Are the dimensions of positive health associated with work participation and receiving disability benefit? A cross-sectional study in outclinic patients with chronic gastrointestinal disorders

Abstract

Objective

To describe work participation (WP) and receiving disability benefit (DB) in persons with chronic gastrointestinal disorders (CGID); to describe associations between the six dimensions of positive health (PH) and WP and DB, respectively; and to assess whether the type of CGID is associated with WP and receiving DB.

Design

Cross-sectional and observational questionnaire study.

Setting

Maastricht University Medical Centre+ (tertiary care hospital), including regional, supra-regional and national referrals.

Participants

441 patients of working age (18–66 years) with CGID who visited the outpatient department of the Gastroenterology-Hepatology Clinic between March 2019 and June 2021 (mean±SD age: 45.0±14.7 years, 68.5% women).

Primary/secondary outcome measures

Associations of PH dimensions and WP and receiving DB, respectively; association of type of CGID and WP, and receiving DB. The main independent variables (PH dimensions) were bodily functions, mental well-being, meaningfulness, quality of life (QoL), participation and daily functioning.

Results

Among 441 working-age patients, 49% worked; 20% received DB (40% of non-workers). All PH dimensions related significantly to WP and DB (bivariate level). After adjusting for demographics and CGID type, better general health (dimension QoL) (OR 1.018; 95% CI 1.001 to 1.035, p=0.040) and daily functioning dimension (OR 1.030; 95% CI 1.013 to 1.047, p<0.001) related to higher WP. In the fully adjusted model, daily functioning showed the strongest association with receiving DB. Each one-point increase in Short Form-36 physical functioning was associated with a 2% lower likelihood of receiving DB (OR 0.975; 95% CI 0.959 to 0.992; p=0.004).

Conclusions

Of the 441 working-age patients with CGID, only 49% worked; most non-workers did not receive DB. Of the PH dimensions, daily functioning appears most directly related to WP and receiving DB; CGID type had minimal impact.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The dimensions of positive health were used as independent predictors of work participation (WP) and receiving disability benefit (DB).

• In our research group, none in the working group received a (partial) DB (as is possible in the Dutch scheme): either they had not met the requirements for the DB scheme because they had not been sick for 2 years or their loss of earning capacity was less than 35%.

• Through the cross-sectional nature of the study, no longitudinal data were captured.

• While unemployment often is linked to poorer health and financial difficulties, the results might be explained by reversed causality.

• While the study focused on chronic gastrointestinal patients and our findings indicate that type of disease had no impact on WP and only limited effect on receiving DB, the assumption that the conclusions may be generalisable to other chronic conditions needs further research.

Introduction

In Europe, 36.1% of people aged 16 or older had a chronic disorder in 2022.¹ People with a chronic disorder are affected by the ‘disability employment gap’, being less likely in paid employment compared with the general population (21.5% in 2023).2,4 Being unemployed has significant consequences for both individuals and society. Accordingly, European data show unemployment to be associated with poverty, poor health⁵ and low quality of life (QoL).⁶ To address these differences and increase work participation (WP) for those with chronic conditions, we need to understand the experiences of patients with specific chronic diseases in a broader sense. The International Classification of Functioning, Disability and Health (ICF) model also emphasises the importance of a broad approach to health when considering and assessing participation. Positive health (PH) is a concept that covers the full range of patients’ experiences and is even broader than the ICF model. We have chosen the PH concept because of the broadness of the model, the emerging interest in this concept in the Netherlands and the availability of the concept.

PH and WP are closely linked. An integrated approach that focuses on well-being, resilience and meaningfulness contributes to a healthy, motivated and productive working environment. This study investigates how these experiences relate to WP and receiving a disability benefit (DB) in patients with chronic gastrointestinal disorders (CGID).

Patients with chronic gastrointestinal disorders

CGID include disorders of the gastrointestinal tract, liver, pancreas and gallbladder. The extensive group of patients with these conditions often experiences symptoms at a relatively young age, often with major consequences for daily functioning and participation in society.^{7 8} These CGID are generally divided into three types. The first type is inflammatory bowel diseases (IBD); these are chronic inflammations of the bowel including M. Crohn and ulcerative colitis. The second type comprises chronic disorders of the liver, pancreas and gallbladder (HPB). The third type constitutes chronic disorders of the neurogastroenterology and motility (NGM) system, consisting of motility disorders and functional disorders. In motility disorders, gastrointestinal transport is affected. Functional disorders relate to changes along the gut-brain axis, such as irritable bowel syndrome etc.

Several studies describe the symptoms, mental complaints and QoL, mainly for those with IBD and HPB.^{9 10} Less information can be found for those with NGM, a broad group of disorders with different underlying causes.^{11 12}

The most common problems for patients with CGID are related to the nature of their disease: they often have abdominal pain and changes in stool frequency or consistency.¹³ Fatigue is another prevalent symptom. For patients with HPB, problems such as jaundice and itching are common. Those with NGM often experience complaints such as abdominal pain, dysphagia, stasis, nausea and vomiting, bloating and distension, constipation, diarrhoea and faecal incontinence.¹⁴

Well-known associated comorbidities in other organ systems include rheumatic disorders and skin diseases.^{9 10 14 15} Mental problems often accompany gut disorders based on interaction through the gut-brain axis.⁸ A further consequence of gastrointestinal (GI) disorders is malnutrition, and the group with CGID was found to perform worse than norm groups over almost all life areas.¹⁶

Work participation and disability benefit

There is some evidence regarding how CGID relate to mental health problems and reduced QoL, but how CGID relate to WP and DB specifically remains unclear. In the Netherlands (place of research), a DB can be obtained if there is a theoretical loss of earning capacity due to disability of more than 35% after 2 years of sickness absence. Throughout the study, no policy changes occurred. The Netherlands is among the European Union (EU) countries with the highest WP.¹⁷ In 2020, 82% of healthy persons aged 15–75 years had paid work, and over 60% of those with a chronic disorder were in this age range—though working fewer hours.¹⁸ Patients with IBD had an overall WP of 61%.³ For HPB and NGM patients, research is scarce, although Van den Burg et al¹⁹ found a WP of 45–59% in liver transplant patients.

Studies on WP of people with chronic disorders tend to focus only on employees with permanent work, ignoring increasing numbers of workers with flexible contracts (including temporary and zero-hour contracts) and self-employed persons, the latter group now constitutes 13% of Dutch workers.²⁰ Other studies focus on DB in persons with a chronic condition, assuming that those not receiving a benefit work.²¹ However, WP and receiving DB are not communicating vessels: not working does not automatically mean receiving a DB, and vice versa. There is a disability gap between them.²² It is thus essential to include both WP and DB as outcomes.

Positive health

The need arises thus to map the health of patients with CGID in a more extensive, encompassing manner. The concept of PH provides one option to achieve this.²³ PH focuses on the individual’s ability to cope and adapt to life’s challenges and, in doing so, maintain control. The concept includes six key dimensions of health: bodily functions, mental well-being, meaningfulness, QoL, participation and daily functioning. Altogether, these provide a promising avenue for better understanding the participative ability of patients with GI disorders.

Patients with CGID experience problems across various domains, with only limited information available on their WP and DB, and even less information on how these outcomes relate to their specific determinants of health and functioning. Individual dimensions of PH have been studied, but—at the time of writing—we are unaware of studies including all dimensions in similar samples. Given the advantages of measuring all six dimensions simultaneously, we used validated questionnaires for each PH dimension.¹⁶ We found that (1) a suitable impression could be obtained about the problems these patients experienced across PH dimensions and (2) patients with GI disorders showed more problematic scores than norm populations (excluding the PH dimension meaningfulness).¹⁶ Building on this and aiming to address the shortcoming of neither WP nor DB being included previously, we set out to investigate these PH dimensions in relation to WP and DB further.

Relevance

To reduce the disability gap in WP of patients with CGID, understanding the determinants of WP and receiving DB is necessary. Previous studies focused on IBD.^{7 24} Some research has been done on patients with HPB, but research on NGM seems lacking.25,27 Therefore, this study focuses on WP and DB in patients with CGID, in which three types of GI disorders are examined to better understand the impact of a chronic disease on the patient’s life and, more specifically, the patient’s participation.

Additionally, we focused on the relationships between the dimensions of PH and both WP and receiving DB.

We further aimed to study whether PH covers the differences between disease types or whether the type of disorder still has a unique additional association with WP and DB. In other words, does PH cover the variations between diseases?

Therefore, the research questions are as follows:

1. What is the prevalence of WP and receiving DB in this group of patients with CGID?

2. How are the six dimensions of PH associated with WP in patients with CGID?

3. How are the six dimensions of PH associated with receiving DB in patients with CGID disorders?

4. Is the type of CGID associated with WP and receiving DB in addition to the association between dimensions of PH with WP and DB?

Hypotheses

(H1) The prevalence of WP is lower and the prevalence of receiving DB is higher in patients with CGID compared with a healthy population. In De Boer et al,³ it was found that 61% of persons with IBD participated in work in the Netherlands. As this subgroup experiences fewer complaints than the other two groups, we hypothesise that those with HPB or NGM participate less in work than those with IBD. Similarly, we hypothesise that patients with GI disorders receive DB more often than the general population, with higher frequencies in those with HPB or NGM than those with IBD.

(H2) We hypothesise that the PH dimensions will be positively associated with WP in patients with CGID.^{26 28}

(H3) We hypothesise that the PH dimensions will be negatively associated with receiving DB in patients with CGID.29,31

(H4) We hypothesise that PH covers all aspects of the type of these disorders and that the type of disorder has no unique additional association with WP and DB. Baanders et al³² found that labour market position is primarily related to ‘common consequences of chronic illness’ without an independent effect of the specific disease diagnosis.

Materials and methods

Design and patients

This cross-sectional study is part of a broader study in which patients attending the outpatient department of the Gastroenterology-Hepatology Clinic of the Maastricht University Medical Centre+ (MUMC+) were invited to participate in this single-centre, prospective and observational questionnaire study, including an interview and patient file data. The population of patients with CGID in the MUMC+ included regional, supra-regional and national referrals.

Patients aged 18 years and older, diagnosed with CGID, and scheduled for an appointment at the outpatient clinic between March 2019 and June 2021, were invited by letter to participate. In answering our research questions, we focused on adult patients of working age (18–66 years, the latter being the national retirement age in the Netherlands).

The letter included an invitation for their regular appointment, an informed consent form and a self-report composite questionnaire that included all the separate domain questionnaires. After obtaining written informed consent, the completed questionnaires were collected. Additional information was obtained during a short interview by the researcher, and clinical data from the electronic patient files (medical diagnoses, duration of complaints and comorbidity) were added.

Variables

Independent variables

As described previously,¹⁶ validated questionnaires covering each PH dimension were grouped to score the health status.

1. Bodily functions was measured with the Gastrointestinal Symptom Rating Scale (GSRS)³³ and the Short Nutritional Assessment Questionnaire (SNAQ),³⁴ and we scored the body mass index (BMI).

2. Mental well-being was measured with the Patient Health Questionnaire (PHQ-9)³⁵ for depressive symptoms and the Generalized Anxiety Disorder (GAD-7)³⁶ for anxious symptoms. The Short Form-36 (SF-36)³⁷ subscales were distributed over the various PH dimensions to best match the specific dimension. The Mental Health subscale of the SF-36 was used to measure the mental well-being dimension.

3. QoL was measured with the EuroQol-5D-5L (EQ-5D-5L) and EuroQol-VAS (EQ-VAS)³⁸ as well as with the SF-36 Vitality and General Health subscales.

4. Meaningfulness was measured using the Disease Experience Questionnaire (DEQ, Meaningfulness subscale).^{16 39} This subscale defines the ‘meaningful’ disease experience as the opportunity to set new life priorities. The questionnaire was introduced in 2019, once data collection had already started, and thus meaningfulness was only measured in a subset of 235 of the 555 participants (181 out of 441).

5. Participation was measured with the Social Functioning, Emotional Role Limitations and Physical Role Limitations subscales of the SF-36. WP was also measured but tested as a dependent variable, as described below.

6. Daily functioning was measured by using the Physical Functioning subscale of the SF-36.

Dependent variables

The dependent variables in this study are ‘work participation’ (WP) and ‘receiving disability benefits’ (DB). In previous studies, we considered WP and DB as part of the Social Participation dimension of PH, but in this study, we consider them as dependent variables and the dimensions of PH as independent variables.

For WP, the following categories were used: permanent or temporary employment, temporary/on-call/flex workers, and/or self-employed with or without employees. Other categories, for example, giving informal care, housewives/men, welfare recipients and people with an early retirement pension, were regarded as not working.

Receiving DB was defined as receiving a full or partial DB under the Disability Insurance Act, Income and Disability Act, or the Disability Insurance for Young Disabled Persons Act, versus not receiving one of these benefits.

Statistical analyses

Statistical analyses were performed with IBM SPSS Statistics V.27.0.1.0. χ² tests were used to assess initial differences between the working group and the receiving DB group. Analysis of Variance (ANOVAs) were used to test the significance of the differences between more subgroups. We employed hierarchical multivariate logistic regression analyses for the bivariate and multivariate testing of PH dimensions. A p value <0.05 was considered significant. Nagelkerke’s R² was determined to estimate the explained variance in the dependent variable. Visual inspection of log-odds across quartiles of the continuous independent variables indicated that the linearity assumption was largely met. To check for effects, categorical independent variables are checked. To correct for the possible role of the subgroups, we used the subgroups (IBD/HPB/NGM) as an independent variable, with the IBD group versus the rest (highest WP) and the NGM group (highest DB) versus the rest. Multicollinearity was assessed using variance inflation factors (VIFs). All VIF values were below 5, indicating no problematic multicollinearity. More specifically, the three highest scores (3.6–4.7) were found among mental health-related variables (PHQ-9, GAD-7, SF-36 Mental Health). All other VIFs ranged between 1.1 and 3.6.

The overall level of missing data was very low (mean 2.1%, maximum 4.9%), although according to Little’s MCAR test, this was significant. Simulation studies suggest that under such conditions, complete-case analysis yields estimates comparable to multiple imputation (eg, Marshall et al⁴⁰). Moreover, our sensitivity analyses excluding variables with higher missingness produced substantively identical results. We therefore assume the risk of bias to be minimal. Since the complete Meaningfulness scale was added later, we indeed had a substantial number of non-random missing data for this variable. We did not regard multiple imputation of the complete scale as a reliable method.

Results

A total of 555 patients participated in the study, out of 1170 patients contacted (response rate: 47.4%). In answering our research questions, we focused on 441 patients (79.5%) of working age (18–66 years). At the time of the study, the retirement age in the Netherlands was 66 years.

Descriptives are presented in table 1. In total, 441 participants participated in the study (IBD n=140, HPB n=105, NGM n=196). Almost 69% were female, mean (M±SD) age=45±14.7 years. Low education refers to people who have not completed any education and have only attended primary school and/or lower preparatory vocational education. Middle education refers to those who have completed secondary vocational education or preparatory education for higher education. Higher education refers to those who have completed higher vocational education or university. The distribution of participants across the three educational levels did not differ between the different disorder groups. The description of the group in a clinical perspective is published elsewhere.¹⁶

Table 1. Demographics and positive health dimensions across chronic gastrointestinal patient subgroups (N=441).

	Total (n=441*)	IBD (n=140*)	HPB (n=105*)	NGM (n=196*)	Pairwise group comparisons
Demographics
Gender (%)
Man	31.5	42.9	47.2	14.5	IBD>NGM (p<0.034); HPB>NGM (p<0.001); IBD=HPB (p=0.501)
Woman	68.5	57.1	52.8	85.5
Age (M±SD)	45.0±14.7	44.9±15.3	51.4±12.3	41.5±14.4	HPB>IBD (p=0.034) > NGM (p<0.001)
Education level (%)	(n=437)	(n=139)	(n=104)	(n=194)
Low	30.4	32.4	26.9	30.9	No overall differences (p=0.705)
Middle	40.5	37.4	40.4	42.8
High	29.1	30.2	32.7	26.3
Bodily functions
Total GSRS score (M±SD)	18.2±7.2	15.1±6.1	14.2±6.3	22.4±5.9	IBD>NGM (p<0.001); HPB>NGM (p<0.001); IBD=HPB (p=0.285)
SNAQ score (%)
0 of 1	56.2	75.0	64.2	39.0	Overall differences (p<0.001)
=2	21.8	12.9	18.9	30.0
>2	19.7	10.7	13.2	29.0
BMI (%)
<18.5	4.5	2.1	2.8	7.0	Overall differences (p=0.023)
18.5–25	49.0	50.7	39.6	52.5
>25	42.2	44.3	50.9	36.5
Mental well-being
PHQ-9 (M±SD)	7.9±6.0	6.3±5.3	7.0±5.9	9.5±6.1	IBD<NGM (p<0.001); HPB<NGM (p<0.001); IBD=HPB (p=0.329)
GAD-7 (M±SD)	5.3±5.2	4.1±4.2	5.0±5.5	6.2±5.6	IBD<NGM (p<0.001); HPB=NGM (p=0.078); IBD=HPB (p=0.198)
SF-36 Mental Health (M±SD)	70.5±20.3	73.9±19.7	71.5±20.0	67.6±20.6	IBD>NGM (p=0.005); HPB=NGM (p=0.107); IBD=HPB (p=0.356)
Quality of life
EQ-5D-5L (M±SD)	0.68±0.26	0.78±0.21	0.73±0.24	0.58±0.27	IBD>NGM (p<0.001); HPB>NGM (p<0.001); IBD=HPB (p=0.085)
EQ-VAS (M±SD)	59.7±21.3	66.0±21.3	62.9±20.3	53.6±20.3	IBD>NGM (p<0.001); HPB>NGM (p<0.001); IBD=HPB (p=0.263)
SF-36 Vitality (M±SD)	42.2±25.6	49.6±25.6	46.3±26.8	35.0±23.4	IBD>NGM (p<0.001); HPB>NGM (p<0.001); IBD=HPB (p=0.322)
SF-36 General Health (M±SD)	38.6±22.3	45.0±22.4	39.1±23.4	33.9±20.4	IBD>NGM (p<0.001); IBD>HPB (p=0.047); HPB=NGM (p=0.054)
Participation
SF-36 Social Functioning (M±SD)	63.7±29.8	73.4±26.8	67.9±30.1	54.5±29.4	IBD>NGM (p<0.001); HPB>NGM (p<0.001); IBD=HPB (p=0.136)
SF-36 Emotional Role Limitations (M±SD)	73.7±39.8	77.9±37.5	75.2±39.7	70.1±41.2	IBD=NGM (p=0.074); HPB=NGM (p=0.314); IBD=HPB (p=0.581)
SF-36 Physical Role Limitations (M±SD)	40.0±42.9	55.3±43.9	44.0±43.3	27.3±37.9	IBD>HPB (p<0.048) > NGM (p<0.001)
Daily functioning
SF-36 Physical Functioning (M±SD)	71.0±25.3	81.1±20.6	68.9±27.3	64.9±25.3	IBD>NGM (p<0.001); IBD>HPB (p<0.001); HPB=NGM (p=0.211)

	Total (n=179)	IBD (n=52†)	HPB (n=29†)	NGM (n=98†)
Meaningfulness
DEQ score (M±SD)	2.84±0.83	2.78±0.88	2.93±0.82	2.84±0.82	IBD=NGM (p=0.690); HPB=NGM (p=0.594); IBD=HPB (p=0.454)

^*All variables were tested on the full dataset with a maximum individual missing percentage of 4.9% (average of all variables=2.1%). Little’s MCAR test: χ2=533.537, DF=475, p=0.032.

^†Limited data were available for the meaningfulness dimension on account of being included later in the study.

BMI, body mass index; EQ-5D-5L, EuroQol-5D-5L; EQ-VAS, EuroQol Visual Analogue Scale; GAD-7, Generalized Anxiety Disorder; GSRS, Gastrointestinal Symptom Rating Scale; HPB, hepatico, pancreatic and biliary disorders; IBD, inflammatory bowel disease; NGM, neurogastro and motility disorders; PHQ-9, Patient Health Questionnaire; SF-36, Short Form-36; SNAQ, Short Nutritional Assessment Questionnaire.

Regarding bodily functions, the total group scored M=18.2±7.2 on the GSRS, and there was moderate to severe malnutrition in 41.3% of the group. Regarding mental well-being, the mean of the total group indicated mild depressive (PHQ-9, M=7.9±6.0) and anxiety (GAD-7, M=5.3±5.2) complaints. The SF-36 Mental Health subscale was scored M=70.5±20.3. Regarding meaningfulness, the total group (n=179) scored M=2.84±0.83 on the DEQ. QoL was scored with M=0.68±0.26 for the total group (EQ-5D-5L); EQ-VAS scored M=59.7±21.3. On the SF-36 Vitality subscale, M=42.2±25.6 was scored, and on General Health subscale, M=38.6±22.3 was scored. For social participation, the total group had M=63.7±29.8 for social functioning (SF-36) and M=40.0±42.9 for physical role limitations. For emotional role limitations (SF-36), the total group scored M=73.7±39.8. The total group scored M=71.0±25.3 for daily functioning on the SF-36 Physical Functioning subscale.

Distribution of work participation and disability benefit

In total, 441 patients reported their work status. Of them, 216 (49%) were working (full-time n=109, part-time n=107). The other 225 participants (51%) were not working: DB (n=90, 20.4% of the total group; 40% of those not working), receiving unemployment benefit (n=45), jobless (n=35), early retirement (n=19), student (n=17), housewife/man (n=16) or on social welfare (n=3). None of those working received DB (figure 1 and table 2).

Figure 1. Flowchart of work participation distribution.

Table 2. Distribution of subgroups across work participation and receiving disability benefit (N=441).

		Total (N=441; %)	IBD (n=140; %)	HPB (n=105; %)	NGM (n=196; %)	P value
Work	No DB (n=216)	49.0	62.1	48.6	39.8	IBD>HPB p=0.038; IBD>NGM p<0.001; HPB=NMG p=0.179
	DB (n=0)	–	–	–	–
No work	No DB (n=135)	30.6	23.6	31.4	35.2
	DB (n=90)	20.4	14.3	20.0	25.0	IBD=HPB p>0.999; IBD=NGM p=0.737; HPB=NGM p=0.867

DB, disability benefit; HPB, hepatico, biliary and pancreatic disorders; IBD, inflammatory bowel disease; NGM, neurogastroenterology and motility disorders.

Of the IBD patients, 62.1% were working; of the HPB patients, 48.6%; and of the NGM patients, 39.8% (differences between the IBD and HPB and NGM groups, respectively, were significant (p<0.001)).

Of the IBD patients, 14.3% received DB, 20% of the HPB patients and 25% of the NGM patients (differences not significant) (table 2).

Determinants of work participation

Bivariate analyses

Model 1 represents bivariate logistic regression analyses (table 3) used to assess associations between all individual variables representing dimensions of PH with WP. All underlying variables except the DEQ (meaningfulness dimension, p=0.115) were significantly related to WP (highest p≤0.029), with PH being positively related to WP.

Table 3. The relation between the predictor variables of positive health (PH) and work participation (WP) in people of work age with chronic gastrointestinal and hepatico-biliary-pancreatico disorders (total 342≤n≤441).

Variables	Model 1 Bivariate			Model 2 Per PH dimension			Model 3 Per PH dimension and disease group
	OR	95% CI	P value	OR	95% CI	P value	OR	95% CI	P value
Gender*	0.75	0.51 to 1.13	0.169				0.86	0.48 to 1.54	0.601
Age*	0.98	0.97 to 0.99	0.003				0.97	0.96 to 0.99	0.008
Level of education*	1.44	1.12 to 1.84	0.004				1.15	0.81 to 1.63	0.447
Bodily functions
GSRS	0.95	0.92 to 0.97	<0.001	0.95	0.92 to 0.98	0.001	1.00	0.95 to 1.05	0.993
SNAQ	0.77	0.60 to 0.97	0.029	0.93	0.71 to 1.21	0.576	1.21	0.84 to 1.74	0.300
BMI
Underweight versus normal weight	4.19	1.36 to 12.94	0.013				2.91	0.70 to 12.06	0.140
Underweight versus overweight	3.71	1.20 to 11.52	0.023				3.55	0.83 to 15.15	0.087
Mental well-being
PHQ-9	0.91	0.88 to 0.94	<0.001	0.89	0.84 to 0.94	<0.001	0.98	0.89 to 1.07	0.625
GAD-7	0.94	0.90 to 0.98	0.001	1.08	1.01 to 1.16	0.020	1.05	0.96 to 1.15	0.300
SF-36 Mental Health	1.02	1.01 to 1.04	<0.001	1.02	1.00 to 1.04	0.054	1.02	1.00 to 1.04	0.129
Meaningfulness
DEQ	0.75	0.52 to 1.07	0.115	0.75	0.52 to 1.07	0.115
Quality of life
EQ-5D-5L	16.38	6.69 to 40.15	<0.001	5.60	1.81 to 17.31	0.003	3.36	0.63 to 17.88	0.155
EQ-VAS	1.03	1.02 to 1.04	<0.001	1.00	0.98 to 1.01	0.671	0.99	0.97 to 1.01	0.353
SF-36 Vitality	1.03	1.02 to 1.03	<0.001	1.01	1.00 to 1.02	0.184	1.00	0.98 to 1.02	0.759
SF-36 General Health	1.03	1.02 to 1.04	<0.001	1.02	1.01 to 1.03	0.003	1.02	1.00 to 1.04	0.039
Participation
SF-36 Social Functioning	1.01	1.01 to 1.02	<0.001	1.00	0.99 to 1.01	0.812	0.99	0.97 to 1.00	0.047
SF-36 Emotional Role Limitations	1.01	1.01 to 1.02	<0.001	1.00	1.00 to 1.01	0.141	0.94	0.99 to 1.01	0.942
SF-36 Physical Role Limitations	1.02	1.01 to 1.02	<0.001	1.01	1.01 to 1.02	<0.001	1.01	1.00 to 1.01	0.243
Daily functioning
SF-36 Physical Functioning	1.05	1.04 to 1.06	<0.001	1.05	1.04 to 1.06	<0.001	1.03	1.01 to 1.05	<0.001
Disease group
IBD versus rest							1.19	0.60 to 2.36	0.614
NGM versus rest							0.78	0.36 to 1.68	0.524

Model 1 is the bivariate analysis of the separate questionnaires distributed over the dimensions of PH according to ‘work participation’. Model 2 is the multivariate analysis per dimension of PH according to ‘work participation’. Model 3 is adjusted for the disease: the IBD subgroup and the NGM subgroup. Meaningfulness was only measured in 179 patients.

^*For gender, the reference category was men (hence, the OR was for women). For age and level of education, the OR reported was for each one-unit increase: years and level of education (low, medium, high), respectively.

BMI, body mass index; DEQ, Disease Experience Questionnaire; EQ-5D-5L, EuroQol 5D-5L; EQ-VAS, EuroQol Visual Analogue Scale; GAD-7, Generalized Anxiety Disorder; GSRS, Gastrointestinal Symptom Rating Scale; IBD, inflammatory bowel disease; NGM, neurogastroenterology and motility disorders; PH, positive health; PHQ-9, Patient Health Questionnaire; SF-36, Short Form-36; SNAQ, Short Nutritional Assessment Questionnaire.

Multivariate analyses

In model 2 (multivariate per PH dimension), multivariate logistic regression analyses were performed per dimension, entering all variables representing one dimension simultaneously (table 3). In these analyses, all dimensions of PH except for Meaningfulness were significantly and positively related to WP, with GSRS, PHQ-9, GAD-7, EQ-5D-5L, SF-36 General Health, SF-36 Physical Role Limitations and SF-36 Physical Functioning remaining significantly related.

In model 3 (hierarchical; table 3), variables were included simultaneously in three hierarchical steps. Here, we found that step 1, sociodemographic variables, provided a significant relation with WP (Nagelkerke R²=0.054, χ²=14.138, p=0.003), as did step 2, PH dimensions (Nagelkerke R²=0.340, χ²=86.322, p<0.001). Step 3, the subgroup classification, did not significantly improve the model (Nagelkerke R²=0.344, χ²=1.498, p=0.473), indicating that this set of subgroup comparisons (ie, IBD vs rest; NGM vs rest) was not statistically relevant towards explaining more WP beyond steps 1 and 2. Meaningfulness was absent because of many missing values due to its later addition in the process of gathering data. A sensitivity analysis for WP on a subset where meaningfulness data were available revealed that meaningfulness was not significant (p=0.527), and second, adding meaningfulness in step 2 showed no substantial improvement in the model (Nagelkerke R²=0.467 → 0.470, χ²=54.562 → 54.963, p=<0.001).

Determinants of receiving disability benefit

Bivariate analyses

Model 1 showed that every PH dimension, but not every variable, was significantly related to receiving DB (highest p≤0.026; table 4): having more complaints and malfunctioning were associated with a higher chance of receiving DB. Only the SNAQ and the BMI were non-significant.

Table 4. The relation between the predictor variables of positive health (PH) and receiving disability benefit (DB) in people of work age with chronic gastrointestinal and hepatico-biliary-pancreatico disorders (343≤n≤441).

Variables	Model 1 Bivariate			Model 2 Per PH dimension			Model 3 Per PH dimension and disease group subtype
	OR	95% CI	P value	OR	95% CI	P value	OR	95% CI	P value
Gender*	0.80	0.49 to 1.30	0.367				0.55	0.26 to 1.17	0.121
Age*	1.03	1.01 to 1.05	<0.001				1.05	1.02 to 1.08	0.001
Level of education*	0.70	0.51 to 0.95	0.022				0.64	0.41 to 0.99	0.046
Bodily functions
GSRS	1.05	1.01 to 1.08	0.007	1.05	1.01 to 1.09	0.008	0.98	0.92 to 1.05	0.512
SNAQ	1.04	0.77 to 1.40	0.805	0.85	0.61 to 1.19	0.341	0.62	0.39 to 1.01	0.052
BMI
Underweight versus normal weight	0.50	0.18 to 1.37	0.176				0.79	0.16 to 3.84	0.768
Underweight versus overweight	0.68	0.25 to 1.87	0.450				0.68	0.14 to 3.37	0.637
Mental well-being
PHQ-9	1.09	1.05 to 1.13	<0.001	1.12	1.05 to 1.20	0.001	1.08	0.97 to 1.21	0.157
GAD-7	1.05	1.01 to 1.10	0.026	0.93	0.86 to 1.01	0.091	0.92	0.82 to 1.02	0.121
SF-36 Mental Health	0.98	0.97 to 0.99	<0.001	0.99	0.97 to 1.01	0.494	1.00	0.97 to 1.03	0.968
Meaningfulness
DEQ	1.97	1.21 to 3.22	0.007	1.97	1.21 to 3.22	0.007
Quality of life
EQ-5D-5L	0.12	0.05 to 0.28	<0.001	0.35	0.11 to 1.10	0.073	0.85	0.13 to 5.83	0.871
EQ-VAS	0.98	0.97 to 0.99	<0.001	1.00	0.98 to 1.02	0.985	1.01	0.98 to 1.03	0.624
SF-36 Vitality	0.98	0.97 to 0.99	<0.001	0.99	0.98 to 1.01	0.214	0.98	0.96 to 1.00	0.095
SF-36 General Health	0.97	0.95 to 0.98	<0.001	0.98	0.96 to 0.99	0.005	0.99	0.96 to 1.01	0.223
Participation
SF-36 Social Functioning	0.98	0.98 to 0.99	<0.001	0.99	0.98 to 1.00	0.083	0.99	0.98 to 1.01	0.299
SF-36 Emotional Role Limitations	0.99	0.99 to 1.00	0.001	1.00	0.99 to 1.00	0.341	1.00	0.99 to 1.01	0.654
SF-36 Physical Role Limitations	0.99	0.98 to 0.99	<0.001	0.99	0.99 to 1.00	0.070	1.01	1.00 to 1.02	0.246
Daily functioning
SF-36 Physical Functioning	0.97	0.96 to 0.98	<0.001	0.97	0.96 to 0.98	<0.001	0.98	0.96 to 0.99	0.004
Disease group
IBD versus rest							2.52	0.94 to 6.81	0.068
NGM versus rest							4.18	1.47 to 11.91	0.007

Model 1 is the bivariate analysis of the separate questionnaires distributed over the dimensions of PH according to ‘receiving disability benefit’. Model 2 is the multivariate analysis per dimension of PH according to ‘receiving disability benefit’. Model 3 is adjusted for the disease: the IBD subgroup and the NGM subgroup. Meaningfulness was only measured in 179 patients.

^*For gender, the reference category was men (hence, the OR was for women). For age and level of education, the OR reported was for each one-unit increase: years and level of education (low, medium, high), respectively.

BMI, body mass index; DEQ, Disease Experience Questionnaire; EQ-5D-5L, EuroQol 5D-5L; EQ-VAS, EuroQol Visual Analogue Scale; GAD-7, Generalized Anxiety Disorder; GSRS, Gastrointestinal Symptom Rating Scale; IBD, inflammatory bowel disease; NGM, neurogastroenterology and motility disorders; PH, positive health; PHQ-9, Patient Health Questionnaire; SF-36, Short Form-36; SNAQ, Short Nutritional Assessment Questionnaire.

Multivariate analyses

In model 2 (multivariate per PH dimension; table 4), we found that higher levels of gastrointestinal and depressive complaints were related to receiving DB more often. Better general health and physical functioning were associated with receiving DB less often.

Model 3 (hierarchical; table 4) showed significant improvements in explaining DB across all three steps: for demographics (Nagelkerke R²=0.069, χ²=14.664, p=0.002); dimensions of PH (Nagelkerke R²=0.282, χ²=49.612, p<0.001); and subgroup classifications (Nagelkerke R²=0.313, χ²=7.973, p=0.019). This indicates that all three steps, including subgroup classifications, contributed significantly to explaining the odds of receiving DB. Specifically regarding step 3, the IBD group did not significantly differ from both other groups (OR=2.52, 95% CI 0.94 to 6.81, p=0.068) but the NGM group showed higher odds of receiving DB (OR=4.18, 95% CI 1.47 to 11.91, p=0.007), compared with both other groups (ie, HPB and IBD).

Discussion

In this study, we set out to achieve four goals. First, investigate the prevalence of WP and receiving a DB among 441 patients with CGID. Second, investigate to what degree the six dimensions of PH were associated with WP and third, with DB. Fourth, determine whether the type of CGID was associated with WP and DB in addition to the association between dimensions of PH with WP and DB, respectively. Below, we will discuss the outcomes.

Participating in paid work and receiving disability benefit

Among patients with CGID, 49% participated in paid work, confirming H1. The WP was significantly lower than that in the general Dutch population (70.5% in 2020, p<0.001), and aligning with previous studies on comparable groups.^{3 19} Our HPB subgroup comprised a heterogeneous group with various liver disorders, whereas the latter research group¹⁹ included only liver transplant patients. For the NGM group, no comparable WP data were found, while we demonstrated that this group, as well as the HPB group, had paid work significantly less often than the IBD subgroup.

In the total CGID group, significantly more patients received DB. Surprisingly, receiving DB did not significantly differ between the three subgroups: we expected more participants to receive DB in the HPB and NGM groups (second part H1). In the IBD subgroup, a substantially lower part of the patients (14.3%) received DB than the 23% reported by De Boer et al.³ This difference may be due to a higher proportion of highly educated participants in the study of De Boer et al³ (with a greater chance of higher income and so more often DB). On the other hand, our group consisted of older patients (45 vs 41 years), which is a risk factor for receiving DB,⁴¹ as also confirmed by our findings (table 4). Among the HPB patients, 20% received DB. Büsch et al⁴² reported a DB rate of 30% in Sweden among chronic hepatitis C patients, with the variation possibly explained by group composition differences (our group had fewer men (47.2% vs 65%) and more highly educated patients (32.1% vs 13%)). Fazekas et al²⁵ found that liver transplant patients had a mental health component and a physical health component which prevented them from actual return to work, despite their wish to return to work. Possibly, individual and systemic factors influence the return to work.²⁵ According to Black et al,¹¹ patients with NGM disorders often experience limitations in their daily functioning because they have many invalidating symptoms (such as pain, diarrhoea, etc). This can lead to social isolation, among other things. Based on the severity of NGM symptoms and their impact, Black et al point to early identification and diagnosis so that appropriate treatment and support can be initiated.¹¹

No comparable research on DB exists for the NGM subgroup. Considering that legislation about DB differs considerably between countries,⁴³ our findings appear in line with other studies.

Of the non-working persons, 40% had alternative sources of income (eg, a partner’s earnings or a study grant). The NGM group had a significantly higher percentage of women, aligning with other publications.^{3 19} Women are not only more at risk for DB but also not working twice as often as men in the Netherlands.⁴⁴ Unemployment is a non-negligible risk of financial vulnerability⁶ and moreover, those not working nor receiving DB have no access to the return-to-work expertise of occupational and insurance physicians (accredited 4-year specialisations in the Netherlands, and access to them is a legal right for employees).

Determinants of work participation

Bivariate analyses showed that all individual scales representing the six dimensions of PH were individually associated with WP, confirming hypothesis 2. Age was significantly associated with WP. Previous studies^{3 45 46} found that older age reduced WP, but other research⁴⁷ suggested that age, in isolation, is not necessarily a barrier to employment; rather, the impact of age on health and daily functioning is what matters.

Only for some scales did the relationship with WP remain significant when entering all scales representing one PH dimension simultaneously. Thus, all PH dimensions are relevant for WP, but each dimension’s different scales show considerable overlap.

Model 2 (table 3) showed fewer complaints, and better general health and physical functioning were associated with higher WP. However, when controlled for sociodemographics and entered simultaneously, only SF-36 General Health, SF-36 Social Functioning and SF-36 Physical Functioning remained significantly related to WP. We hypothesised that the PH dimensions would positively correlate with WP in patients with CGID (H2). This was partly confirmed. Four out of six PH dimensions were related to WP. The relationship with Mental Well-Being disappeared when age was taken into account: older respondents participated less in work.

The type of disorder did not relate significantly to WP. This demonstrates the importance of PH for WP over type of disorder. The importance of PH aligns with the findings from Huber et al⁴⁸ and was further validated in 2022 in the context of mental health.⁴⁹ Our finding that all PH dimensions, except Meaningfulness, are related to WP is new.

Determinants of receiving disability benefit

Regarding determinants of DB, bivariate analyses showed significant associations between all PH dimensions and DB. Hypothesis 3 was confirmed: the worse the participants feel, the more often they receive DB.

When entering all scales representing one dimension simultaneously (table 4, model 2), all PH dimensions were relevant for DB, and the different scales within the dimensions overlapped, just as with WP. We hypothesised that the PH dimensions would correlate negatively with DB in patients with GI disorders (H2). This was confirmed for all six PH dimensions.

In the stepwise model, after accounting for sociodemographics (table 4, model 3), only SF-36 Physical Functioning, higher age and lower level of education remained significantly related to DB. For the explanation of DB, age and education were underlying the other PH dimensions, except for daily functioning. Furthermore, and different from the findings regarding WP, type of disorder had an independent added value in relation to DB: the NGM group was more than four times more likely to receive DB than other subgroups. The NGM group included significantly more women and experienced significantly more physical and mental complaints, and reported significantly lower QoL, physical functioning and social functioning and more physical role limitations (see table 2). Given the limited added explained variance (3.1%), the type of CGID is however of less importance for receiving DB than daily functioning. Curiously, higher education was associated with a lower likelihood of receiving DB. Typically, higher education correlates with higher income, and because the Dutch disability assessment represents loss of income, there is a greater chance of getting DB. An explanation for our findings might be that lower-educated workers might work more often in physically demanding jobs and left unable to continue working when disabled, unlike those in white-collar jobs that require higher education.

Positive health

We can conclude that PH represents the holistic experience of health, but some dimensions overlap. In relation to WP and DB, daily functioning appeared to be the strongest overarching dimension, representing almost all explained variance of the other dimensions. Daily functioning includes evaluating limitations in one’s ability to perform various physical activities, such as running, heavy lifting, vacuuming, climbing stairs, walking and self-grooming. If a person is limited in these basic activities, their daily functioning will be severely restricted, and therefore their ability to work. This aligns with how disability is assessed in the Netherlands: querying the daily story is key in testing claimed limitations.

However, we cannot rule out reversed causality. Not working may lead to poorer health and thus to reduced daily functioning, and in this way to receiving DB more often.

Finally, it might be sufficient to have one scale per PH dimension. We might recommend the GSRS for bodily functions in case of chronic IBD/NGM patients, but to get a more general insight, a more generic questionnaire could be used. The PHQ-9 and GAD-7 for mental well-being, the EQ-5D-5L and SF-36 General Health for the QoL, SF-36 Physical Role Limitations for participation, and SF-36 Physical Functioning for daily functioning could be used to gain insight into the different dimensions.

Strengths, limitations and recommendations for further research

To our knowledge, this is the first study that explored PH dimensions as independent predictors of WP and receiving DB. The importance of the medical specialist adopting a broader perspective than just assessing the physical condition is reiterated here. Our study provides valuable insights into the challenges related to WP and DB among working-age GI patients with different GI subdiagnoses. Results demonstrate that the most evaluated group, the IBD group, has a relatively higher WP rate. This may be due to better organised care in which evidence-based telemedicine already is implemented. This might have led to underestimating the WP difficulties of the whole group of CGID patients and calls for more research on this topic in the HPB and NGM subgroups. Moreover, more research is needed to determine the effects of WP and getting DB on the financial status of patients and their (economic) independence. British research shows that policy makers assume that a decrease in DB leads to an increase in WP, but in fact, this is not the case.^{22 50}

One limitation is the cross-sectional nature of the study, which does not capture longitudinal data and precludes assessing causality. Moreover, the results might be explained by reversed causality. Unemployment is often linked to poorer health and financial difficulties, warranting further exploration of the direction of the relationship between PH and WP and receiving DB, respectively, in future longitudinal research. Data collection overlapped with the COVID-19 period, but pre-pandemic versus during-pandemic data collection was not distinguished, and this heterogeneity might have influenced WP and DB outcomes. The response rate to our study was 47%. No further investigation was conducted into why patients decided not to take part in the study so there could potentially be a selection bias.

This study focused exclusively on patients with CGID. Our findings indicate that disease type had no significant impact on WP and only a limited effect on receiving DB. These results suggest that the study’s conclusions may be generalisable to individuals with other chronic conditions, and so are in line with other studies.³² However, further research is required to validate this assumption.

The PH dimension daily functioning might be a substitute for other correlated dimensions due to the intertwining of parameters from other dimensions. Further research should clarify whether this is explained by psychometric characteristics of the measures or by its content. But this research is beyond the scope of the current research. The strength of this outcome is that with one parameter, it may be possible to make a prediction about WP and DB.

Finally, concerning the role of meaningfulness: the scale was only administered in a subset of participants. Based on our findings, we recommend further studies including the Meaningfulness scale.

Recommendations for practice

Although daily physical functioning is the primary factor associated with both WP and receiving DB, it seems important to discuss all PH dimensions during consultations with patients with CGID to acknowledge the patient’s experiences from a holistic approach. Decuman et al⁴⁵ found that among patients with systemic sclerosis, disability was often associated with limited employment, with over 70% having to leave work or stop studying. But importantly, 70% had not discussed this with their specialist, suggesting that insufficient attention to the consequences of the condition might lead to lower WP. Butink et al⁵¹ developed an intervention that facilitates clinical healthcare professionals providing work-related support to patients. Here, too, the implementation was somewhat reduced by other medical priorities and patients not sharing work-related problems. The earlier during the course of the disease, there is attention for WP, the better it will be for the WP of the patient. Healthcare specialists should regard WP as an important health outcome and encourage patients to openly discuss (work)-related concerns. This does not necessarily require extra time or effort beyond the physician’s medical expertise: during a consultation, the physician can ask questions about work and, if necessary, refer to, for example, the hospital’s occupational therapist to improve daily functioning. PH assessment could be integrated by using the questionnaire by sending it to the patient before the consultation. The consultation itself could then focus on dimensions that are experienced as positive and as problematic compared with norm scores.

Conclusion

Of the working-age patients with CGID, 49.4% had some form of paid work, 20.4% of the participants received DB and the others had other sources of income, mainly unemployment benefit. PH dimensions, and not the GI subgroup, related to both WP and receiving DB; only the NGM group still received DB more often than the other groups after controlling for PH dimensions.

Although we found that the type of GI subgroup was less important to WP and receiving DB, we do not know whether this is generalisable to other patient groups. Patients of different ages and education present different levels of WP and DB: higher age relates to lower WP and higher DB and a lower education relates to higher DB. The PH dimension daily functioning remained stably related to WP and DB, irrespective of controlling for other determinants. Further research on the dominance of this PH dimension is recommended but our findings show the potential of the concept of PH for medical practice and insurance medicine in particular.

Data availability statement

Data are available upon reasonable request.