Functional gastrointestinal disorders, mental health, genetic susceptibility, and incident chronic kidney disease

Authors: Mengyi Liu, Panpan He, Ziliang Ye, Sisi Yang, Yanjun Zhang, Qimeng Wu, Chun Zhou, Yuanyuan Zhang, Fan Fan Hou, Xianhui Qin

Abstract

Background:

Whether functional gastrointestinal disorders (FGIDs) are associated with the long-term risk of chronic kidney disease (CKD) remains unclear. We aimed to investigate the prospective association of FGIDs with CKD and examine whether mental health mediated the association.

Methods:

About 416,258 participants without a prior CKD diagnosis enrolled in the UK Biobank between 2006 and 2010 were included. Participants with FGIDs (including irritable bowel syndrome [IBS], dyspepsia, and other functional intestinal disorders [FIDs; mainly composed of constipation]) were the exposure group, and non-FGID participants were the non-exposure group. The primary outcome was incident CKD, ascertained from hospital admission and death registry records. A Cox proportional hazard regression model was used to investigate the association between FGIDs and CKD, and the mediation analysis was performed to investigate the mediation proportions of mental health.

Results:

At baseline, 33,156 (8.0%) participants were diagnosed with FGIDs, including 21,060 (5.1%), 8262 (2.0%), and 6437 (1.6%) cases of IBS, dyspepsia, and other FIDs, respectively. During a mean follow-up period of 12.1 years, 11,001 (2.6%) participants developed CKD. FGIDs were significantly associated with a higher risk of incident CKD compared to the absence of FGIDs (hazard ratio [HR], 1.36; 95% confidence interval [CI], 1.28–1.44). Similar results were observed for IBS (HR, 1.27; 95% CI, 1.17–1.38), dyspepsia (HR, 1.30; 95% CI, 1.17–1.44), and other FIDs (HR, 1.60; 95% CI, 1.43–1.79). Mediation analyses suggested that the mental health score significantly mediated 9.05% of the association of FGIDs with incident CKD and 5.63–13.97% of the associations of FGID subtypes with CKD. Specifically, the positive associations of FGIDs and FGID subtypes with CKD were more pronounced in participants with a high genetic risk of CKD.

Conclusion:

Participants with FGIDs had a higher risk of incident CKD, which was partly explained by mental health scores and was more pronounced in those with high genetic susceptibility to CKD.

Introduction

Chronic kidney disease (CKD) imposes health and socioeconomic burdens worldwide and may substantially increase the risk of renal failure, cardiovascular disease (CVD), and all-cause mortality.^[1–3]^ The age-standardized global prevalence rates of CKD in adults are 11.8% in women and 10.4% in men, and the prevalence is increasing with the global aging trend and increase in obesity.^[4]^ Therefore, it is important to identify modifiable risk factors for CKD at an early stage to establish primary preventive measures.

Functional gastrointestinal disorders (FGIDs) are a group of disorders characterized by chronic gastrointestinal symptoms (such as abdominal pain, dysphagia, dyspepsia, diarrhea, constipation, and bloating) with no demonstrable pathology or anatomical or physiological abnormalities on conventional testing. Its most common subtypes are irritable bowel syndrome (IBS) and functional dyspepsia.^[5,6]^ FGIDs are very common, with a worldwide prevalence of 40%, accounting for 12% of primary care workloads and 30% of gastroenterology outpatient consultations.^[6]^ Of note, psychiatric disorders such as depression and anxiety often co-occur with FGIDs due to overlapping genetic, physiological, early-life, and psychosocial risk factors that alter the function of the gut–brain axis.^[7,8]^ Since the presence of depression or anxiety was associated with a higher incidence of CKD,^[9,10]^ it is plausible that FGIDs may also be a risk factor for incident CKD, potentially mediated by mental health. However, although the prevalence of FGIDs was higher in participants with renal disease,^[11]^ to our knowledge, until February 2023, only one study has demonstrated that the constipation status and severity were associated with a higher risk of incident CKD.^[12]^ To date, it remains unclear whether other FGIDs are associated with the long-term risk of incident CKD and whether mental health plays a possible mediating role in the association between FGIDs and incident CKD.

To address these aforementioned gaps in knowledge, the present study aimed to evaluate the association of FGIDs with the risk of CKD in the general population and examine the possible effect modifiers of the association using population-based cohort data from nearly half a million adults enrolled in the UK Biobank. In addition, we also examined whether mental health may mediate the association between FGIDs and CKD.

Methods

Data source and study population

The UK Biobank is a large prospective, observational, and population-based cohort of half a million adult residents (aged 37–73 years) in the United Kingdom, with data collected at 22 assessment centers across England, Wales, and Scotland between 2006 and 2010. Participants were asked to complete a touchscreen questionnaire, a face-to-face interview, and a series of physical measurements as well as to provide biological samples for laboratory analysis. Since recruitment, the participants have been followed to determine clinical outcomes via hospital inpatient records, death certificates, and primary care records. A detailed description of the study design has been provided previously.^[13,14]^ The UK Biobank was approved by the North West Research Ethics Committee (No.06/MRE08/65), and all the participants signed informed consent forms.

In this study, we restricted our analysis to participants with complete information available on kidney function. Of the 453,688 participants, those with prior CKD (determined by a self-reported history of CKD, a CKD diagnosis prior to the date of baseline assessment, an estimated glomerular filtration rate [eGFR] <60 mL·min^–1^·1.73 m^–2^, or a urine albumin-to-creatinine ratio [UACR] ≥30 mg/g) (n = 32,770) were excluded. Additionally, considering the overall similarity in the symptoms of inflammatory bowel disease (IBD) and IBS, as well as the potential challenge of differential diagnosis between IBD and IBS, we further excluded 4660 participants who had a previous IBD diagnosis prior to baseline. Therefore, a total of 416,258 participants were enrolled in the present analysis [Supplementary Figure 1, http://links.lww.com/CM9/B664].

Definition of FGIDs

FGIDs, composed of IBS, dyspepsia, and other functional intestinal disorders (FIDs), were ascertained by self-report data and data linkage with primary care, hospital admissions, and death registry records based on the International Classification of Diseases, 10th revision (ICD-10), codes of K58 (IBS), K30 (dyspepsia), and K59 (other FIDs). Individuals with a diagnosis of FGIDs prior to baseline were considered the exposure group, and those without a diagnosis of FGIDs prior to baseline were considered the non-exposure group, regardless of whether they were diagnosed during the follow-up.

Ascertainment of covariates

Detailed information on covariates was collected through standardized questionnaires, including age, sex, race, Townsend Deprivation Index (TDI), smoking status, alcohol consumption, and comorbidities (hypertension, diabetes, and high cholesterol). Body mass index (BMI) was calculated as weight (kg)/height (m)^2^. Prevalent hypertension was defined as a self-reported history of hypertension, the use of antihypertensive drugs, a systolic blood pressure of 140 mmHg or higher, or a diastolic blood pressure of 90 mmHg or higher. Prevalent diabetes at baseline was identified through multiple procedures considering the type of diabetes and sources of the diagnosis.^[15]^ Optimal physical activity was defined as more than 4 days of vigorous/moderate physical activity in a typical week.^[16]^ A healthy diet score was evaluated using a recent dietary recommendation for cardiovascular health, which considered adequate consumption of fruits, vegetables, whole grains, fish, shellfish, dairy products, and vegetable oils and reduced consumption of refined grains, processed meats, unprocessed meats, and sugar-sweetened beverages.^[17]^ Serum creatinine levels and the UACR were measured at a dedicated central laboratory, and the eGFR was calculated by the Chronic Kidney Disease Epidemiology Collaboration equation.^[18]^

Information on psychological factors and mental health was collected through a touchscreen questionnaire. The mental health score was calculated by adding the participant's answers to the Eysenck Personality Questionnaire-Revised Short Form items, which included the mood swings, misery, irritability, sensitivity to hurt feelings, frustration, nervousness, worry, being tense/highly strung, excessive worry after embarrassment, suffering from nerves, feelings of loneliness/isolation, and feelings of guilt; additionally, participant's responses regarding risk-taking were recorded.^[19]^ Higher scores represented more mental health-related symptomatology.

Definition of genetic risk scores of CKD

Detailed information about genotyping and quality control in the UK Biobank study has been provided previously.^[20]^ The genetic risk score (GRS) of eGFR was calculated using a weighted method including 263 single nucleotide polymorphisms (SNPs) that showed independent, significant genome-wide association with eGFR.^[21]^ A higher score represents a lower genetic risk of CKD. Participants were divided into low, intermediate, or high genetic risk for CKD according to the tertiles of eGFR GRS values.

Ascertainment of CKD

The primary outcome of the study was the incidence of CKD, which was identified using linkage with hospital admission and death registry records data according to ICD-9 codes, ICD-10 codes, and the Office of Population Censuses and Surveys Classification of Interventions and Procedures-4 (OPCS-4) code [Supplementary Table 1, http://links.lww.com/CM9/B664]. The follow-up person-time for each participant was calculated from the date of the first assessment until the date of death, the first date of outcome diagnosis, the date of loss to follow-up, or the end of follow-up (September 30, 2021 for centers in England, July 31, 2021 for centers in Scotland, and February 28, 2018 for centers in Wales), whichever came first.

Statistical analysis

Population characteristics are presented as the mean ± standard deviation if normally distributed and medians (interquartile ranges [IQR]) if not normally distributed for continuous variables and as proportions for categorical variables. Comparisons of characteristics according to FGID status (yes or no) were performed by chi-squared tests for categorical variables and t-tests or Mann–Whitney U tests, as appropriate, for continuous variables.

Cox proportional hazards regression models were used to estimate the hazard ratio (HR) and 95% confidence interval (CI) of incident CKD for individuals with FGIDs and FGID subtypes. In multivariable models, potential confounders that are traditional or suspected risk factors for CKD were adjusted for, including age, sex, race, TDI, BMI, smoking status, alcohol consumption, healthy diet score, physical activity, comorbidities (hypertension, diabetes, and high cholesterol), and baseline renal function (the eGFR and UACR).

Mediation analysis was performed using the mediation package to investigate the extent to which mental health contributed to the FGID-CKD association. In the mediation analysis, two models with adjustments for the same covariates included in the above Cox proportional hazards regression model were first, a multivariate linear regression model for the association of mediator (the mental health score) with the exposure (FGIDs and FGID subtypes), and second, a multivariate Cox regression model for the association of the mediator and exposures with the outcome (incident CKD). The direct effects (DE) represent the effect of exposures on the outcome, independent of the mediator, while the indirect effects (IE) represent the proportion of the exposure that can be explained by its association with the mediator. To quantify the magnitude of the mediation effect, the mediation proportions (IE/[DE + IE]) were estimated. Additionally, we estimated the joint effect of FGIDs and CKD-GRS on the risk of incident CKD.

As an exploratory analysis, stratified analysis was conducted to assess potential modifiers of the association between FGIDs and incident CKD according to age (<60 years or ≥60 years), sex (female or male), TDI (<–2.2 or ≥–2.2), BMI (<30 kg/m^2^ or ≥30 kg/m^2^), smoking status (never or ever), alcohol consumption (<1 time/week or ≥1 times/week), diabetes (yes or no), hypertension (yes or no), high cholesterol (yes or no), healthy diet score (<4 or ≥4), and optimal physical activity (yes or no). Potential modifying effects were assessed by modeling the cross-product term of the stratifying variable with FGIDs.

Several sensitivity analyses were performed. First, we further excluded participants who had a diagnosis of FGIDs after recruitment from the non-exposure group. Second, we further excluded participants who developed CKD in the first 2 years of follow-up. Third, we restricted the analyses to a subsample of participants who completed the ancillary online gastrointestinal health self-assessment questionnaire issued in 2017 and defined the control group as participants without self-reported IBS or abdominal discomfort or pain in the last 3 months. Fourth, given that the influence of FGIDs may be small with a short time interval between exposure and outcome, we excluded participants with a diagnosis of FGIDs 2 years before baseline and those with less than 2 years of follow-up to ensure that there was at least a 4-year time gap between exposure and outcome. Fifth, incident CKD (field ID: 132033) was defined by the well-validated algorithm of "first occurrence" data fields across self-report, primary care, hospital inpatient data, and death data mapped to a 3-digit ICD-10 code in the UK Biobank. Sixth, we further excluded participants with a previous history of peptic ulcer disease, intestinal tuberculosis, or chronic liver disease.

A two-tailed P <0.05 was considered to be statistically significant in all analyses. Analyses were performed using R 4.1.1 software (R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org).

Results

Study participants and baseline characteristics

Of the 416,258 participants in the current study, the mean age was 56.3 ± 8.1 years and 224,837 (54.01%) were females. Overall, 33,156 (7.97%) participants had a diagnosis of FGIDs before or at enrollment (median [interquartile range] of time since diagnosis at 8.2 [3.8–14.2] years), including 21,060 (5.1%), 8262 (2.0%), and 6437 (1.6%) cases of IBS, dyspepsia, and other FIDs, respectively. As shown in Table 1, participants with FGIDs were older, more likely to be female, and more likely to be White; had higher TDI and BMI values; had lower alcohol intake, physical activity, and eGFR; had higher UACR values, healthy diet scores, and mental health scores, and had a lower prevalence of hypertension and higher prevalence of high cholesterol.

Association of FGIDs and FGID subtypes with risk of incident CKD

During a mean follow-up period of 12.1 years, a total of 11,001 (2.6%) participants developed CKD (median [interquartile range] time to CKD: 9.3 [6.7–11.1] years).

FGIDs were significantly associated with a higher risk of incident CKD compared to the absence of FGIDs (adjusted HR, 1.36; 95% CI, 1.28–1.44) [Table 2]. Similar results were observed for FGID subtypes (IBS: adjusted HR, 1.27, 95% CI, 1.17–1.38; adjusted HR, 1.30, 95% CI, 1.17–1.44; other FID: adjusted HR, 1.60, 95% CI, 1.43–1.79 [Table 2]. After Bonferroni correction for multiple comparisons, P <0.01 was set to be statistically significant, and the relationships of FGIDs and FGID subtypes with the risk of CKD remained significant.

Mediation analysis of the mental health score

The mediation analysis showed that the mental health score significantly mediated the association of FGIDs (mediation proportion, 9.05%; 95% CI, 6.59%–12.09%) with incident CKD [Figure 1]. Moreover, the mental health score significantly mediated the associations of IBS, dyspepsia, and other FIDs with incident CKD, with a mediating proportion of 13.97% (95% CI, 9.61%–22.95%), 8.27% (5.51%–13.81%), and 5.63% (4.02%–8.03%), respectively [Figure 1].

Joint effect of FGIDs and FGID subtypes with genetic risks of CKD on the risk of incident CKD

Although the effect of the interaction between FGIDs and CKD GRS values on the risk of CKD was not significant (P-interaction = 0.078; Figure 2A), compared with participants without FGIDs and with a low genetic risk of CKD, participants with FGIDs and a high genetic risk of CKD had the highest risk of CKD (HR, 1.58; 95% CI, 1.44–1.74; Figure 2A). Similar trends were found for IBS [Figure 2B], dyspepsia [Figure 2C], and other FIDs [Figure 2D].

Stratified analyses of the association between FGIDs and incident CKD

None of the variables, including age, sex, TDI, BMI, smoking status, alcohol consumption, diabetes, hypertension, high cholesterol, healthy diet score, and optimal physical activity, significantly influenced the association of FGIDs with incident CKD (all P for interaction ≥0.05; Supplementary Table 2, http://links.lww.com/CM9/B664).

Sensitivity analyses of the association between FGIDs and incident CKD

In the sensitivity analyses, the results did not change substantially after excluding participants who had a diagnosis of FGIDs after recruitment in the non-exposure group (sensitivity analysis 1); excluding participants who developed CKD in the first 2 years of follow-up (sensitivity analysis 2); restricting analyzed data to a subsample of participants who completed the ancillary online gastrointestinal health questionnaire issued in 2017 and defining the control group as participants without self-reported IBS or abdominal discomfort or pain in the last 3 months (sensitivity analysis 3); excluding those with a FGIDs diagnosis 2 years before baseline and those with less than 2 years of follow-up (sensitivity analysis 4); defining incident CKD according to the well-validated algorithm of "first occurrence" data fields across self-report, primary care, hospital inpatient data, and death data mapped to a 3-digit ICD-10 code in the UK Biobank (sensitivity analysis 5); or excluding participants with a previous history of peptic ulcer disease, intestinal tuberculosis, or chronic liver disease (sensitivity analysis 6) [Supplementary Table 3, http://links.lww.com/CM9/B664].

Discussion

In this large-scale population-based cohort, FGIDs and FGID subtypes were significantly associated with a higher risk of incident CKD in the general population, and the observed positive associations were partly mediated by mental health. Furthermore, the positive association between FGIDs and incident CKD was more pronounced in participants with a higher genetic risk of CKD.

A cohort study conducted among US veterans reported that constipation, defined based on diagnostic codes and laxative use, was associated with a higher risk of CKD (HR, 1.13; 95% CI, 1.11–1.14) than non-constipation.^[12]^ Our current study, which included all FGID subtypes (IBS, dyspepsia, and other FIDs [mainly composed of constipation]), demonstrated that all FGID subtypes were independently and significantly associated with the risk of incident CKD. Interestingly, we also found that there was a stronger association between FGIDs and incident CKD in participants with higher genetic risks of CKD. That is, FGIDs and high genetic risks of CKD may jointly increase the risk of incident CKD.

Another finding worth mentioning is that mental health may mediate the relationship between FGIDs and incident CKD. FGIDs occur as a result of the interaction of biological, psychological, and social factors in a biopsychosocial framework.^[22]^ Previous studies have shown that individuals with FGIDs have a higher prevalence of psychiatric disorders,^[7,8]^ and IBS has even shown a strong genome-wide overlap with psychological traits^[23]^ such as anxiety, neuroticism, depression, and schizophrenia. In addition, considering that the presence of depression or anxiety was associated with a higher incidence of CKD,^[9,10]^ the relationship between FGIDs and CKD may exist through an indirect link with mental health. Consistent with this idea, further mediation analysis showed that the mental health score significantly explained 9.05% and 5.63–13.97% of the associations of FGIDs and FGID subtypes with CKD, respectively.

Moreover, emerging evidence suggests that the disruption of the normal gut microbiota is linked to the development of various diseases through chronic inflammation and/or altered metabolite profiles.^[24–26]^ In the metabolism-dependent pathway, dysbiosis leads to the overproduction and accumulation of p-cresyl and indoxyl sulfates in the intestines, which may disrupt the gut barrier and increase gut permeability, resulting in the accumulation of endotoxins and uremic toxins, and leading to renal fibrosis, inflammation, and oxidative stress as well as the accelerated progression of kidney disease.^[24,25,27]^ In the immune pathway, bone marrow-derived mature immune cells in the gut are overactivated by the dysbiotic microbiota, followed by the production of inflammatory cells, cytokines, and soluble urokinase plasminogen activator surface receptor in the gut, which are associated with the onset and progression of CKD.^[24,28]^ Given that FGIDs encompass a multifaceted complex of changes in the gut microbiota and immunology,^[5,29]^ these pathophysiological mechanisms could be the potential explanation for the association between FGIDs and CKD. However, other underlying mechanisms may also be involved and need to be further examined.

Several limitations should be mentioned. First, although we carefully controlled for potential confounding factors and performed a range of stratified analyses, the observed associations might be driven by unmeasured or unknown residual confounders. Second, FGIDs and CKD were identified by ICD codes across several sources, including self-reports, primary care, hospital inpatient data, and death data, rather than by structured questionnaires (i.e., Rome III or IV scales), interviews, or repeated biological measurements. Thus, the exact diagnostic data for each FGID and CKD case were unavailable. However, our definitions of FGIDs and CKD were similar to those used in previous studies analyzing data from the UK Biobank.^[30–33]^ Third, as functional conditions, FGIDs might be underdiagnosed and might not be adequately identified in the UK Biobank, which may lead to potential misclassification and attenuate the results, thereby underestimating the true associations. Indeed, in a sensitivity analysis that defined the control group as participants without IBS or abdominal discomfort or pain over the past 3 months, we observed a larger effect-size estimate of the association between FGIDs and incident CKD (sensitivity analysis 3 in Supplementary Table 3, http://links.lww.com/CM9/B664). Fourth, although we observed similar results for the main subtypes of FGIDs (i.e., IBS and dyspepsia), a cautious interpretation is warranted given that FGIDs are a relatively broad set of gastrointestinal conditions. Owing to these limitations, further confirmation of the reported findings in future studies is necessary.

In conclusion, this large-scale prospective study showed that participants with FGIDs had a higher risk of incident CKD, which was explained in part by mental health scores. The positive association of FGIDs and incident CKD was more pronounced in those with a high genetic susceptibility to CKD. Given the high prevalence of FGIDs, we emphasize the necessity and importance of monitoring renal function trajectory in participants with FGIDs.

Acknowledgments

This research has been conducted using the UK Biobank Resource under application number 73201. We specially thank all the participants of UK Biobank and all the people involved in building the UK Biobank study.

Funding

The study was supported by the National Key Research and Development Program (Nos. 2022YFC2009600 and 2022YFC2009605), National Natural Science Foundation of China (Nos. 81973133 and 81730019), National Natural Science Foundation of China (Key Program) (No. 82030022), Program of Introducing Talents of Discipline to Universities, 111 Plan (No. D18005), Guangdong Provincial Clinical Research Center for Kidney Disease (No. 2020B1111170013), and Key Technologies R&D Program of Guangdong Province (No. 2023B1111030004).

Availability of data and material

The UK Biobank data are available on application to the UK Biobank, and the analytic methods and study materials that support the findings of this study will be available from the corresponding authors on request.

Conflicts of interest

None.

Supplementary Material