Impact of Maternal Smoking, Offspring Smoking, and Genetic Susceptibility on Crohn’s Disease and Ulcerative Colitis

Abstract

Background and Aims

The long-term impact of maternal smoking during pregnancy [MSDP] on the risk of Crohn’s disease [CD] and ulcerative colitis [UC] in adult offspring remains uncertain. The present study aimed to investigate the individual and combined effects of early life exposure [MSDP], offspring personal behaviour [smoking], and genetic risk on the development of CD and UC in adult offspring.

Methods

We conducted a prospective cohort study using UK Biobank data, including 334 083 participants recruited between 2006 and 2010, with follow-up until December 31, 2021. Multivariable Cox regression models were used to evaluate the associations of genetic factors, maternal and personal smoking, and their combination with CD and UC.

Results

Participants exposed to MSDP had an 18% increased risk of CD compared to those without MSDP (hazard ratio [HR] = 1.18, 95% confidence interval [CI] = 1.01–1.39). However, no significant association was found between MSDP and UC risk [HR = 1.03, 95% CI = 0.92–1.16]. Personal smoking increased the risk of CD and UC, and had a numerically amplified effect with MSDP. Participants with high genetic risk and MSDP had a 2.01-fold [95% CI = 1.53–2.65] and a 2.45-fold [95% CI = 2.00–2.99] increased risk of CD and UC, respectively, compared to participants without MSDP and with low genetic risk.

Conclusions

Our prospective cohort study provides evidence that MSDP increases the risk of CD in adult offspring, whereas no evidence supports their causal association. Additionally, smoking and genetic susceptibility had a numerically amplified effect with MSDP on CD and UC, but the interaction lacked statistical significance.

Maternal smoking, Crohn’s disease, ulcerative colitis

1. Introduction

Inflammatory bowel disease [IBD] constitutes a range of chronic, idiopathic inflammatory conditions that affect the gastrointestinal tract, primarily manifesting as Crohn’s disease [CD] and ulcerative colitis [UC].¹ Recent epidemiological trends indicate a significant surge in the global incidence of IBD, particularly in the USA and UK.^2,3 According to data from the 2019 Global Study of Diseases, Injuries, and Risk Factors, IBD has a profound impact on patients’ quality of life, requiring long-term care and costly medical management and resulting in 1.62 million disability-adjusted life years in 2019.^4,5

Maternal and personal smoking are prevalent and controllable risk factors for adverse health outcomes.⁶ Notably, ~50% of smokers in Europe and the USA continue to smoke during pregnancy.⁷ Maternal smoking during pregnancy [MSDP] has been implicated in having adverse consequences such as sudden unexpected infant death,⁸ lung injury,⁹ and childhood hypertension,¹⁰ as well as potential long-term effects on offspring health, including gastrointestinal disease.¹¹ Existing studies have also reported that MSDP may increase the risk of IBD. Two case-control studies have demonstrated a dose–response relationship between maternal active and passive smoking during pregnancy and an increased risk of developing IBD, especially CD, in children.^12,13 Moreover, MSDP has also been demonstrated to influence the phenotype and disease course of paediatric CD.¹⁴ However, these studies were limited by their retrospective design, small sample sizes, and lack of causal confirmation. Currently, few studies have systematically evaluated the individual and combined effects of maternal and personal smoking on the risk of developing adult-onset IBD.

In addition to environmental and behavioural factors, genetic susceptibility may play a crucial role in the development of IBD. Epidemiological evidence indicates a genetic predisposition in the development of CD, with 15% of patients having a family history with IBD.¹⁵ To further investigate this, polygenic risk score [PRS] analysis has been used to identify individuals with a higher genetic risk for IBD, potentially serving as a screening tool.¹⁶ However, the interaction between genetic susceptibility to IBD and MSDP remains insufficiently explored. Furthermore, it is unclear whether the effects of MSDP have a combined effect with the presence of inherent risk factors.

To address these research gaps, we conducted a prospective cohort study utilizing the UK Biobank database, with the objectives of [1] examining the association between maternal and personal smoking and the incidence of new-onset CD and UC in the general population, and [2] assessing the individual and combined effects of genetic risk as defined by the PRS and maternal smoking on the incidence of CD and UC.

2. Methods

2.1. Study population

The UK Biobank is a large prospective population-based cohort study that recruited over half a million participants aged 37–73 years from 22 assessment centres across the UK in 2006–2010. The study amassed comprehensive data on environmental factors, medical history, genetics, phenotypes, and biological samples, all accessible through application to the UK Biobank [www.ukbiobank.ac.uk/].

Our study utilized a sample of 334 083 participants for analyses, which comprised individuals with available genetic data and excluded those with self-reported or diagnosed IBD or cancer at baseline [Supplementary Figure S1]. We also excluded participants with missing genetic data, or maternal or personal smoking data [see Supplementary Methods for more details]. The study was approved by the National Research Ethics Committee [REC ID: 16/NW/0274], and electronic informed consent was obtained from all participants.

2.2. Exposure assessment

In this study, information on maternal and personal smoking was collected through a self-administered questionnaire. Participants were asked, ‘Did your mother smoke regularly around the time when you were born?’ The answers were likely to be ‘No’, ‘Yes’, ‘Don’t know’ and ‘Prefer not to answer’. Those who did not answer or answered ‘Don’t know do not want to answer’ or ‘Prefer not to answer’ whether their mother smoked during pregnancy were excluded. Furthermore, pack-years of personal smoking were calculated from self-reported data on smoking status, age of smoking initiation, age of cessation, age at recruitment, and average daily cigarette consumption. Personal smoking status was categorized based on pack-years into three groups: never-smoking, 0–30 pack-years, and >30 pack-years.

2.3. Polygenic risk scores

In our study, PRS values for CD and UC were constructed using 217 single-nucleotide polymorphisms [SNPs] identified from genome-wide association studies [GWAS] for individuals of European ancestry.¹⁷ The selection of SNPs was based on ‘clumped’ results, with the most significant variant per linkage disequilibrium block meeting the criteria of r² > 0.8 and p < 5 × 10⁻⁸. The PRS was calculated as the sum of an individual’s risk alleles, weighted by their effect sizes derived from GWAS data, according to equation [1]:¹⁸

$${\displaystyle \begin{array}{l}{\displaystyle PRS={\displaystyle \sum _{i=1}^n}{\beta }_i\ast SN{P}_i}\end{array}}$$

[1]

where SNP_i is the risk allele number [0, 1, or 2] of each SNP. The PRS was categorized into three genetic risks, defined as ‘low risk’ [the lowest tertile], ‘medium risk’ [second tertile], and ‘high risk’ [the highest tertile], with a higher PRS indicating a greater genetic predisposition to IBD. Additional details on the genotyping and imputation procedures of the SNPs in the UK Biobank are provided in the Supplementary Methods.

2.4. Outcome assessment

Outcome indicators were defined as CD [ICD-10 code K50] and UC [ICD-10 code K51]. Participants were identified using hospital inpatient records from the Hospital Episode Statistics for England and Wales, and Scottish Morbidity Record data for Scotland. Person-years of follow-up were calculated from the recruitment date to the earliest occurrence of an IBD diagnosis, death, or the end of follow-up [December 31, 2021].

2.5. Assessment of covariates

Sociodemographic characteristics (sex, age, assessment centre, ethnicity, Townsend deprivation index, and body mass index [BMI]), medication [vitamin, mineral, aspirin, and statin use], lifestyle factors [physical activity, drinking status, smoking status, and dietary patterns], and baseline disease [high blood pressure, high cholesterol, diabetes, asthma, rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis, and irritable bowel syndrome] were obtained from standardized questionnaires, physical measurements, and biological sample collection. BMI was calculated by measuring height in metres and weight in kilograms. Serum concentrations of cholesterol and glucose were quantified using a Beckman Coulter AU5800 analyser.

2.6. Statistical analysis

Multivariable Cox regression models were used to assess the association of genetic factors, maternal and personal smoking, and their combination on the incidence of CD and UC, with enrolment time as the time scale. The Cox proportional hazards assumption was tested using the UK Biobank’s Schoenfeld test. We excluded participants who had IBD during the first 2 years of follow-up to reduce the possibility of reverse causality. In Model 1, we adjusted for multiple variables including Townsend deprivation index, BMI, medication [vitamin, mineral, aspirin, and statin use], and baseline disease [high blood pressure, high cholesterol, and diabetes], and stratified the analyses jointly by age, sex, and assessment centres [22 UK Biobank centres]. To further reduce potential confounding effects by other lifestyle and genetic factors, we additionally adjusted lifestyle factors [physical activity, drinking status, smoking status, and dietary patterns] and genetic factors [first ten principal components of ancestry] in Model 2. To investigate the potential modifying effect of personal smoking and genetic susceptibility on the association of maternal smoking with CD and UC, we conducted the analysis stratified by personal smoking status [never-smoking, 0–30 pack-years, and >30 pack-years] and PRS [low risk, medium risk, and high risk]. In addition, a logistic analysis was performed to investigate the effect of MSDP on personal smoking in the offspring. We also calculated p values for trend and P for interaction.

To assess the robustness and reliability of the results, we performed sensitivity analyses as follows: [1] we stratified the analyses by age [<60, ≥60 years], sex, and BMI [<25, ≥25 kg/m²], drinking status [no, yes], pack years of personal smoking, and personal smoking status [no, yes]; [2] we controlled for immune-mediated inflammatory diseases including asthma, rheumatoid arthritis, systemic lupus erythematosus, and multiple sclerosis at baseline; [3] to mitigate potential bias arising from shared gastrointestinal symptoms, we accounted for the presence of irritable bowel syndrome at baseline; and [4] we constructed a competing risk model to control for the competitive risk bias caused by death.

We carried out all analyses using R software, version 4.2.0.

3. Results

3.1. Population baseline characteristics

After excluding individuals with IBD and cancer at baseline, 334 083 participants from the UK Biobank were ultimately included. During a median follow-up period of 12.60 years, 230 921 participants reported exposure to MSDP. Table 1 presents the baseline characteristics of participants stratified by MSDP. Participants exposed to MSDP were more likely to be male, have higher BMI and Townsend deprivation index, and have baseline conditions such as hypertension, high cholesterol, and diabetes mellitus. Furthermore, participants with MSDP demonstrated a propensity for unhealthy lifestyles, characterized by higher rates of smoking and lower consumption of fruits and vegetables. They were also more likely to use medications such as aspirin and statins. The characteristic distribution of participants is shown in Supplementary Table S1.

Table 1.

Baseline characteristics of participants by maternal smoking

Baseline characteristics	Overall [N = 334 083]	Maternal smoking [N = 230 921]	Non-maternal smoking [N = 103 162]
Age, mean [SD], years	56.46 [8.01]	55.90 [7.68]	56.70 [8.14]
Male, N [%]	154 330 [46.2]	49 285 [47.8]	105 045 [45.5]
White, N [%]	334 083 [100.0]	103 162 [100.0]	230 921 [100.0]
BMI, mean [SD], kg/m2	27.35 [4.76]	27.93 [4.97]	27.10 [4.64]
Townsend deprivation index, mean [SD]	16.48 [13.52]	18.20 [14.63]	15.72 [12.92]
Lifestyle factor
Physical activity, mean [SD], MET h/week	44.64 [45.31]	45.88 [47.44]	44.09 [44.31]
Fruit and vegetable intake, mean [SD], single/day	4.57 [2.96]	4.41 [2.89]	4.64 [2.99]
Personal smoking, N [%]
Never-smoking	192 344 [57.6]	57 713 [55.9]	134 631 [58.3]
0–30 packs/year	112 758 [33.8]	33 684 [32.7]	79 074 [34.2]
30+ packs/year	28 981 [8.7]	11 765 [11.4]	17 216 [7.5]
Drinking, N [%]
Never drinking	56 525 [16.9]	17 380 [16.9]	39 145 [17.0]
Occasional drinking	36 846 [11.0]	11 517 [11.2]	25 329 [11.0]
Regular drinking	240 522 [72.0]	74 189 [72.0]	166 333 [72.1]
Comorbidity, N [%]
Hypertension	194 540 [58.2]	60 930 [59.1]	133 610 [57.9]
High cholesterol	39 910 [11.9]	12 815 [12.4]	27 095 [11.7]
Diabetes	26 762 [8.0]	9494 [9.2]	17 268 [7.5]
Medications, N [%]
Vitamin	49 093 [14.7]	14 773 [14.3]	34 320 [14.9]
Mineral supplement	71 084 [21.3]	21 766 [21.1]	49 318 [21.4]
Aspirin	47 067 [14.1]	15 075 [14.6]	31 992 [13.9]
Statin	52 560 [15.7]	17 050 [16.5]	35 510 [15.4]

BMI, body mass index; SD, standard deviation; MET h/week, hours of physical activity per week.

3.2. Association of maternal smoking and personal smoking with the incidence of CD and UC

Maternal smoking was associated with the incidence of IBD [Table 2]. Compared to those whose mothers did not smoke during pregnancy, participants with MSDP had a 18% increased risk of CD (hazard ratio [HR] = 1.18, 95% confidence interval [CI] = 1.01–1.39). However, we did not find any significant association between MSDP and the risk of developing UC [HR = 1.03, 95% CI = 0.92–1.16]. Sensitivity analysis adjusted for immune-mediated inflammatory diseases, irritable bowel syndrome, and competing risk of death showed that MSDP increased the risk of CD [Supplementary Table S2].

Table 2.

Associations between maternal smoking with risk of Crohn’s disease and ulcerative colitis

	CD		UC
	Non-maternal smoking	Maternal smoking	Non-maternal smoking	Maternal smoking
Cases	450	242	951	469
Person-years	2 597 018	1 158 109	2 594 397	1 157 033
Incidence ratea	17.3	20.9	36.7	40.5
Multivariate-adjusted model 1b, HR [95% CI]	1.00 [Reference]	1.19 [1.01, 1.40]*	1.00 [Reference]	1.05 [0.94, 1.18]
Multivariate-adjusted model 2c, HR [95% CI]	1.00 [Reference]	1.18 [1.01, 1.39]*	1.00 [Reference]	1.03 [0.92, 1.16]

^aPer 100 000 person-years.

^bMultivariable adjusted model 1: multivariate-adjusted hazard ratios and 95% CIs were estimated from Cox regression models, stratified by sex, age, and assessment centre, and additionally adjusted for sociodemographic characteristics [Townsend deprivation index and BMI], medication [vitamin, mineral, aspirin, and statin use], and baseline disease [high blood pressure, high cholesterol, and diabetes].

^cMultivariable adjusted model 2: fully adjusted model stratified by sex, age, and assessment centre, and additionally adjusted for sociodemographic characteristics [Townsend deprivation index and BMI], medication [vitamin, mineral, aspirin, and statin use], lifestyle factor [physical activity, drinking status, smoking status, and dietary patterns], baseline disease [high blood pressure, high cholesterol, and diabetes], and gene [first ten principal components of ancestry].

Asterisks indicate: *p < 0.05.

When we stratified participants by offspring smoking status [non-smoking and smoking] in subgroup analyses [Supplementary Table S3], we found that MSDP had a numerically amplified effect with personal smoking on CD, but there was no evidence of an interaction between the two factors [P for interaction >0.01].

Supplementary Table S4 shows the correlation between personal smoking status and the risk of CD and UC. Personal smoking was significantly associated with an increased risk of CD in all individual smoking profiles. Participants with more than 30 pack-years of personal smoking had a significantly higher relative incidence of CD [adjusted HR = 2.12, 95% CI = 1.68–2.67] and UC [adjusted HR = 2.25, 95% CI = 1.92–2.64] compared to never smokers. There was a tendency to increase the risk of CD and UC as personal smoking levels increased [p for trend <0.001]. The effects of quitting time on CD and UC are shown in Supplementary Table S5.

3.3. Effect of maternal and personal smoking on subsequent risk of CD and UC

We further evaluated the effect of maternal and personal smoking on the incidence of UC and CD [Figure 1]. Compared to participants without maternal and personal smoking, the relative risk of developing CD and UC increased 1.60-fold [95% CI = 1.24–2.08] and 1.46-fold [95% CI = 1.21–1.75] in participants exposed to MSDP and 0–30 pack-years of personal smoking, respectively. Indeed, the estimate increased 2.66-fold [95% CI = 1.95–3.63] and 2.46-fold [95% CI = 1.97–3.07] in those exposed to MSDP and more than 30 pack-years of personal smoking, respectively. Figure 1 also shows that the hazard ratio of MSDP on the risk of IBD was numerically higher in those offspring who were personal smokers; however, there was no statistical interaction between MSDP and personal smoking.

Figure 1.

Hazard ratios of Crohn’s disease and ulcerative colitis based on the combined classification of maternal smoking and offspring smoking dose.^a

^aMultivariate adjusted hazard ratios and 95% CIs were estimated from Cox regression models, stratified by sex, age, and assessment centre, and additionally adjusted for sociodemographic characteristics [Townsend deprivation index and BMI], medication [vitamin, mineral, aspirin, and statin use], lifestyle factor [physical activity, drinking status, and dietary patterns], baseline disease [high blood pressure, high cholesterol, and diabetes], and gene [first ten principal components of ancestry].

3.4. Effect of maternal smoking on IBD incidence in individuals predisposed to IBD genetic risk.

Supplementary Table S6 illustrates the associations between genetic susceptibility and the incidence of CD and UC. The overall HRs for medium PRS compared to low PRS for the incidence of CD and UC were 0.98 [95% CI: 0.80–1.21] and 1.67 [95% CI: 1.43–1.93], respectively. By contrast, the corresponding HRs for high PRS were 1.76 [95% CI: 1.46–2.12] and 2.26 [95% CI: 1.96–2.60], respectively, indicating a numerically greater effect. There was a trend toward an increased disease risk as individual genetic risk increased [p < 0.001 for trend].

We estimated the effect of genetic risk and MSDP on the risk of developing IBD [Figure 2]. The risk of CD and UC increased 2.01-fold [95% CI = 1.53–2.65] and 2.45-fold [95% CI = 2.00–2.99] in participants exposed to high genetic risk and MSDP, respectively, compared to participants without MSDP and those at low genetic risk. Although there seemed to be a numerically increased effect of MSDP on CD and UC in individuals with high genetic risk compared to medium or low genetic risk, Supplementary Table S7 shows that there was no significant interaction between MSDP and genetic risk [CD: P for interaction = 0.059; UC: P for interaction = 0.598]. Supplementary Table S7 highlights that MSDP elevates the risk of CD in individuals with low genetic risk. However, the impact of MSDP was not statistically significant in individuals with medium and high genetic risk. This could be attributed to the fact that the influence of MSDP tends to be less pronounced when genetic factors play a more significant role.

Figure 2.

Hazard ratios of Crohn’s disease and ulcerative colitis based on the combined classification of maternal smoking and offspring genetic risk.^a

^aMultivariate adjusted hazard ratios and 95% CIs were estimated from Cox regression models, stratified by sex, age, and assessment centre, and additionally adjusted for sociodemographic characteristics [Townsend deprivation index and BMI], medication [vitamin, mineral, aspirin, and statin use], lifestyle factor [physical activity, drinking status, and dietary patterns], baseline disease [high blood pressure, high cholesterol, and diabetes], and gene [first ten principal components of ancestry].

4. Discussion

In this prospective cohort study involving 334 083 participants, we observed that maternal smoking increased the risk of offspring developing CD. Moreover, the smoking behaviour of offspring elevated the risk of both CD and UC, and had a numerically amplified impact with MSDP on IBD. Maternal smoking-related CD and UC risk were found across individuals with varying genetic susceptibility. Specifically, participants with MSDP and a higher genetic risk had an estimated 2- to 3-fold greater risk of developing CD and UC. This study provided valuable insights into the combined effects of maternal smoking, personal smoking, and genetic susceptibility on the development of IBD and emphasized the importance of considering early life exposures, personal behaviours, and genetic factors in IBD risk assessment and prevention strategies.

MSDP had short- and long-term negative implications on the health of offspring. Robust evidence supported a significant association between MSDP and the development of IBD in children. A multicentre study conducted in Japan further confirmed the dose–response relationship, revealing a significantly elevated risk of CD in children exposed to maternal passive smoking during pregnancy.¹² In addition to affecting children’s health, MSDP may also increase long-term gastrointestinal morbidity, particularly IBD, in offspring.¹¹ Notably, current studies have indicated that MSDP mainly elevates the risk for IBD, notably CD.¹⁹ Fewer studies have identified a correlation between MSDP and UC, and one matched case-control study suggested that MSDP had a stronger impact on CD compared to UC.¹³ It was believed that DNA methylation served as the biological mechanism through which MSDP increased the risk of IBD.²⁰ Our study found that MSDP increased the risk of CD among adult offspring, but there was no statistically significant correlation with UC, which aligned with previous studies in this field.

Existing research on the effects of smoking on UC and CD presented contradictory findings. Some studies suggested that smoking increased the risk of CD but decreased the risk of UC in Western populations.²¹ One Israeli study found no association between smoking and CD, with a higher proportion of current smokers and a lower proportion of quitters in UC cases.²² A Korean national study showed that former smokers had a significantly higher risk of developing UC compared to non-smokers, which may be proportional to cumulative smoking exposure.²³ There were no uniform conclusions about the effects of smoking on IBD, and there was no convincing evidence to explain these contradictory phenomena. In contrast, our study found that smoking significantly increased the risk of UC and CD. This finding differed from another study based on UK Biobank data, probably due to differences in our definition of smoking status, inclusion/exclusion criteria, and especially the covariates controlled for.²⁴ The negative effect of smoking on CD could be mediated by non-nicotine substances. It was reasonable to assume that the specific mechanism for the negative effect could be mediated by humoral and cell immunity.²⁵ In addition, some potential influencing mechanisms have also been proposed, such as cytokine and eicosanoid levels, changes in intestinal motility, permeability and blood flow, colonic mucus, and oxygen-free radicals,²⁶ and further studies are needed to explore the mechanisms regarding the effects of smoking on offspring IBD. However, there was no significant evidence for a causal relationship between smoking with CD and UC development using Mendelian randomization [Supplementary Table S8].

In the present study, personal smoking had a numerically amplified impact with MSDP on the risk of IBD in adult offspring. Previous studies have shown that MSDP and personal smoking have a combined influence on other diseases.^27,28 In addition, MSDP serves as a risk factor for smoking initiation in adult offspring, thereby implying that MSDP can impact the health of offspring by influencing their behaviour.²⁹ Our study also observed that offspring exposed to MSDP were 10% more likely to have smoking behaviours. Even after considering all potential confounding factors, the adjusted risk of having smoking behaviours was still 5% higher among individuals exposed to MSDP [Supplementary Table S9]. In addition, we found a significant causal relationship between MSDP and increased risk of personal smoking habits in offspring through Mendelian randomization [Supplementary Table S8]. However, the results of the mediation analysis indicated that insufficient evidence was found to prove that self-smoking mediated the effect of MSDP on CD and UC [Supplementary Table S10].

IBD is a difficult-to-treat disease induced by a combination of genetic and environmental factors. Genes and genetic loci associated with IBD can influence the development of IBD by affecting intestinal homeostasis through barrier function, microbial defence, innate immune regulation, reactive oxygen species production, autophagy, adaptive immune regulation, and endoplasmic reticulum stress.³⁰ Large-scale GWAS conducted on European cohorts have identified over 200 genetic loci associated with IBD, and the PRS derived from these SNPs was used to predict genetic heterogeneity of IBD in different populations.^17,31 In addition, polygenic components play a role in exploring pathogenesis, localizing disease subtypes, and identifying potential novel variants.^32–34 In our study, we stratified participants by PRS tertile and observed a significant association between high PRS and increased risk of CD and UC, which is consistent with previous studies.³⁵ Importantly, our study revealed that there was a numerically amplified effect of MSDP and high genetic risk on CD and UC. This suggested that there might be a superimposed effect of early life exposure and genetic susceptibility on the development of IBD, which need to be explored in additional studies. Evidence indicates a potential causal relationship between DNA methylation and IBD.²⁰ Genetic variation affects DNA methylation levels through the combination of transcription factors, with an effect potentially more pronounced than MSDP.³⁶ This might explain the possible superimposed effect of MSDP and genetic factors to some extent. Therefore, we believe that attention to early life exposures, especially in populations at high genetic risk, is important for the prevention and control of IBD.

This study possesses several key strengths. First, unlike previous studies primarily conducted as case-control studies with limited sample sizes, our study employs a prospective cohort design utilizing the extensive UK Biobank dataset, offering stronger cause–effect validation. Second, it controlled for a comprehensive set of covariates, notably including genetic factors, minimizing the influence of confounding factors on the observed associations. Third, our study examined the numerically amplified effect of early-life factors, individual behaviour, and genetic factors on the development of CD and UC in adult offspring. Fourth, we used Mendelian randomization to investigate the causal associations between smoking and IBD, as well as between MSDP and personal smoking. In addition, the robustness of our findings is further enhanced by the consistent results obtained from subgroup and sensitivity analyses.

This study also has some limitations. First, there is potential for recall bias and misclassification inherent in the reliance on self-reported smoking information, particularly the recollections of MSDP by older participants. Second, although we corrected for many factors, there may be other residual confounding. Third, the study’s participants were limited to white British people, restricting the extrapolation and usefulness of the results. Fourth, in the case of MSDP, IBD risk may vary by maternal smoking cessation status. Fifth, while there appears to be an amplified effect when both exposures co-occur, their interaction was not statistically significant, potentially due to insufficient power in the dataset to detect this interaction. However, our study was unable to assess the association due to a lack of this information. Future studies that include detailed information on maternal smoking cessation status are needed to investigate its impact. More studies are needed to explore the mechanism of action.

5. Conclusion

Our large prospective cohort study provides evidence that maternal smoking increases the risk of developing CD in adult offspring. Furthermore, smoking [personal behaviour] and genetic susceptibility can amplify the impact of MSDP on the development of IBD. However, due to the observational nature of our study design, further studies are still needed to verify causality and explore mechanisms of action.

Funding

This work was supported by the National Natural Science Foundation of China [grant numbers 82003408, 82003524, and 82103913], the Startup Fund for the 100 Top Talents Program, SYSU [392012], Beijing Key Laboratory of Indoor Air Quality Evaluation and Control [No. BZ0344KF21-05], and the Fundamental Research Funds for the Central Universities, Sun Yat-sen University [23hytd005].

Conflict of Interest

The authors declare that they have no known competing financial interests or personal relationships that could appear to have influenced the work reported in this paper.

Acknowledgments

The authors would like to thank the participants from the UK Biobank studies for their contributions.

Author Contributions

Linmin Hu, Siqing Wu: conceptualization, methodology, software, data curation, formal analysis, supervision, writing—original draft, writing—review & editing. All Authors: writing—review & editing. Jinqiu Yuan, Jianbang Xiang, Bin Xia: statistical analysis, funding acquisition, conceptualization, supervision, writing—review & editing.

Data Availability

This research received access to the UK Biobank [https://www.ukbiobank.ac.uk] under application number 51671. Further information is available from the corresponding author upon request.

Author notes

Linmin Hu and Siqing Wu contributed equally and share first authorship.