Association between “Life’s Critical 9” health score and the risk of inflammatory bowel disease: a prospective cohort study from the UK biobank

Abstract

Background

Inflammatory bowel disease (IBD), including ulcerative colitis (UC) and Crohn’s disease (CD), is a chronic inflammatory condition of the gastrointestinal tract. The “Life’s Critical 9” (LC9) health score, proposed by the American Heart Association (AHA), incorporates nine modifiable factors including psychological health. This study aims to evaluate the association between the LC9 score and the risk of IBD, particularly UC and CD.

Methods

This prospective cohort study utilized UK Biobank data from 370,645 participants. Cox proportional hazards models assessed associations between LC9 score and IBD, UC, and CD risk. Subgroup analyses examined interactions with age, sex, ethnicity, education, smoking, alcohol consumption, and body mass index (BMI). Mediation analyses evaluated the role of triglycerides, neutrophils, albumin, and C-reactive protein (CRP). Sensitivity analyses replaced the overall LC9 score with the mental-psychological subscale to evaluate robustness.

Results

Higher LC9 scores were significantly associated with reduced UC risk (hazard ratio [HR] 0.909, 95% confidence interval [CI] 0.863–0.958) and overall IBD risk (HR 0.976, 95% CI 0.953–1.001). No significant association was observed for CD (HR 0.928, 95% CI 0.858–1.003, P = 0.059). Subgroup analyses revealed stronger inverse associations for UC among individuals aged ≥ 60 years, females, European descent, lower education levels, current drinkers, and BMI 18.5–25. Mediation analyses showed triglycerides, neutrophils, albumin, and CRP accounted for 10.78%, 26.74%, 4.77%, and 22.95% of the LC9-UC association, respectively. Sensitivity analyses using the mental-psychological subscale demonstrated consistent protective associations across all models for IBD (HR 0.991, 95% CI 0.990–0.992), UC (HR 0.995, 95% CI 0.993–0.997), and CD (HR 0.996, 95% CI 0.994–0.998), all P < 0.001.

Conclusion

Higher LC9 scores were significantly associated with reduced UC risk, with notable interactions across demographic and behavioral characteristics. The comprehensive LC9 framework, incorporating cardiovascular, metabolic, and psychological factors, may provide a holistic approach to IBD risk assessment and prevention strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26434-0.

Introduction

Inflammatory bowel disease (IBD), encompassing ulcerative colitis (UC) and Crohn’s disease (CD), is recognized as a chronic and relapsing inflammatory disorder affecting the gastrointestinal tract, with millions of new cases diagnosed annually worldwide and over 3.2 million cases reported in Europe alone [1]. The pathogenesis of IBD is complex and not yet fully understood; however, genetic predisposition, environmental factors, and immune responses are known to play critical and interrelated roles in its development and progression [2]. Given the projected substantial increase in the disease burden of IBD over the coming decades, identifying modifiable factors for its prevention holds significant public health importance.

In 2010, the American Heart Association (AHA) introduced “Life’s Simple 7,” a set of seven modifiable health behaviors and factors aimed at promoting cardiovascular health (CVH) [3]. In 2022, the AHA expanded upon this concept by incorporating sleep health and redefining the parameters as “Life’s Essential 8” (LE8), which includes diet, physical activity, nicotine exposure, sleep health, body mass index (BMI), blood lipids, blood glucose, and blood pressure [4]. Previous studies have demonstrated a close association between adherence to LE8 and reduced incidence of all-cause mortality, cardiovascular disease (CVD), and various chronic conditions linked to immune regulation [5, 6]. A recent study published in Circulation suggested that psychological health could be considered as an addition to LE8, proposing a new framework termed the “Life’s Critical 9” (LC9) [7]. However, this has not yet been formally incorporated into the established cardiovascular health metrics. Psychological health is regarded as a fundamental component in maintaining CVH, with research indicating that depression and anxiety are significant risk factors for cardiovascular diseases [8]. IBD, including UC and CD, frequently coexists with psychological health issues such as depression and anxiety. Therefore, incorporating psychological health into the LC9 scoring system may facilitate a more comprehensive assessment of the relationship between these conditions and overall health.

The relationship between the LC9 score and IBD, particularly UC and CD, has not been extensively investigated to date. Therefore, this study aims to further quantify the psychological health status score, which includes symptoms of depression and anxiety as components of the LC9, by utilizing a large prospective cohort from the UK Biobank. This will elucidate the complex relationship between the LC9 score and the risk of developing IBD, including UC and CD. Additionally, the study seeks to explore the potential mediating role of specific biomarkers, such as triglycerides, neutrophils, albumin, and C-reactive protein (CRP), in the relationship between the LC9 score and UC.

Methods

Data resource and study population

The UK Biobank is a large-scale prospective cohort study that conducted baseline assessments of over 500,000 participants aged 40 to 69 years at 22 assessment centers across the United Kingdom between 2006 and 2010. The detailed study protocol has been comprehensively described in previous reports [9]. This extensive database collected a wide range of baseline data through questionnaires, face-to-face interviews, physical measurements, and the collection of blood and urine samples. The data encompass a broad spectrum of sociodemographic information, electronic health records, physical examinations, genetic biospecimens, and imaging data. This study was conducted in accordance with the Declaration of Helsinki. The UK Biobank study received ethical approval from the North West Multi-Centre Research Ethics Committee (REC reference 11/NW/0382), and informed consent was obtained from all participants using a touchscreen signature capture device. The analyses in this study were conducted under the UK Biobank application ID 84,347.

This study includes data from a total of 502,411 participants. A total of 110,851 participants were excluded due to the inability to calculate LC9 resulting from missing relevant data; an additional 20,915 participants were excluded due to missing baseline covariate data. Ultimately, 370,645 participants were included in the final analysis. Further specific details can be found in Fig. 1.

Fig. 1.

Flowchart of the Experimental Design. Note: Abbreviations: BMI, Body Mass Index; CD, Crohn’s disease; IBD, Inflammatory Bowel Disease; LC9, Life’s Critical 9; TDI, Townsend Deprivation Index; UC, Ulcerative Colitis

Assessment of LC9

The LC9 algorithm is composed of five health factors—BMI, blood glucose, blood lipids, blood pressure, and mental health—and four health behaviors: diet, physical activity, sleep health, and nicotine exposure. The calculation of LC9 was performed using the updated definitions provided by the American Heart Association [7]. Detailed scoring criteria are presented in Supplementary Table 1. Each CVH metric was assigned a score ranging from 0 to 100 points, and the final LC9 score was calculated as the arithmetic mean of the nine CVH metrics. Participants with LC9 scores between 80 and 100 points were categorized as the high CVH group, those with scores between 0 and 49 points were categorized as the low CVH group, and scores between 50 and 79 points defined the intermediate CVH group. In this study, we applied the same cutoff points for the evaluation of LC9 scores.

In the UK Biobank, depression and anxiety symptoms at baseline were assessed using the Patient Health Questionnaire (PHQ)-4 [10]. During the follow-up surveys, depression and anxiety were evaluated using the PHQ-9 (an updated version of PHQ-4) [11] and the Generalized Anxiety Disorder (GAD)-7 questionnaires [12], respectively. Dietary quality was assessed using the Healthy Eating Index 2015 (HEI-2015). The construction of this index utilized dietary intake data obtained from two 24-hour dietary recall questionnaires completed by participants, which were then combined with food pattern equivalents data from the United States Department of Agriculture (USDA). The HEI-2015 score was calculated using SAS code provided by the National Cancer Institute [13]. Information on participants’ blood glucose levels, glycated hemoglobin, and blood lipids was obtained from their hematology samples. Data on height, weight, blood pressure, and other relevant anthropometric measurements were obtained from participants’ physical examination reports. Information regarding physical activity, smoking history, diabetes history, medication use, and sleep patterns was collected through self-reported questionnaires. Specific details on the calculation of LC9 can be found in Table S1. For further information, please refer to https://biobank.ndph.ox.ac.uk/.

Definition of IBD

The identification of CD and UC was based on the International Classification of Diseases, Tenth Revision (ICD-10) codes. The primary endpoint of this study is IBD. Participants were categorized into three subgroups: ‘CD,’ ‘UC,’ and ‘IBD-U’ based on specific diagnostic criteria. The term IBD-unclassified (IBD-U) refers to cases where the type of inflammatory bowel disease could not be clearly determined. Diagnostic records were sourced from hospital inpatient records, primary care databases, and self-reported medical histories. To enhance diagnostic validity, IBD case confirmation required at least one corresponding ICD-10 diagnostic code (CD: K50; UC: K51; IBD-U: K52.9) documented in either hospital inpatient records or primary care databases. Cases identified solely through self-report without corroborating medical records were excluded from the analysis. Participants were recruited between 2006 and 2010, with the date of baseline assessment serving as the start of follow-up for each individual. Follow-up continued until the diagnosis of IBD (CD, UC, or IBD-U), withdrawal from the study, death, or the censoring date of December 31, 2022, whichever occurred first. The median follow-up duration was 13.01 years.

Ascertainment of covariates

In this study, we adjusted for the following covariates, which were selected based on established potential risk factors for IBD. Given that the LC9 already encompasses various health factors and behaviors, we limited our adjustments to the following variables to avoid over-adjustment, increased model complexity, and the potential for multicollinearity: sex, age, ethnicity (European, Asian, African, Chinese, Mixed-race, Others), educational attainment indicators, Townsend Deprivation Index (TDI), and drinking status (Current, Never, Previous). Educational attainment indicators were derived from data related to “Child and Adolescent” and “Adult Skills,” which reflect the “flow” and “stock” of educational disadvantage within a region, respectively. The TDI is a measure of socioeconomic deprivation, used to assess the level of deprivation in material and social resources among residents of a particular area, thus reflecting the socioeconomic status of the region. This information was primarily collected via touchscreen interfaces at the assessment center. Missing data for continuous covariates were imputed using mean substitution, and continuous covariates were standardized to align their scales with that of the primary exposure variables.

Statistical analysis

In this study, continuous baseline variables were expressed as mean ± SD and compared using t-tests and Wilcoxon rank-sum tests, while categorical variables were presented as numbers (percentages) and compared using chi-square tests. Multiple Cox proportional hazards models were constructed to evaluate the association between LC9 (both continuous and categorical variables) and IBD, using HR and 95% CI, and trend tests were performed to assess the presence of linear trends between these variables. Model 1 did not adjust for any confounders; Model 2 adjusted for age, race, and sex; Model 3 further adjusted for educational level and the TDI based on Model 2. Most exposures and covariates were assessed only at baseline. For psychological health, the PHQ-4 was administered at baseline, while PHQ-9 and GAD-7 were used during follow-up surveys. Additionally, restricted cubic spline (RCS) models were constructed to further evaluate the dose-response relationship between LC9 and IBD, including UC, CD, and IBD-U. Sensitivity analyses were conducted by replacing the overall LC9 score with its mental-psychological subscale to evaluate the robustness of the primary associations.

Subgroup analyses were conducted based on age, gender, race, educational attainment, smoking status, alcohol consumption status, and BMI (< 18.5, 18.5–25, ≥ 25), and multiplicative interactions between LC9 and these variables were assessed. Furthermore, mediation analyses were performed treating LC9 as a continuous variable to evaluate whether triglycerides, neutrophils, albumin, and CRP mediated the relationship between LC9 and UC. The total effect (TE) represents the overall relationship between LC9 and UC, which is the sum of the direct and indirect effects. The direct effect (DE) represents the influence of LC9 on UC that does not pass through the mediating variables (triglycerides, neutrophils, albumin, and CRP). The indirect effect (IE) indicates the influence of LC9 on UC that operates through these mediating variables. A significant IE indicates a significant mediation effect, with the proportion of mediation calculated as (IE/TE) × 100%.

All statistical analyses and visualizations were performed using R (version 4.4.1). All statistical tests were two-sided, and a p-value of < 0.05 was considered statistically significant.

Results

Baseline characteristics

Table 1 summarizes the baseline characteristics of participants with incident IBD, CD, and UC. The median follow-up period in this study was 13.01 years, during which there were 15,780 cases of IBD, 1,548 cases of CD, and 3,410 cases of UC. Compared to participants in the low CVH group (LC9 score ranging from 0 to 50), those in the moderate and high CVH groups were more likely to be female, younger in age, of mixed-race, Asian, or Chinese descent, have lower educational scores, lower TDI scores, have a history of never or former smoking, never consume alcohol, have a lower BMI, and have lower prevalence rates of UC and IBD. Supplementary Table 2 provides detailed baseline characteristics comparing participants without IBD (No IBD) and those diagnosed with CD and UC. Compared to participants without IBD, those with UC had a higher proportion of males (53% vs. 46%), were older (mean age 57.6 vs. 56.5 years), more likely to be former smokers (46% vs. 35%), and had lower mean LC9 scores (539.4 vs. 550.5). Participants with CD showed similar patterns, with a notably higher proportion of current smokers (16% vs. 10%) and lower physical activity scores (43.3 vs. 48.9) compared to those without IBD. Notably, psychological health scores were higher (indicating poorer psychological health) in both CD (33.9 vs. 29.9) and UC (33.6 vs. 29.9) groups compared to the No IBD group.

Table 1.

Baseline characteristics stratified by life’s critical 9 health score

Characteristic	Overall, N = 370,645	Low CVH [0,50), N = 49,090	Moderate CVH [50,80), N = 312,223	High CVH[80,101), N = 9,332	P-value*
Sex, n (%)					< 0.001
Female	199,423 (54%)	19,294 (39%)	172,612 (55%)	7,517 (81%)
Male	171,222 (46%)	29,796 (61%)	139,611 (45%)	1,815 (19%)
Age, Mean (SD)	56.572 (8.076)	58.579 (7.420)	56.354 (8.109)	53.317 (8.220)	< 0.001
Race, n (%)					< 0.001
European	352,478 (95%)	46,989 (96%)	296,505 (95%)	8,984 (96%)
Mixed-race	2,117 (0.6%)	269 (0.5%)	1,785 (0.6%)	63 (0.7%)
Asian	6,942 (1.9%)	610 (1.2%)	6,193 (2.0%)	139 (1.5%)
African	4,887 (1.3%)	804 (1.6%)	4,039 (1.3%)	44 (0.5%)
Chinese	1,138 (0.3%)	58 (0.1%)	1,044 (0.3%)	36 (0.4%)
Others	3,083 (0.8%)	360 (0.7%)	2,657 (0.9%)	66 (0.7%)
Educational Score, Mean (SD)	14.561 (16.004)	17.955 (18.033)	14.120 (15.660)	11.442 (13.455)	< 0.001
TDI, Mean (SD)	-1.402 (3.004)	-0.943 (3.173)	-1.462 (2.977)	-1.803 (2.780)	< 0.001
Smoking status, N (%)					< 0.001
Never	203,083 (55%)	6,398 (13%)	187,943 (60%)	8,742 (94%)
Previous	129,676 (35%)	39,396 (80%)	90,246 (29%)	34 (0.4%)
Current	37,886 (10%)	3,296 (6.7%)	34,034 (11%)	556 (6.0%)
Drinking status, n (%)					< 0.001
Never	15,806 (4.3%)	1,510 (3.1%)	13,854 (4.4%)	442 (4.7%)
Previous	12,948 (3.5%)	2,124 (4.3%)	10,485 (3.4%)	339 (3.6%)
Current	341,891 (92%)	45,456 (93%)	287,884 (92%)	8,551 (92%)
BMI, Mean (SD)	27.397 (4.746)	31.520 (5.214)	26.875 (4.334)	23.178 (2.608)	< 0.001
LC9, Mean (SD)	550.163 (89.592)	401.925 (36.407)	567.767 (66.744)	740.979 (20.321)	< 0.001
HEI-2015 diet score , Mean (SD)	71.332 (21.544)	60.265 (22.288)	72.626 (20.949)	86.267 (14.641)	< 0.001
Physical activity score, Mean (SD)	48.737 (34.718)	23.612 (27.822)	51.665 (33.990)	82.910 (19.287)	< 0.001
Tobacco/nicotine exposure score, Mean (SD)	60.069 (45.161)	16.766 (34.577)	65.796 (43.167)	96.236 (10.097)	< 0.001
Sleep health score, Mean (SD)	89.115 (18.841)	80.343 (24.277)	90.265 (17.633)	96.796 (9.452)	< 0.001
Body mass index score, Mean (SD)	68.964 (28.544)	43.384 (27.332)	72.216 (26.674)	94.732 (12.873)	< 0.001
Blood lipids score, Mean (SD)	58.703 (10.644)	59.461 (11.436)	58.553 (10.513)	59.726 (10.439)	< 0.001
Blood glucose score, Mean (SD)	42.185 (9.099)	42.859 (9.398)	42.074 (8.984)	42.360 (10.962)	< 0.001
Blood pressure score, Mean (SD)	39.467 (34.152)	17.061 (22.324)	41.571 (33.811)	86.953 (20.271)	< 0.001
Psychological health score, Mean (SD)	69.334 (28.940)	81.917 (25.978)	68.077 (28.754)	44.908 (23.897)	< 0.001
IBD, n (%)					< 0.001
CD	1,548 (0.4%)	234 (0.5%)	1,269 (0.4%)	45 (0.5%)
IBD	15,780 (4.3%)	2,328 (4.7%)	13,033 (4.2%)	419 (4.5%)
None	349,907 (94%)	45,955 (94%)	295,182 (95%)	8,770 (94%)
UC	3,410 (0.9%)	573 (1.2%)	2,739 (0.9%)	98 (1.1%)

*P-values were calculated using chi-square (χ2) tests for categorical variables and one-way Analysis of Variance (ANOVA) for continuous variables to compare differences across low CVH, moderate CVH, and high CVH groups. The ‘Overall’ column is presented for descriptive purposes only and was not included in statistical comparisons. Abbreviations: BMI Body Mass Index, CD Crohn’s Disease, CVH Cardiovascular Health, IBD Inflammatory Bowel Disease, LC9 Life’s Critical 9, SD Standard Deviation, TDI Townsend Deprivation Index, UC Ulcerative Colitis

Association between LC9 score and the risk of incident UC, CD, and IBD

Table 2 presents the associations between LC9 scores and the risk of UC, CD, and IBD. For UC, when LC9 was examined as a continuous exposure, all three models demonstrated a significant inverse association with UC risk (Model 1: HR 0.872, 95% CI 0.830–0.915; Model 2: HR 0.907, 95% CI 0.862–0.955; Model 3: HR 0.909, 95% CI 0.863–0.958; all P < 0.001). When LC9 was analyzed as a categorical variable, compared to the low CVH group, the moderate CVH group showed a significant risk reduction across all models (Model 1: HR 0.755, 95% CI 0.670–0.851; Model 2: HR 0.804, 95% CI 0.712–0.907; Model 3: HR 0.805, 95% CI 0.712–0.910). However, the high CVH group did not show statistically significant differences compared to the low CVH group in any model (Model 1: HR 0.948, 95% CI 0.718–1.251, P = 0.705; Model 2: HR 1.114, 95% CI 0.841–1.476, P = 0.453; Model 3: HR 1.146, 95% CI 0.862–1.523, P = 0.349). Despite this, trend tests confirmed a significant dose-response relationship across all models (P for trend: Model 1 = 0.001, Model 2 = 0.027, Model 3 = 0.036).

Table 2.

Hazard ratios (95% CIs) of the risk of incident UC, CD and IBD in relation to LC9

		Continuous		Low CVH (0–49)	Moderate CVH (50–79)		High CVH (80–100)
		HR 95%CI	P-value		HR 95%CI	P-value	HR%95CI	P-value	P for trend
UC	Model 1	0.872(0.830 ~ 0.915)	< 0.001	Ref.	0.755(0.670 ~ 0.851)	< 0.001	0.948(0.718 ~ 1.251)	0.705	0.001
	Model 2	0.907(0.862 ~ 0.955)	< 0.001	Ref.	0.804(0.712 ~ 0.907)	< 0.001	1.114(0.841 ~ 1.476)	0.453	0.027
	Model 3	0.909(0.863 ~ 0.958)	< 0.001	Ref.	0.805(0.712 ~ 0.910)	0.001	1.146(0.862 ~ 1.523)	0.349	0.036
CD	Model 1	0.912(0.847 ~ 0.981)	0.014	Ref.	0.867(0.720 ~ 1.045)	0.135	1.051(0.690 ~ 1.601)	0.818	0.358
	Model 2	0.904(0.838 ~ 0.976)	0.010	Ref.	0.868(0.718 ~ 1.050)	0.144	1.042(0.680 ~ 1.598)	0.849	0.361
	Model 3	0.928(0.858 ~ 1.003)	0.059	Ref.	0.904(0.745 ~ 1.097)	0.305	1.146(0.746 ~ 1.761)	0.535	0.684
IBD	Model 1	0.956(0.934 ~ 0.979)	< 0.001	Ref.	0.874(0.824 ~ 0.928)	< 0.001	0.835(0.720 ~ 0.967)	0.016	< 0.001
	Model 2	0.960(0.937 ~ 0.983)	0.001	Ref.	0.883(0.831 ~ 0.937)	< 0.001	0.850(0.732 ~ 0.987)	0.033	< 0.001
	Model 3	0.976(0.953 ~ 1.001)	0.060	Ref.	0.911(0.857 ~ 0.969)	0.003	0.902(0.775 ~ 1.050)	0.184	0.005

Model 1, No covariates were adjusted; Model 2 included additional adjustments for age, sex and race. Model 3 further adjusted for educational score and TDI based on Model 2 adjustments

Abbreviations: CI Confidence Intervals, CVH Cardiovascular Health, HR Hazard ratios, IBD Inflammatory Bowel Disease, LC9 Life’s Critical 9, SD Standard Deviation, TDI Townsend Deprivation Index; UC, Ulcerative Colitis;

Bold P-values indicate statistical significance (P < 0.05)

For CD, the continuous LC9 score showed a significant inverse association in Models 1 and 2 (Model 1: HR 0.912, 95% CI 0.847–0.981, P = 0.014; Model 2: HR 0.904, 95% CI 0.838–0.976, P = 0.010), but this association was attenuated and no longer statistically significant in the fully adjusted Model 3 (HR 0.928, 95% CI 0.858–1.003, P = 0.059). The categorical analyses did not reveal significant associations between CVH groups and CD risk, and trend tests were not significant across all models (P for trend > 0.05).

For overall IBD, the continuous LC9 score was significantly associated with reduced IBD risk in Models 1 and 2 (Model 1: HR 0.956, 95% CI 0.934–0.979, P < 0.001; Model 2: HR 0.960, 95% CI 0.937–0.983, P = 0.001), while the association in Model 3 was attenuated and borderline non-significant (HR 0.976, 95% CI 0.953–1.001, P = 0.060). When analyzed categorically, compared to the low CVH group, the moderate CVH group showed significant risk reductions across all models (Model 1: HR 0.874, 95% CI 0.824–0.928; Model 2: HR 0.883, 95% CI 0.831–0.937; Model 3: HR 0.911, 95% CI 0.857–0.969). The high CVH group showed significant risk reductions in Models 1 and 2 (Model 1: HR 0.835, 95% CI 0.720–0.967, P = 0.016; Model 2: HR 0.850, 95% CI 0.732–0.987, P = 0.033), but this was attenuated in Model 3 (HR 0.902, 95% CI 0.775–1.050, P = 0.184). Trend tests remained significant across all models (P for trend < 0.01).

Nonlinearity analysis using RCS

An RCS model was constructed in this study to further assess the dose-response relationships between LC9 scores and the incidence of UC, CD, and IBD. Following established methodological recommendations for large sample sizes [14], four knots were employed to balance model flexibility with the risk of overfitting. Knots were placed at the 5th, 35th, 65th, and 95th percentiles of the LC9 score distribution, which ensures adequate observations between knots for stable estimation while allowing sufficient flexibility to capture potential non-linear patterns. The 5th percentile served as the reference value, and the multivariable model adjustments were consistent with those in Model 3. As shown in Fig. 2A, there was a significant overall association between LC9 and UC risk (P-overall < 0.001), with a significant nonlinear relationship (P-non-linear = 0.001): HRs were higher at lower LC9 levels, decreased as LC9 increased, reached the lowest point/closest to the reference level at approximately 5.5, and subsequently plateaued or showed mild elevation. For CD, the overall association between LC9 and disease risk remained significant (P-overall = 0.002), but the test for nonlinearity was not significant (P-non-linear = 0.133), suggesting that this association approximates a linear (or decline-then-plateau) protective trend (Fig. 2B). For overall IBD, as illustrated in Fig. 2C, both the overall and nonlinear associations between LC9 and risk were significant (P-overall < 0.001, P-non-linear < 0.001).

Fig. 2.

Restricted Cubic Spline Curves Illustrating Non-linear Relationships between LC9 score and A), UC; B), CD; C), IBD. Note: Graphs show HRs for end according to LC9 score adjusted for age, sex race, educational score and TDI. Data were fitted by a cox regression model, and the model was conducted with 4 knots at the 5th, 35th, 65th, 95th percentiles of fatty acids(reference is the 5th percentile). Solid lines indicate HRs, and shadow shape indicate 95% CIs. Abbreviations: CD, Crohn's disease; IBD, Inflammatory Bowel Disease; LC9, Life's Critical 9; TDI, Townsend Deprivation Index; UC, Ulcerative Colitis; HR, hazard ratio; CI, confidence interval

Mediation analysis

In this study, a mediation analysis model was constructed to further examine whether triglycerides (TG), neutrophils, albumin, and CRP mediate the relationship between LC9 and UC. In this mediation model, LC9 was considered the independent variable, UC the dependent variable, and triglycerides, neutrophils, albumin, and CRP were treated as mediating variables. As illustrated in Fig. 3A, triglyceride levels demonstrated a significant mediating effect on the association between LC9 and UC, accounting for 10.78% of the total correlation, with an IE of 1.010 (95% CI: 1.008 to 1.012). Figure 3B shows that neutrophil levels also had a significant mediating effect on the association between LC9 and UC, explaining 26.74% of the total correlation, with an IE of 1.022 (95% CI: 1.019 to 1.025). Similarly, as indicated in Fig. 3C, serum albumin levels exhibited a significant mediating effect in the association between LC9 and UC, accounting for 4.77% of the total correlation, with an IE of 1.005 (95% CI: 1.003 to 1.006). Figure 3D illustrates that serum CRP levels significantly mediated the relationship between LC9 and UC, explaining 22.95% of the total correlation, with an IE of 1.019 (95% CI: 1.007 to 1.031). Further specific details can be found in Fig. 3.

Fig. 3.

Path diagram of the mediation analysis models. Note: A), triglycerides; B), neutrophils; C), albumin; D), CRP partially mediates the relationship between LC9 Score and UC. Abbreviations: CRP, C-Reactive Protein; LC9, Life’s Critical 9; UC, Ulcerative Colitis

Subgroup analysis and interaction

Subgroup analyses were conducted based on age, sex, race, educational attainment, smoking status, drinking status, and BMI. As shown in Fig. 4A, LC9 was negatively associated with the risk of developing UC across most subgroups. This negative association was more pronounced among participants aged ≥ 60 years, females, individuals of European descent, those with lower educational attainment, current drinkers, and those with a BMI between 18.5 and 25. Figure 4B demonstrates a significant interaction between LC9 and both smoking and drinking status. As indicated in Fig. 4C, LC9 was negatively associated with the risk of developing IBD in certain subgroups. This negative association was more evident among participants aged < 60 years, females, and individuals of European descent, with significant interactions observed between LC9 and factors such as sex, race, smoking status, and drinking status.

Fig. 4.

Subgroup analysis of the association of the LC9 Score and IBD. Note: Each stratifcation was adjusted for age, sex, race, educational score, smoking status, drinking status and BMI. A), UC; B), CD; C), CD, Crohn’s disease; IBD. Abbreviations: BMI, Body Mass Index; IBD, Inflammatory Bowel Disease; LC9, Life’s Critical 9; UC, Ulcerative Colitis

Sensitivity analysis of psychological component

As shown in Table 3, sensitivity analysis using the mental-psychological subscale of LC9 in place of the overall LC9 score revealed a consistent protective association with overall IBD risk in Cox regression models, with HRs of 0.991 (95% CI approximately 0.990–0.992) across Model 1–Model 3, all reaching statistical significance (P < 0.001). For UC, this subscale similarly demonstrated a significant protective effect, with consistent HRs of 0.995 (95% CI 0.993–0.997, P < 0.001) across all three models. A consistent protective association was also observed in CD (Model 1–Model 3: HR approximately 0.996, 95% CI 0.994–0.998, P < 0.001), indicating good robustness of the primary analysis findings.

Table 3.

Sensitivity analysis of psychological component score

Outcomes	Model	HR 95%CI	P-value
UC	Model 1	0.995(0.993 ~ 0.997)	< 0.001
UC	Model 2	0.995(0.993 ~ 0.997)	< 0.001
UC	Model 3	0.995(0.993 ~ 0.997)	< 0.001
CD	Model 1	0.996(0.995 ~ 0.998)	< 0.001
CD	Model 2	0.996(0.994 ~ 0.997)	< 0.001
CD	Model 3	0.996(0.995 ~ 0.997)	< 0.001
IBD	Model 1	0.991(0.990 ~ 0.991)	< 0.001
IBD	Model 2	0.991(0.990 ~ 0.992)	< 0.001
IBD	Model 3	0.991(0.991 ~ 0.992)	< 0.001

Model 1, No covariates were adjusted; Model 2 included additional adjustments for age, sex and race. Model 3 further adjusted for educational score and TDI based on Model 2 adjustments

Abbreviations: CI Confidence Intervals, HR Hazard ratios, IBD Inflammatory Bowel Disease, TDI Townsend Deprivation Index UC Ulcerative Colitis

Bold P-values indicate statistical significance (P < 0.05)

Discussion

The concept of LC9 is based on the well-established association between mental health factors (such as depression, anxiety, chronic and traumatic stress, and social integration) and CVD risk, underscoring the necessity of incorporating mental health assessments into overall health evaluations [7]. To the best of our knowledge, this is the first study to evaluate the impact of LC9 on the risk of IBD. In this large prospective cohort study, we found a significant inverse relationship between higher LC9 scores and the risk of UC, suggesting that maintaining good cardiovascular health may play a protective role in reducing the risk of developing UC. This finding was consistent across multiple models that adjusted for various confounding factors, including demographic variables, socioeconomic status, and baseline health behaviors.

A notable finding of this study is the differential association between LC9 and IBD subtypes. While LC9 demonstrated a significant protective effect against UC (HR = 0.909, 95% CI: 0.863–0.958 in Model 3), the association with CD did not reach statistical significance in the fully adjusted model (HR = 0.928, 95% CI: 0.858–1.003, P = 0.059). This divergence is consistent with the understanding that UC and CD, although both classified as IBD, have distinct pathophysiological characteristics that may respond differently to lifestyle modifications.

Our findings extend prior work by Yang et al., who reported significant protective associations between LE8 and reduced risk of IBD, CD, and UC in the same UK Biobank cohort [15]. In our study using LC9—which incorporates psychological health in addition to LE8 components—the association with CD was attenuated and no longer statistically significant (HR = 0.928, 95% CI: 0.858–1.003, P = 0.059), whereas the protective association with UC remained robust (HR = 0.909, 95% CI: 0.863–0.958, P < 0.001).

This apparent attenuation warrants careful interpretation. The incorporation of psychological health into LC9 redistributes the relative weight of each component, potentially diluting individual LE8 contributions. However, our sensitivity analysis demonstrated that the psychological component independently showed significant protective associations with all IBD outcomes, including CD (HR = 0.996, 95% CI: 0.995–0.997, P < 0.001), suggesting that psychological health is independently associated with IBD risk rather than attenuating the overall association. The discrepancy may reflect complex interactions among LC9 components—for instance, the paradoxical effects of smoking on UC (protective) versus CD (harmful) [16, 17] may partially offset benefits of other components in the composite score.

Several biological mechanisms may explain this differential effect. UC inflammation is confined to the colonic mucosa in a continuous pattern, whereas CD presents with transmural, skip lesions throughout the gastrointestinal tract [18]. This localized nature of UC suggests greater susceptibility to factors directly influencing the intestinal microenvironment, including diet, gut microbiota composition, and mucosal barrier integrity [19, 20]. Components of LC9, such as healthy dietary patterns, regular physical activity, and optimal metabolic parameters, can directly modulate these local factors. In contrast, the multifocal and transmural characteristics of CD inflammation may require more complex systemic interventions, potentially attenuating the observable effects of cardiovascular health optimization [21, 22].

Furthermore, smoking—a key component of LC9—has paradoxically different effects on UC and CD. While smoking is a well-established risk factor for CD, it appears to have a protective effect against UC [23, 24]. This phenomenon may partially explain the differential associations observed in our study, as the inclusion of tobacco exposure in LC9 may contribute to the stronger inverse association observed with UC compared to CD.

The RCS analysis further elucidated these relationships. For UC, we observed a significant nonlinear association (P-nonlinear = 0.001), with the protective effect plateauing at higher LC9 scores (approximately 55 points). In contrast, the LC9-CD relationship approximated a linear protective trend without significant nonlinearity (P-nonlinear = 0.133). These patterns suggest that while moderate improvements in cardiovascular health may substantially reduce UC risk, additional benefits diminish at higher LC9 levels, whereas the relationship with CD follows a more gradual, linear pattern.

Our sensitivity analysis examining the psychological health component independently revealed significant protective associations with all IBD outcomes, including CD (HR = 0.996, 95% CI: 0.995–0.997, P < 0.001). This finding is particularly noteworthy because while the overall LC9 score did not show a significant association with CD in the fully adjusted model, the psychological component alone demonstrated a consistent protective effect. This suggests that mental health may represent an independent risk factor for IBD that warrants specific attention.

The mechanisms linking psychological health to IBD risk likely involve the gut-brain axis. Studies have shown that anxiety and depression are associated with increased IBD disease activity, reduced quality of life, and higher rates of hospitalization [25, 26]. Psychological stress may directly affect neuroendocrine regulation of the gut and indirectly influence IBD progression by altering behaviors such as dietary habits and sleep quality [27]. Furthermore, the bidirectional relationship between psychological distress and intestinal inflammation suggests that effective management of mental health may both reduce IBD risk and improve disease outcomes [28–30].

Our mediation analysis identified four biomarkers that significantly mediated the LC9-UC relationship: neutrophils (26.74%), CRP (22.95%), triglycerides (10.78%), and albumin (4.77%). The substantial mediation by inflammatory markers (neutrophils and CRP) supports the hypothesis that LC9 influences UC risk primarily through modulation of systemic inflammation.

Neutrophils accumulate at sites of intestinal inflammation and play a key role in mediating inflammatory responses [31]. A healthy cardiovascular state, as reflected by higher LC9 scores, may help maintain appropriate neutrophil responses, thereby preventing excessive inflammatory reactions that contribute to UC pathogenesis [32]. Similarly, CRP elevation reflects chronic inflammatory states commonly observed in IBD patients [33], and the significant mediation effect suggests that cardiovascular health optimization may reduce UC risk partly through attenuation of systemic inflammation.

The mediation by triglycerides (10.78%) highlights the role of metabolic pathways in UC pathogenesis. Triglycerides are essential for maintaining intestinal epithelial integrity and barrier function [34, 35]. A healthy lifestyle encompassing dietary quality, physical activity, and sleep health is strongly associated with optimal TG levels [36–38], which may in turn protect against the intestinal barrier dysfunction implicated in UC development.

Albumin mediation (4.77%) underscores the interconnection between nutritional status, inflammation, and UC risk. Hypoalbuminemia in IBD patients is often associated with both malnutrition and chronic inflammation [39, 40]. These findings collectively suggest that the protective effect of LC9 on UC operates through integrated metabolic, nutritional, and inflammatory pathways, with systemic inflammation representing the dominant mechanism.

Our subgroup analyses revealed significant interactions between LC9 and several demographic and behavioral factors. For UC, significant interactions were observed with smoking status and drinking status (Fig. 4B). The interaction with smoking is particularly relevant given the complex relationship between smoking and IBD subtypes [41]. Our findings suggest that the benefits of high LC9 scores may be modified by smoking behavior, potentially through differential effects on immune function and inflammatory pathways.

The interaction with alcohol consumption underscores the need to consider drinking behaviors when assessing cardiovascular health effects on IBD risk. Alcohol has complex effects on the gut microbiota, mucosal immunity, and systemic inflammation [42], all of which may modulate the LC9-IBD relationship. Similarly, for overall IBD, significant interactions were observed with sex, race, smoking status, and drinking status (Fig. 4C), suggesting that the benefits of cardiovascular health optimization may vary across demographic subgroups.

The stronger protective association observed in older adults (≥ 60 years) for UC aligns with the hypothesis that cumulative lifetime exposure to cardiovascular health factors influences disease risk [43]. Women also showed more pronounced protective associations, which may reflect sex-specific differences in immune responses, gut microbiota composition, or hormonal factors that modulate susceptibility to lifestyle modifications [44].

The mechanism linking LC9 with IBD likely involves the integration of metabolic, inflammatory, and immunological pathways. One potential mechanism is through the regulation of gut microbiota, which plays a crucial role in maintaining intestinal homeostasis [45, 46]. Components of LC9, including healthy diet, regular physical activity, and smoking avoidance, have been shown to promote a diverse and balanced gut microbiota protective against IBD development [23, 24, 47]. Additionally, oxidative stress and endothelial dysfunction, hallmarks of both cardiovascular diseases and IBD, may represent shared pathophysiological pathways linking cardiovascular health to intestinal inflammation [48, 49].

This study has several strengths, including the novel application of the LC9 score in the context of IBD, utilization of data from over 370,000 UK Biobank participants with a median follow-up of 13.01 years, use of validated instruments (PHQ-4, PHQ-9, and GAD-7) for psychological health assessment, and comprehensive mediation and subgroup analyses that provide mechanistic insights.

However, several limitations warrant consideration. First, despite comprehensive covariate adjustment, residual confounding from unmeasured factors cannot be excluded, and the observational design limits causal inference. Second, LC9 components were primarily assessed at baseline, and temporal changes in cardiovascular health status were not captured. Third, while we analyzed the composite LC9 score, individual components may differentially influence IBD pathogenesis; future research should examine component-specific effects. Fourth, the predominantly European ancestry of UK Biobank participants may limit generalizability to other populations. Finally, although we conducted sensitivity analysis for the psychological component, separate analyses for all nine components were not performed due to concerns about standardization and potential complex interactions among components.

Conclusion

In conclusion, this study provides robust evidence of a significant inverse association between LC9 and the risk of UC, with systemic inflammation and immune dysregulation potentially acting as mediators in this relationship. Innovatively, this research quantifies the mental health status scores within the components of LC9, including symptoms of depression and anxiety. The findings suggest that maintaining optimal cardiovascular health, as reflected by a high LC9 score, may confer a protective effect against the development of UC. However, the relationship between LC9 and CD appears to be more complex, warranting further investigation. This study also underscores the importance of considering demographic and lifestyle factors, such as age, sex, smoking, and alcohol consumption status, when evaluating the impact of cardiovascular health on the risk of IBD. Future research should aim to elucidate the underlying biological mechanisms and explore the potential of targeted prevention strategies based on the optimization of cardiovascular health.

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Yu Chang. The first draft of the manuscript was written by Yu Chang, Fan Li, Chanjiao Yu, Zhihao Wang, Yuqin Li and Tongyu Tang. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The present study was supported by The Science and Technology Agency Jilin Province (grant nos. 20210402013GH and 20200201343JC).

Data availability

The data can be downloaded here: https://www.ukbiobank.ac.uk/.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki. The UK Biobank received ethical approval from the North West Multi-Centre Research Ethics Committee, Manchester, U.K. (REC reference 11/NW/0382), and all participants provided written informed consent.The analyses in this study were conducted under the UK Biobank application ID 84347.

Consent for publication

All participants provided written informed consent before participation.

Competing interests

The authors declare no competing interests.