward a more comprehensive evaluation of well-being appears warranted [17]. The ICEpop CAPability measure for Adults (ICECAP-A) assesses quality of life through the lens of individual capabilities (what an individual can do) rather than function (what they actually do), focusing on well-being in a broader sense rather than health [18]. When used alongside HrQoL measures, the ICECAP-A may provide a more comprehensive and patient-centered assessment by capturing broader aspects of well-being that are often overlooked, yet highly valued by individuals living with FI [19]. To date, several validated translations of the ICECAP-A questionnaire are available, and its psychometric properties have been increasingly studied across various populations and health conditions [20, 21]. However, its applicability in the context of FI has not yet been explored. In particular, no studies have assessed the ICECAP-A in parallel with established HrQoL measures in this patient group. Although the capability-based measure ICECAP-A and HrQoL instruments are conceptually distinct, some subscales across the instruments may overlap. We therefore hypothesized that the strength of associations between the measures would vary according to the degree of conceptual overlap between their respective subscales. Therefore, the aim of this study is twofold: [1] to examine the construct validity, specifically convergent and known-group validity, of the ICECAP-A in comparison with the EQ-5D-5L and FIQL questionnaires; and [2] to assess its acceptability together with its floor and ceiling effects in an outpatient population of individuals with FI. Methods Study design and population Data were collected from participants involved in a randomized controlled trial (RCT) on the efficacy of an anal plug compared to care as usual for treating FI (The CONFIDEnCE trial, NCT04657588). Baseline measurements were considered for testing validity. The duration of the treatment period was 8 weeks. Follow-up measurements after treatment were considered for testing acceptability and floor/ceiling effects. Participants were recruited between 2020 and 2024 in a secondary care hospital (Rijnstate hospital, Arnhem) and a secondary/tertiary care hospital (Maastricht University Medical Center +, Maastricht) in the Netherlands. Patients (aged 16–90 years) who met the Rome IV criteria for FI were eligible for inclusion in the study. Patients could be at any stage of the treatment pathway for FI, ranging from those starting basic options (e.g. anti-diarrhea medication, or pelvic floor therapy) to those who had already undergone more invasive treatments including surgery. Patients were excluded in case of Inflammatory bowel disease (IBD), pregnancy, anorectal anatomical abnormalities such as fistulas, congenital anorectal malformations, previous anorectal surgery, or radiotherapy to a pelvic organ, and insufficient cognitive function to complete questionnaires. The study protocol was approved by the Maastricht University Medical Center+ (MUMC+) ethics committee and was applicable to both participating hospitals. All study procedures were performed in compliance with good clinical practice guidelines and according to the revised Declaration of Helsinki. Patient inclusion was completed after receiving a written informed consent before participation. All authors had access to the study data and reviewed and approved the final manuscript. Data collection All participants completed digital questionnaires at two time points: at baseline prior to randomization (T0) and one week after completion of the 8-week study period (T1). In addition to collecting demographic data, questionnaires assessing symptom severity, condition-specific and generic HrQoL, overall well-being, and mental health, were used. Demographics Extracted information on demographics was: age in years at baseline, sex, body mass index (BMI), employment status, and education level (ranging from ‘no education’ to ‘university’). ICECAP-A The ICECAP-A is designed to assess an individual’s capabilities—what they are able to do—rather than their actual functioning, thereby highlighting the freedom to choose. It includes five dimensions or subscales: stability, attachment, autonomy, achievement, and enjoyment, each rated on a four-point scale (1 = unable to achieve; 4 = fully able). Scores were then converted to capability values using the Dutch tariff, ranging from 0 (no capability) to 1 (full capability) [18, 19] . EQ-5D-5L Generic HrQoL was assessed using the EQ-5D-5L-questionnaire, which comprises five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is rated on a five-point scale. In addition, the EQ visual analogue scale (EQ-VAS) captures the participant’s self-rated overall health on a scale from 0 (worst imaginable health state) to 100 (best imaginable health state). Dutch utility index scores were calculated for each participant, ranging from 1 (perfect health) to − 0.446 (worst possible health state) [14, 22]. FIQL Condition-specific HrQoL was assessed using the Fecal Incontinence Quality of Life (FIQL)-scale. This measure consists of 29 questions, subdivided into 4 subscales: lifestyle, coping/behavior, depression/self-perception, and embarrassment, each rated on a five-point scale (1 = strongly agree; 5 = strongly disagree) [15, 16]. Apart from mean subscale scores a total mean score was calculated. St. Marks (Vaizey) Clinical symptom severity was assessed using the St. Mark’s (Vaizey) severity scale for FI. This questionnaire includes three questions regarding FI and fecal consistency (solid or liquid stools, gas), and one question regarding lifestyle alterations due to FI (range 0–4; 0 = never; 4 = daily). Furthermore, additional questions on incontinence material and constipation medicine utilization, and the ability to defer stools are included. A total severity score can be calculated (range 0–24; 0 perfect consistency and no incontinence; 24 = severe incontinence) [23]. GAD-7 and PHQ-9 Comorbid symptoms of anxiety and depression were assessed using the generalized anxiety disorder-7 (GAD-7; Likert scale 0–7; ≥10 moderate anxiety; ≥15 severe anxiety) and the patient health questionnaire-9 (PHQ-9; Likert scale 1–4; ≥10 moderate depression; ≥15 severe depression) [24, 25]. Both questionnaires are validated for application in a Dutch population [26]. Statistical analysis Statistical analysis was performed using SPSS 28.0 (IBM, USA) and GraphPad Prism 8.0.2 (GraphPad Software, Inc, USA). Continuous data is presented as means and interquartile range (IQR), categorical data as proportions (%). A p-value of < 0.05 was deemed statistically significant. Psychometric evaluation was conducted in accordance with the COSMIN (COnsensus-based Standards for the selection of health Measurement INstruments) taxonomy of measurement properties [27]. Within this framework, we assessed construct validity (including convergent and known-group validity) and interpretability (floor and ceiling effects) of the ICECAP-A in comparison with the EQ-5D-5L and FIQL. Internal consistency was not evaluated, as the ICECAP-A comprises five conceptually distinct subscales (stability, attachment, autonomy, achievement, and enjoyment) that represent separate dimensions of capability well-being rather than reflective indicators of a single latent construct. Consequently, the instrument can be considered formative or multidimensional in nature. In accordance with the COSMIN guidelines, the assessment of internal consistency was therefore not considered applicable [28]. Responsiveness was not assessed due to the absence of an appropriate external anchor or validated criterion to determine clinically meaningful change over time. Convergent validity Convergent validity was assessed using Spearman’s rank correlation coefficients. Correlation coefficients were interpreted as follows: absent (< 0.20); weak (0.20–0.34); moderate (0.35–0.49); and strong (≥ 0.50) [29]. Analyses included subscale scores and capability values of the ICECAP-A and subscale, utility scores, and EQ-VAS health score of EQ-5D-5L, as well as subscale and total mean scores of the FIQL. Convergent validity was assessed using an a priori hypothesis-driven approach, in line with COSMIN recommendations, specifying the expected direction and magnitude of correlations based on conceptual overlap. For subscales with limited conceptual similarity, weak to moderate correlations were hypothesized, whereas stronger associations were expected where conceptual overlap was greater. A list of all hypotheses on convergent validity can be found in Table S1 of the supplementary information. We hypothesized moderate to strong positive correlations between the capability values of the ICECAP-A and the utility scores of the EQ-5D-5L and EQ-VAS (H1-H2), in which higher HrQoL is expected to be associated with greater overall well-being. Given the ICECAP-A’s intended applicability to individuals with mental problems [30], we expected moderate to strong negative correlations of all ICECAP-A subscales to the anxiety/depression subscale of EQ-5D-5L (H3-H7). Based on prior findings, we further hypothesized a higher score on autonomy subscales of ICECAP-A would be associated with fewer reported problems—and thus lower scores—on the EQ-5D-5L subscales for mobility, self-care, and usual activities [19, 20]. Accordingly, we expected to observe moderate to strong negative correlations between these measures. Furthermore, we anticipated that experiencing problems on the EQ-5D-5L subscales for pain/discomfort and usual activities would be associated with lower scores on the ICECAP-A subscales for achievement and enjoyment. These correlations were expected to manifest as moderate negative correlations (H11-H14). We hypothesized a moderate to strong positive correlation between the ICECAP-A capability score and the FIQL total score (H15), as higher HrQoL is expected to align with greater overall well-being. We expected that experiencing fewer problems—reflected in higher scores—on the FIQL subscales lifestyle, coping/behavior, and depression/self-perception would be associated with increased capability, particularly on the ICECAP-A subscales stability and achievement, however with slight conceptual overlap of these subscales. These associations were therefore expected to result in weak to moderate positive correlations (H15-H21). Due to higher conceptual overlap moderate to strong correlations were expected between FIQL subscales lifestyle, coping/behavior, and depression/self-perception and ICECAP-A subscale enjoyment (H22-24). Given that experiencing embarrassment is often linked to impaired social functioning, we hypothesized a moderate negative correlation between the ICECAP-A subscale autonomy and the FIQL subscale embarrassment (H25). Finally, we expected that fewer problems on the FIQL subscale depression/self-perception would be associated with higher scores on the ICECAP-A subscale autonomy, resulting in a moderate positive correlation (H26). Known-group validity The hypotheses on known-group validity are summarized in Table S2 of the supplementary information. A priori, confirmation of at least 75% of the predefined hypotheses was considered indicative of adequate known-group validity, in line with commonly applied psychometric evaluation criteria. Known-group validity was evaluated by examining the extent to which the questionnaires could differentiate between predefined subgroups, as informed by prior research. First, previous psychometric studies comparing the ICECAP-A and EQ-5D-5L have demonstrated discriminative capacity for both instruments, depending on the outcome and subgroup. These subgroups included EQ-VAS health status (< 65 as poor-intermediate health vs. ≥65 good-perfect health), employment status (no employment vs. employment, education level (lower vs. secondary/tertiary), and mental health status—specifically, the presence or absence of anxiety (GAD-7 < 10 vs. ≥10), and depression (PHQ-9 < 10 vs. ≥10) [19, 20]. In addition, based on existing literature identifying determinants of reduced HrQoL in individuals with FI, we included the following subgroups for further analysis: sex (male vs. female), body mass index (BMI < 25 vs. ≥25), symptom severity measured by the Vaizey score (< 13 vs. ≥13), comorbid irritable bowel syndrome (IBS yes vs. no), and (urogenital or gynecologic) surgery (yes vs. no) [3, 6, 31–33]. In addition, both anxiety and depression are also described as determinants to influence HrQoL outcomes in FI. Group differences in ICECAP-A, EQ-5D-5L, and FIQL scores were assessed using the Mann–Whitney U test (two groups) and Kruskal–Wallis test (≥ three groups), as none of the variables followed a normal distribution. To correct for multiple testing, the Holm–Bonferroni method was applied. Effect sizes (ES) were calculated using Z-values for the Mann–Whitney U test and H-values for the Kruskal–Wallis test. For the Mann–Whitney U, ES were interpreted as small (0.1–0.3), moderate (0.3–0.5), and large (≥ 0.5). For the Kruskal–Wallis test, epsilon squared (ε²) was used, with thresholds of 0.01–0.06 for a weak effect, 0.06–0.14 for a moderate effect, and > 0.14 for a strong effect [34, 35]. Additionally, the standard error of the mean (SEM) of the ICECAP-A capability values was calculated to quantify measurement error in the original unit of the instrument. The SEM was used to contextualize observed between-group differences by indicating the extent to which these differences may attributable to measurement variability. Importantly, the SEM was not interpreted as a threshold for meaningful or clinically important differences, but rather as an estimate of absolute measurement error [36]. In contrast to the the minimal important difference (MID), which is primarily intended for interpreting longitudinal change, and the intraclass correlation coefficient (ICC), which reflects relative reliability, the SEM provides an estimate of absolute measurement error in the original unit of the instrument. This makes it useful for assessing whether observed cross-sectional between-group differences are likely to be distinguishable from measurement noise [37]. The SEM is defined as the square root of the error variance: SEM = √σ2error. Acceptability, floor and ceiling effects In reflection of the perceived relevance of outcomes of the ICECAP-A, EQ-5D-5L, and FIQL, acceptability was assessed through completion rates at T0 and T1 of all three instruments. Rates above 95% indicated high acceptability. In addition, floor and ceiling effects were examined to determine whether the ICECAP-A, EQ-5D-5L, and FIQL adequately captured the full range of variation within the study population. Specifically, score distributions, the prevalence of minimum (floor), and maximum (ceiling) scores, were evaluated for each instrument. A floor or ceiling effect was considered present if ≥ 15% of respondents achieved the lowest or highest possible score, respectively [38]. Sample size The sample size was primarily determined by the clinical efficacy of the anal plug (results to be reported elsewhere). For convergent validity, a correlation of r = 0.35 required a minimum of 62 participants (α = 0.05, 80% power). For known-groups validity, an ES of r = 0.30 translated into 46 participants after applying a 15% inflation to account for the Mann–Whitney U test. Results Between May 2020 and June 2024, 72 patients were enrolled for randomization in the CONFIDEnCE trial. During the 8-week follow-up period, 11 patients discontinued participation: 8 out of 35 in the treatment group and 3 out of 37 in the control group, therefore data from 61 patients was available at follow-up. Baseline characteristics of the study sample can be found in Table 1. The mean age of the study sample was 67.33 years and 83.3% were female. The mean symptom severity score was 14.72 (SD 4.04), 72.2% of the participants had a symptom severity score higher than 13. The mean HrQoL measured by the EQ-5D-5L was 0.71 (SD 0.27), and 2.34 (SD 0.61) measured by the FIQL. Lastly, the mean overall (capability) ICECAP-A score was 0.86 (SD 0.15). Table 1Baseline characteristics study sampleFI (n = 72)Sex, female, n (%)60 (83.3)Age, years, mean (SD)67.33 (9.81) Age ≥ 40 years, n (%)71 (98.6)BMI, mean (SD)27.14 (6.01) BMI ≥ 25, n (%)45 (62.5)FI symptoms onset, years, mean (SD)11.3 (11.7)Urogenital/gynecological surgery, n (%)7 (9.7)Comorbid irritable bowel syndrome, n (%)8 (11.1) Education level Lower, n (%)29 (40.3) Secondary, n (%)21 (29.2) Tertiary, n (%)22 (30.6) Employment status No employment, n (%)27 (37.5) Incapacitated to work, n (%)1 (1.4) Employment, n (%)32 (44.4) Other, n (%)12 (16.6)Symptom severity, Vaizey score, mean (SD)14.72 (4.04) Vaizey score ≥ 13, n (%)52 (72.2)Anxiety symptoms, GAD-7, mean (SD)4.36 (4.52) GAD-7 score ≥ 10, n (%)11 (15.3)Depressive symptoms, PHQ-9, mean (SD)5.44 (4.27) PHQ-9 score ≥ 10, n (%)10 (13.9)EQ-5D-5L, utility score, mean (SD)0.71 (0.27) EQ-VAS health score, mean (SD)68.54 (15.82) EQ-VAS health score ≥ 65, n (%)47 (65.3)FIQL, total mean score, mean (SD)2.34 (0.61)ICECAP-A, capability value, mean (SD)0.86 (0.15)FI, fecal incontinence; BMI, body mass index; GAD-7, generalized anxiety disorder-7 questionnaire; PHQ-9, patient health questionnaire-9; EQ-5D-5L, Euroqol-5D-5 L questionnaire; EQ-VAS, EQ-5D-5L visual analogue scale; FIQL, FI quality of life questionnaire; ICECAP-A, ICEpop CAPabality measure for adults Baseline characteristics study sample FI, fecal incontinence; BMI, body mass index; GAD-7, generalized anxiety disorder-7 questionnaire; PHQ-9, patient health questionnaire-9; EQ-5D-5L, Euroqol-5D-5 L questionnaire; EQ-VAS, EQ-5D-5L visual analogue scale; FIQL, FI quality of life questionnaire; ICECAP-A, ICEpop CAPabality measure for adults Convergent validity Table 2 presents the correlation matrix assessing the convergent validity of the ICECAP-A in relation to the EQ-5D-5L and FIQL. Overall, twenty out of twenty-six predefined hypotheses (76.9%) were confirmed. ICECAP-A capability value showed strong correlations with the EQ-VAS (r = 0.50, 95% CI [0.300, 0.661]) and the FIQL total mean score (r = 0.52, 95% CI [0.319, 0.673]), and a moderate correlation with the EQ-5D-5L utility score (r = 0.41, 95% CI [0.187, 0.588]). Strong correlations were also found between ICECAP-A subscale stability and the anxiety/depression subscale of the EQ-5D-5L (r = -0.53, 95% CI [-0.681, -0.331]), as well as ICECAP-A’s enjoyment to FIQL’s depression/self-perception subscale (r = 0.53, 95% CI [0.340, 0.686]). No significant correlations were observed between ICECAP-A autonomy and any FIQL subscales, or the total mean score. Similarly, ICECAP-A subscale attachment was only significantly correlated with the depression/self-perception subscale of the FIQL (r = 0.28, 95% CI [0.045, 0.486]). Table 2Convergent validity in Spearman’s correlation coefficientsICECAP-AStabilityAttachmentAutonomyAchievementEnjoymentCapability EQ-5D-5L Mobility-0.080.08 -0.37 * -0.29 *0.15-0.07 Self-care-0.25 *-0.04 -0.43 ** -0.48 **-0.15-0.33 * Usual activities-0.24 *-0.06 -0.44 ** -0.27 * -0.25 * -0.31 * Pain/discomfort-0.140.08-0.37 * -0.32 * -0.16-0.25 * Anxiety/depression -0.53 ** -0.35 * -0.22-0.20 -0.41 ** -0.51 ** EQ-VAS Health0.33 *0.120.29 *0.37 **0.39 ** 0.50 ** Utility score0.33 *0.140.440.32 *0.26 * 0.41 ** FIQL Lifestyle 0.37 * 0.150.14 0.32 * 0.42 ** 0.47 ** Coping/behavior 0.36 * 0.160.090.20 0.36 * 0.37 ** Depression/self-perception 0.46 ** 0.28 * 0.21 0.36 * 0.53 ** 0.54 ** Embarrassment0.150.050.060.24 *0.100.23 Total mean score0.43 **0.190.230.36 *0.46 ** 0.52 ** * p < 0.05; ** p < 0.001Correlations in bold indicate a confirmed hypothesis. ICECAP-A, ICEpop CAPability measure for adults; EQ-5D-5L, Euroqol-5D-5 L questionnaire; FIQL, fecal incontinence quality of life questionnaire Convergent validity in Spearman’s correlation coefficients * p < 0.05; ** p < 0.001 Correlations in bold indicate a confirmed hypothesis. ICECAP-A, ICEpop CAPability measure for adults; EQ-5D-5L, Euroqol-5D-5 L questionnaire; FIQL, fecal incontinence quality of life questionnaire Known-group validity Table 3 presents the results of the known-group validity analysis. Three out of seven predefined hypotheses (42.8%) were confirmed. The SEM for the ICECAP-A capability values was calculated at 0.072, indicating that score differences smaller than this threshold are likely attributable to measurement error, while larger differences may reflect true group differences. All observed group differences in ICECAP-A scores exceeded the SEM. There were statistically significant differences for EQ-VAS health scores, anxiety symptom scores, and for educational level. Parallel analyses conducted for the EQ-5D-5L and FIQL revealed similar patterns, with significant group differences in EQ-VAS health score and anxiety for both instruments, and an additional significant difference in educational level for the FIQL. When considering the magnitude of mean differences between known groups in relation to the SEM, as well as the corresponding effect size, the ICECAP-A does not appear to demonstrate superior discriminatory ability compared to the EQ-5D-5L and the FIQL. All three instruments showed comparable significant performance in distinguishing between anxiety levels (GAD-7), in which a large effect size was seen for the EQ-5D-5L (ES = -0.621, 95% CI [-0.877,-0.248], p < 0.001). Although not statistically significant, the EQ-5D-5L showed larger ES than the ICECAP-A and FIQL for depression level (PHQ-9) and symptom severity (Vaizey score). The FIQL, in turn, showed a significant difference across groups with different educational levels, albeit with only a small effect size (ES = -0.154, 95% CI [-0.292, -0.016], p = 0.002), which was however larger than the ES observed for ICECAP-A. Table 3Known-group validityICECAP-A capability valueEQ-5D-5L utility scoreFIQL total mean score n Meanp-valueES AMeanp-valueES AMeanp-valueES ASexMale120.840.711-0.0430.610.809-0.0282.610.116-0.185Female600.870.722.29BMI (kg/m2)< 25270.880.330-0.1150.780.152-0.1692.510.091-0.199≥ 25450.850.662.24EQ-VAS health< 65250.80 0.005 -0.3290.52 < 0.001 -0.6212.01 < 0.001 -0.408≥ 65470.900.812.52GAD-7< 10620.90 < 0.001 -0.4310.75 < 0.001 -0.3992.42 0.003 -0.355≥ 10100.630.401.83PHQ-9< 1060 C0.860.412-0.0970.710.179-0.1582.340.886-0.017≥ 1011 C0.880.682.31Vaizey score< 13200.840.446-0.0900.690.266-0.1312.320.990-0.002≥ 13520.870.712.35EmploymentNo350.850.416-0.0960.690.453-0.0882.330.826-0.026Yes370.880.722.35Education levelLower290.83 0.033 -0.069B0.660.305-0.005B2.04 0.002 -0.154BSecond210.860.722.63Tertiary220.910.762.46Comorbid IBSYes80.870.8860.0170.700.5480.0702.320.3420.112No640.860.772.53SurgeryYes70.860.753-0.0370.700.165-0.1642.110.291-0.124No650.870.802.36Bold p-values indicate statistical significance considering Bonferroni correction. ICECAP-A, ICEpop CAPability measure for adults; EQ-5D-5L, Euroqol-5D-5 L questionnaire; FIQL, fecal incontinence quality of life questionnaire; ES, effect size; BMI, body mass index; GAD-7, generalized anxiety disorder-7 questionnaire; PHQ-9, patient health questionnaire-9; IBS, irritable bowel syndromeA Effect sizes (ES) were calculated using Z-values (Mann-Whitney-U); interpretation: small (0.1–0.3), moderate (0.3–0.5), large (≥ 0.5)B Effect sizes (ES) were calculated using H-values (Kruskal-Wallis); interpretation: 0.01–0.06 weak, 0.06–0.14 moderate, > 0.14 strongC Data missing in one subject Known-group validity Bold p-values indicate statistical significance considering Bonferroni correction. ICECAP-A, ICEpop CAPability measure for adults; EQ-5D-5L, Euroqol-5D-5 L questionnaire; FIQL, fecal incontinence quality of life questionnaire; ES, effect size; BMI, body mass index; GAD-7, generalized anxiety disorder-7 questionnaire; PHQ-9, patient health questionnaire-9; IBS, irritable bowel syndrome A Effect sizes (ES) were calculated using Z-values (Mann-Whitney-U); interpretation: small (0.1–0.3), moderate (0.3–0.5), large (≥ 0.5) B Effect sizes (ES) were calculated using H-values (Kruskal-Wallis); interpretation: 0.01–0.06 weak, 0.06–0.14 moderate, > 0.14 strong C Data missing in one subject Acceptability, floor and ceiling effects Figure 1 presents the distribution of ICECAP-A capability values, EQ-5D-5L utility scores, and FIQL total mean scores at T0 (n = 72) and at T1 (n = 61). At T0, no missing data were observed for any of the three measures. At T1, 15.3% of data were missing, which was entirely attributable to trial dropout (n = 11) rather than incomplete questionnaires among participants who remained in the study. Missing data were evenly distributed across subscales. Among participants who completed the trial (n = 61), no missing data were observed, resulting in a 100% completion rate at the measure level (overall completion rate: 84.7%). ICECAP-A scores ranged from 0.24 (n = 1) to 1.00 (n = 1) at baseline and from 0.18 (n = 1) to 1.00 (n = 4) at follow-up. EQ-5D-5L utility scores ranged from − 0.34 (n = 1) to 1.00 (n = 5) at baseline and from 0.34 (n = 1) to 1.00 (n = 7) at follow-up. FIQL total mean scores ranged from 1.16 to 3.70 (both n = 1) at baseline and from 1.21 to 4.06 (both n = 1) at follow-up. None of the instruments showed floor or ceiling effects exceeding the conventional 15% threshold at either time point, indicating an acceptable distribution of scores across the full range of each measure. Distributions of scores at baseline (T0; n = 72) and after the 8-weeks follow-up period at 9 weeks (T1; n = 61, due to trial drop-out). (A) Capability scores of the ICEpop CAPability score for adults (ICECAP-A). (B) Utility scores of the Euroqol-5d-5 L (EQ-5D-5L). (C) Total mean scores of the fecal incontinence quality of life (FIQL) questionnaire Discussion Despite clear theoretical distinctions between the capability-based ICECAP-A and HRQoL instruments such as the EQ-5D-5L and FIQL, our findings demonstrate similarity in psychometric performance in patients with fecal incontinence. The ICECAP-A, grounded in the capability approach, assesses what individuals are able to be and do rather than focusing solely on health status or condition-specific limitations. Nevertheless, convergent and known-group validity were comparable across instruments, and ICECAP-A did not show superior discriminative ability. All three measures demonstrated acceptable degree of acceptability, without notable floor or ceiling effects. These results suggest that, in this population, broader capability well-being is closely intertwined with HRQoL, leading to partial overlap rather than clearly distinct or incrementally informative constructs. The moderate to strong correlations of the ICECAP-A with the EQ-5D-5L utility scores, its EQ-VAS health score, and the FIQL total mean scores, as well as several expected subscale correlations suggest some conceptual overlap between these two questionnaires in this group of patients. As anticipated, the ICECAP-A subscale attachment showed minimal significant correlations with EQ-5D-5L and FIQL subscales, except for a significant correlation with the EQ-5D-5L anxiety/depression subscale. Similarly, the autonomy subscale did not correlate significantly with any FIQL subscales, in line with theoretical expectations. In accordance with our predefined hypotheses, stronger correlations were expected only where conceptual overlap was expected, whereas weaker or absent associations were anticipated for subscales capturing distinct aspects of capability well-being. Our findings suggest a degree of conceptual overlap between ICECAP-A, EQ-5D-5L, and FIQL, while also indicating distinctions in the specific constructs each instrument is designed to capture, this is a phenomenon previously reported across various clinical populations, including individuals with lower urinary tract symptoms, and chronic kidney disease [19, 39, 40]. The ICECAP-A subscales autonomy and achievement did not show significant correlations with the EQ-5D-5L anxiety/depression subscale, despite prior evidence suggesting a moderate to strong association in populations with mental health problems [20, 30]. A potential explanation lies in the characteristics of the present sample, in which only a small proportion of participants reported moderate to severe anxiety or depressive symptoms (GAD-7 and PHQ-9 scores ≥ 10), and mean scores on both scales were relatively low. Additionally, the presence of mild ceiling effects, with 5.5% of participants scoring the maximum ICECAP-A value and 9.7% scoring the highest EQ-5D-5L utility value at follow-up. This may have attenuated the strength of correlations, particularly in subscales related to mental health. Of the seven predefined hypotheses for known-group validity, three were confirmed. ICECAP-A did not demonstrate superior discriminative ability compared to EQ-5D-5L or FIQL; EQ-5D-5L showed larger effect sizes for anxiety, with similar trends for depression and symptom severity, while education level was more strongly associated with FIQL. Given the limited power of several subgroup analyses (e.g., small sample sizes for anxiety) and the proportion of confirmed hypotheses, these findings should be interpreted with caution and are best considered inconclusive rather than indicative of comparable performance between instruments. Although some overlap in outcomes was observed, evidence for known-group validity of the ICECAP-A remained inconsistent across hypotheses, in. A potential explanation is that the ICECAP-A, despite its broader conceptual scope, reflects some health-related aspects already captured by traditional HrQoL measures [19, 20]. Mean ICECAP-A capability scores remained consistently high (≥ 0.84) across subgroups, except among those with severe anxiety, suggesting a relatively preserved sense of well-being in this population. This contrasts with findings in other clinical groups, such as individuals with irritative lower urinary tract symptoms, chronic kidney injury, or those with opiate dependence, where ICECAP-A demonstrated greater sensitivity to the psychological or functional impact [21, 39–41], The relatively low baseline EQ-5D-5L and FIQL scores align with previously reported values in clinical FI populations, suggesting that our findings reflect disease-inherent impairment rather than an unusually mildly affected study sample. From a capability perspective, individuals may preserve valued life domains—such as autonomy and attachment—despite persistent symptoms, reflecting adaptation and coping processes, and possibly response shift. Nonetheless, ceiling clustering, sample homogeneity, and limited sensitivity of capability measures to clinical severity cannot be excluded. In addition, while the SEM provided an estimate of measurement error to contextualize observed differences, it should not be interpreted as a threshold for clinically meaningful differences. The relevance of between-group differences therefore requires consideration beyond measurement precision alone. Well-being as captured by the ICECAP-A may not align directly with clinical symptom trajectories in this population. From a capability perspective, individuals may preserve valued aspects of life—such as autonomy, attachment, or enjoyment—despite persistent symptoms, reflecting adaptation and coping. While such processes may explain the relatively high capability scores observed, reduced sensitivity of capability measures to clinical severity cannot be excluded. This could explain why condition-specific HRQoL instruments appear more sensitive to the direct impact of fecal incontinence, whereas capability measures potentially capture broader resilience. This perspective underscores the importance of considering psychosocial determinants of quality of life alongside clinical symptoms. This interpretation is supported by findings of a Norwegian outpatient cohort, demonstrating that psychosocial factors (e.g., social isolation), rather than physical symptoms (e.g., bowel leakage), were more strongly associated with impairments in HrQoL and plays a central role in patients’ adaptation processes and altering coping strategies [42]. Similarly, two studies by Assmann et al. on patient preferences and lived experiences in FI showed that non-physical dimensions outweighed symptom severity in shaping perceived burden, although physical symptoms remain of considerable impact on daily life [11, 17]. On that note, it is possible that the predefined subgroups used (based on earlier psychometric research on the ICECAP-A in general populations), are less relevant for assessing well-being and HrQoL in patients with FI. For example, known determinants of HrQoL in this population such as employment status and history of surgery, showed non-significant and minimal effect sizes, whereas weak to moderate associations had been anticipated. This may be partly explained by the relative homogeneity of the current sample and the generally high levels of reported well-being, potentially limiting the discriminatory capacity of all three instruments examined. The psychometric performance of the ICECAP-A was further evaluated through the assessment of floor and ceiling effects, as well as overall acceptability. Despite differences in underlying conceptual frameworks, both the ICECAP-A and FIQL include more response categories than the EQ-5D-5L, which may enhance sensitivity and potentially reduce floor and ceiling effects [18, 22]. In the present study, floor effects were minimal, occurring in only one participant across all three instruments at both baseline and follow-up. Ceiling effects were most pronounced for the EQ-5D-5L, observed in 6.9% of participants at baseline and 9.7% at follow-up, but remained below the commonly accepted threshold of 15%. For the ICECAP-A and FIQL, ceiling effects were lower (ICECAP-A: 1.4% at baseline and 5.5% at follow-up; FIQL: 1.4% at both time points), indicating no clear advantage of the ICECAP-A in this regard. However, visual inspection of the score distribution suggested clustering toward higher ICECAP-A values at both time points. Although this did not meet the formal criteria for ceiling effects, such clustering may still reflect reduced variability at the upper end of the scale and could limit sensitivity and discriminative ability in this population. These findings may partly be attributable to the small sample size. These findings may partly be attributable to the small sample size. Importantly, the overall completion rate of 84.7% resulted from trial dropout rather than limited measurement acceptability, as no missing data were observed among participants who completed the study. While differential dropout between groups occurred, no systematic differences in baseline characteristics or outcome scores were identified between completers and non-completers, indicating a limited risk of attrition bias. Taken together, these findings provide no evidence of psychometric superiority of the ICECAP-A over the EQ-5D-5L or FIQL with respect to distributional properties or respondent burden, and furthermore, may be explained by adaption and altered coping strategies to this chronic condition. A key strength of this study lies in the use of a well-defined clinical sample of patients with FI, reflecting the typical population encountered in daily clinical practice. This enhances the external validity and clinical relevance of the findings. Furthermore, by incorporating a diverse set of validation instruments, including the generic EQ-5D-5L, the condition-specific FIQL, and the capability-based ICECAP-A, this study enabled a comprehensive evaluation of multiple conceptual dimensions of health and well-being. The use of multiple psychometric analyses, combined with the inclusion of evidence-based subgroup comparisons and known determinants of HrQoL in FI, further strengthened the methodological rigor and interpretability of the results. Nonetheless, several limitations must be acknowledged. Despite reaching the pre-calculated required sample size, the overall study sample remained relatively small, as it was determined by the parent clinical trial rather than psychometric considerations. While adequate for the predefined correlation analyses, small subgroup sizes limited statistical power, meaning non-significant findings should be interpreted cautiously rather than as evidence of equivalence. Furthermore, it consisted exclusively of participants enrolled in a clinical trial evaluating an anal plug for FI, which required patients to actively apply and demonstrate a certain level of mobility, which may have introduced selection bias and limits generalizability. In addition, the sample showed limited heterogeneity, as reflected by restricted variability in ICECAP-A capability scores and a predominance of participants with higher symptom severity. These factors may have reduced the instruments’ ability to discriminate across subgroups. Moreover, not all COSMIN-defined measurement properties were evaluated. Specifically, internal consistency (within the reliability domain) was not assessed, as the ICECAP-A subscales represent distinct dimensions rather than reflective indicators of a single latent construct [28, 43, 44], and the available data were derived from a clinical trial that was not specifically designed to support a comprehensive reliability assessment. Furthermore, responsiveness (as defined by COSMIN as the ability to detect change over time) could not be examined due to the absence of an appropriate external anchor and sufficiently detailed longitudinal data on clinically meaningful change; future longitudinal studies specifically designed to assess these properties are therefore warranted. To ensure generalizability, such studies should be conducted across different clinical settings, including nursing homes and regular outpatient clinics in both primarily and secondary/tertiary care). Comparative evaluations with established HrQoL instruments, as well as well-being measures such as the EQ Health and Well-being (EQ-HWB) and the newly developed 10-item Well-being instrument (WiX), will be essential to clarify its added value [45–47]. Overall, the findings in this study indicate that, among patients with FI, the ICECAP-A performs comparably to established HrQoL instruments with respect to acceptability, convergent validity, and known-group validity. Its broader conceptualization of well-being however, does not appear to offer additional clinical utility for treatment evaluation or patient monitoring. The relative stability of ICECAP-A scores, irrespective of symptom severity, may reflect underlying adaptation and coping processes that preserve subjective well-being despite substantial disease burden. Future studies should investigate whether ICECAP-A adds meaningful patient-centered information beyond traditional HrQoL measures, particularly when aiming to capture broader aspects of well-being alongside HrQoL. Such evidence would strengthen the case for its integration into holistic outcome assessment and inclusive economic evaluations in this patient population. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Not applicable. Abbreviations Fecal incontinence Health-related quality of life Euroqol 5 level EQ5Dquestionnaire Fecal incontinence quality of life questionnaire ICEpop CAPability measure for Adult Inflammatory bowel disease Maastricht University Medical Center+ EQ-5D-5L visual analogue scale Generalized anxiety disorder-7 questionnaire Patient health questionnaire-9 Body mass index Effect size Standard error of the mean Standardizedresponse means EQ-health and well-being 10-item well-being instrument Author contributions Conceptualization: SG, DK, and BE; Methodology: SG, BE; Formal analysis and investigation: SG and BE; Writing - original draft preparation: SG; Writing - review and editing: SG, DK, ZW, SA, SB, and BE; Funding acquisition: DK and BE; Resources: DK and BE; Supervision: DK, ZW, and BE. All authors read and approved the final version of the manuscript. Funding The authors did not receive support from any organization for the submitted work. Data availability Data and SPSS syntax’s are available from the corresponding author upon reasonable request. Declarations Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The need for ethical approval was waived by the medical ethics committee of Maastricht University Medical Center+ (MUMC+). Consent to participate A written informed consent was obtained from all individuals participating in the original clinical trial, of which the data in the current cross-sectional study was obtained. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement Data and SPSS syntax’s are available from the corresponding author upon reasonable request.