|Home | About | Journals | Submit | Contact Us | Français|
Adipose tissue in mesenteric fat plays a key role in systemic and luminal inflammation. However, little is known about the role of visceral adipose tissue (VAT) and its interaction with genetic predisposition in Crohn’s disease (CD) progression.
Our study population included CD patients enrolled in Prospective Registry in Inflammatory Bowel Disease Study at Massachusetts General Hospital (PRISM). VAT volume was measured from CT scans using Aquarius 3D. We used logistic regression models to estimate the multivariable-adjusted odds ratio (MV-adjusted OR) and 95% confidence interval (CI). We tested for effect modification by genetic predisposition using the log-likelihood ratio test.
Among 482 CD patients with available data on VAT, 174 developed penetrating disease, 132 developed stricturing disease, 147 developed perianal disease, and 252 required surgery. Compared to individuals in the lowest quartile of VAT volume, the MV-adjusted OR of surgery among individuals in the highest quartile was 2.02 (95% CI, 1.09 – 3.76; Ptrend = 0.006). Similarly, the risk of penetrating disease appeared to increase with greater VAT volume (Ptrend = 0.022) but not stricturing or perianal disease (all Ptrend > 0.23). The associations between VAT volume and CD complications were not modified by genetic predisposition (all Pinteraction > 0.12).
Visceral adiposity as measured by VAT volume may be associated with a significant increase in risk of penetrating disease and surgery in CD. Our data suggest that visceral adiposity as measured by VAT may negatively impact long-term progression of CD regardless of genetic predisposition.
Crohn’s disease (CD) is a chronic inflammatory disorder of the gastrointestinal tract with a broad spectrum of clinical presentations. Nearly 50% of CD patients develop complications such as stricturing or penetrating disease, while over 40% require surgery in their lifetime.1,2 Because of this significant heterogeneity in disease course, identification of genetic and environmental factors that allow for better patient stratification may afford opportunities for personalized medicine. Although, recent studies have identified a number of novel genetic and environmental factors that are associated with risk of disease onset, little is known about the role of environmental and genetic factors in disease progression.3–6
Obesity, as defined by body mass index (BMI), has been linked to increased risk of CD, likely through its association with intestinal inflammation, permeability, and changes in gut microbiota.7–11 However, we previously found that BMI was not associated with increased risk of CD complications.12 Additionally, we did not observe an interaction between BMI and genetic predisposition and risk of CD complications. Nonetheless, as evidenced by the presence of creeping fat and mesenteric fat hypertrophy among CD patients, it is possible that visceral adipose tissue (VAT), which also contains the more biologically active adipocytes, and not total obesity, is linked to CD progression.13
VAT has been shown to play an active role in systemic inflammation through production of pro-inflammatory mediators such as Tumor Necrosis Factor – α (TNF-α), and anti-inflammatory cytokines such as Interleukin – 10 (IL-10).14,15 In addition, compared to subcutaneous fat, there is a higher expression of leptin mRNA in VAT which in turn can induce TNF-α and lead to increased intestinal inflammation.16 Compared to subcutaneous adipose tissue, VAT secretes three-fold more IL-6.17 A number of prior studies have evaluated the impact of VAT on disease incidence and progression, however these studies have been limited by their sample size and lack of detailed information on important confounders such as smoking.16,18–21 In addition, prior studies have not accounted for the potential influence of genetic variants on the association between VAT and CD progression. The importance of genetics in disease progression has been highlighted by twin studies demonstrating similar disease phenotypes across identical twins as well as a recent study demonstrating the association between genetic variants and disease phenotypes in CD.22,23 Therefore, a more comprehensive examination of the influence of environment on CD progression may be best characterized in the context of genetic predisposition. We therefore sought to examine the effect of visceral adiposity as measured by VAT volume and its interaction with genetic predisposition on risk of CD complications using a large cohort of CD patients with comprehensive information on lifestyle and genetic factors.
Starting in 2004, adult patients, age ≥ 18 years, with a diagnosis of CD, ulcerative colitis, or indeterminate colitis were recruited in the Prospective Registry in IBD Study at Massachusetts General Hospital (PRISM). At the time of recruitment, patients were interviewed by a study coordinator to collect detailed information on their disease characteristics according to the Montreal classification, lifestyle factors including smoking, BMI, and other comorbid conditions. In addition, approximately 10 cc of blood was drawn from each participant at the time of enrollment for extraction of the buffy coat. For this study, among 787 patients with CD with available genetic data in our cohort, 482 patients had at least one available abdominal CT scan and therefore were eligible for inclusion. There were no significant differences in age at diagnosis, sex, race, BMI, history of smoking and disease duration comparing individuals with at least one CT scan to those without (Supplementary Table 1). The Institutional Review Board at the Massachusetts General Hospital (MGH) approved this study.
Although both MRI and CT scans have been used for VAT measurements, due to wide use of CT scans at our institution during the study period, VAT measurements were only done using CT scans.24 Suitable scans were assessed with Aquarius 3D iNtuition viewer (TeraRecon Inc., version 4.4.12, San Mateo, CA). This software traces the abdominal muscular layer and separates subcutaneous from visceral adipose tissue (including mesenteric adipose tissue), after which their volumes are measured within predetermined reference points, applied by the researcher, by using the Fat Analysis 3D option. Using previously reported methods, adipose tissue volumes were measured from the 11th thoracic vertebra (T11) to the 5th sacral vertebra (S5).18,25 As some individuals had a CT scan after the date of diagnosis of CD complications, we compared VAT volumes before and after CD complications among individuals with multiple CT scans before and after the date of disease complications, and observed no significant differences (Supplementary Table 2).
We defined CD complication as the presence of penetrating disease, stricturing disease, perianal disease, or the need for having a bowel resection. At the time of enrollment, information on disease behavior (penetrating, stricturing, and perianal disease) and previous surgeries was collected and confirmed by review of medical records and further verified by primary gastroenterologists.
Participants reported their height at the time of enrollment while information on weight was obtained from the review of medical records at the time of the CT scan. BMI was calculated from height and weight. We also collected information on sex, race, smoking status at the time of enrollment (never, past, or current), age at diagnosis, disease location (ileal, colonic, and ileocolonic), and ever use of CD-related medications (mesalamines, thiopuines, methotrexate, and anti-TNF). Data on date of diagnosis, age of diagnosis, disease location, and medications were confirmed by review of medical records and verified by primary gastroenterologists.
We have previously reported in detail information on genotyping and computation of genetic risk score.12 Briefly, genotyping was performed on the Illumina Immunochip at the Broad Institute (Cambridge, MA). We used the most recent meta-analysis of genome-wide association studies to identify 140 single nucleotide polymorphisms (SNPs) associated with risk of CD.6 Genetic risk score (GRS) was calculated on the basis of these SNPs by using a previously reported weighted method.26,27 Each SNP was recoded as 0, 1, or 2 according to the number of CD increasing risk alleles, and each SNP was weighted by its relative effect size (regression β coefficient) derived from the previously reported meta-analysis data. We created the GRS for CD using the equation of GRSCD= Σ(βi×SNPi) / Σ (βi), where βi is the β coefficient of SNPi on susceptibility to CD. The GRSCD was previously shown to be associated with higher risk of CD complications in PRISM.28 We categorized individuals according to their genetic risk score into high and low risk groups based on the median genetic risk score value of the entire cohort.
We compared the baseline characteristics of study participants across quartiles of VAT residuals using Chi-square tests for categorical variables and Kruskal-Wallis H test for continuous variables. Because VAT and BMI were highly correlated (Spearman correlation, r = 0.69, P < 0.001), we obtained the component of VAT volume that is not explained by BMI, i.e. VAT-residuals, by regressing VAT volume on BMI. These VAT-residuals were used as the exposure variables in our study in order to (a) examine the association between VAT and CD complications independent of BMI, and (b) to avoid collinearity in the multivariable association models. VAT volumes were entered into the models as quartiles where the cutoff for each quartile was determined from the distribution of VAT in patients without complications. We estimated the risk of CD complications by calculating the odds ratio (OR) and 95% confidence interval (CI) using logistic regression modeling adjusting for potential confounders. P-trends were calculated by using the median value of each quartile. All models were adjusted for sex, history of smoking, time between disease diagnosis and CT scan, and age at diagnosis. We evaluated effect modification by genetics using cross-classified categories of genetic risk score and VAT volumes. We tested the significance of this interaction using the log likelihood ratio test to compare the fit of a model containing cross-classified categories with that of a model where VAT volume and GRSCD were independent variables. Same methods were used to evaluate possible interactions with individual CD SNPs. We used SPSS Statistics version 20 (IBM Corp., Armonk, NY) to perform all analyses. All P-values were 2-sided and P < 0.05 was considered significant.
Among 482 patients with available information on VAT volume, 174 developed penetrating disease, 132 developed stricturing disease, 147 developed perianal disease, and 252 required surgery. Compared to the participants in the lowest quartile of VAT volume, patients in the highest quartile of VAT volume were older at the time of diagnosis, had a higher BMIs, and more likely to be smokers (Table 1).
The risk of surgery appeared to increase with increasing VAT volume (Ptrend = 0.006). Compared to participants in the lowest quartile of VAT volume, the MV-adjusted risk of surgery among participants in the highest quartile was 2.02, 95% CI 1.09 – 3.76 (Table 2). These results were not significantly altered in analyses where participants with penetrating disease were excluded (Ptrend = 0.003). Similarly, the risk of penetrating disease appeared to be positively associated with VAT volume (Ptrend = 0.02). Compared to participants in the lowest quartile of VAT volume, the MV-adjusted risk of penetrating disease was 1.95, 95% CI 1.04 – 3.67 in the highest quartile. Conversely, we did not observe an association between VAT volume and risk of perianal or stricturing disease (all Ptrend > 0.20). Compared to CD patients in the lowest quartile of VAT volume, participants in the highest quartile had a MV-adjusted OR of 0.68 (95% CI 0.35 – 1.32) for stricturing disease and 0.82 (95% CI 0.42 – 1.59) for penetrating disease. In exploratory analyses, we did not observe an effect modification by disease duration on the association between VAT and CD progression (All Pinteraction > 0.26).
Although VAT volume did not appear to differ significantly between imaging obtained before and after the occurrence of disease complications, we performed sensitivity analysis restricted to cases where a CT scan predated disease complications and obtained similar results. Among 223 eligible participants for these analyses, of which 70 patients required surgery, the risk of surgery appeared to increase with greater VAT volume (Ptrend = 0.027). Compared to participants in the lowest quartile of VAT volume, the multivariable (MV)-adjusted risk of surgery among participants in the highest quartile was 2.43, with a 95% CI 0.95 – 6.21. Since VAT volume appeared to vary according to age of diagnosis (Table 1), we also explored whether the association between VAT volume and risk of surgery and penetrating disease is modified by age of diagnosis and observed no interaction (Pinteraction = 0.09 and 0.99, respectively).
We explored the possibility that the association between VAT volume and risk of CD complications may be modified by genetic predisposition, and therefore evaluated for effect modification by genetic risk score (Figure 1). We failed to observe any evidence to support effect modification by genetic risk score (as a binary variable) on the associations between VAT volume and risk or surgery, or risk of stricturing, penetrating, or perianal disease, all Pinteraction > 0.20). Similar results were obtained when genetic risk score was treated as a continuous variable (all Pinteraction > 0.12). We also evaluated for potential interaction between CD SNPs within the risk score and VAT volume on risk of CD complications (Supplementary Table 3). Although there appeared to be trends toward interaction for some SNPs, none of these met the threshold for statistical significant after accounting for multiple testing (Bonferroni-corrected P = 0.00036).
In a large cohort of CD patients, we show that visceral adiposity as measured by VAT volume is associated with increased risk of surgery and penetrating disease but not perianal or stricturing disease. In addition, genetic predisposition does not appear to modify the association between VAT volume and risk of CD complications.
These findings are party supported by previous studies. Uko et al. evaluated the impact of VAT volume in a small cohort of pediatric IBD cases (n = 114), and showed that IBD patients have greater VAT volume values compared to healthy controls.18 In addition, at the time of diagnosis greater VAT volume was associated with increased risk of fistulizing disease. Similarly, Bunning and colleagues demonstrated a good correlation between index of VAT accumulation (VAT/total fat mass) and disease activity and complications in small cohort of adult women with CD (N = 31).19 Interestingly, a number of recent studies have also demonstrated an association between visceral adiposity and post-operative morbidity and disease recurrence in CD.20,21 Because these prior studies have small sample sizes and limited information on other key risk factors, our analysis which benefited from nearly 500 well-characterized CD patients and carefully adjusted for important confounders significantly extends these prior findings. Moreover, to our knowledge this is the first study evaluating the interaction between visceral adiposity and genetic predisposition on risk of CD complications.
Our findings are biologically plausible. Adipocytes secrete tumor necrosis factor-α (TNF- α) and express Toll-like receptor (TLR) suggesting a potential role for adipose tissue in regulating innate immunity.29 In addition, numerous proinflammatory and anti-inflammatory cytokines including IL-6, adiponectin, and resistin have been identified in adipocytes, further suggesting a link between adipocyte tissues and innate immunity.14,15,29–34 Interestingly, in comparison to subcutaneous adipose tissue, visceral adipose tissue appears to play a pivotal role in regulating systemic inflammation.35,36 Consistent with these observations, compared to total obesity, there appears to be a stronger association between visceral adiposity and risk of metabolic and inflammatory disorders.37–39 This is also consistent with our prior and current observations that visceral adiposity as measured by VAT and not total obesity as measured by BMI is likely associated with CD progression.12 Finally, there is growing data suggesting infiltration of visceral adipose tissue by macrophages and gut microbiota, raising the possibility of local crosstalk between immune function, visceral adipose tissue, and the gut microbiota in development and progression of CD.13,40,41
Our study has several important strengths that are worth noting. First, our study represents the largest and most comprehensive study evaluating the impact of visceral adipose tissue on CD progression in the context of genetic predisposition. Second, we confirmed all cases of CD and diagnoses of CD complications through review of medical records, rather than relying on diagnostic codes or self-reported information. Third, in our analyses, we were able to adjust for confounding using detailed information on relevant lifestyle factors, including smoking and BMI. Fourth, we used a previously validated tool to measure VAT volume in all patients minimizing the possibility of measurement errors.
We acknowledge several limitations. First, participants were not systematically screened for presence of disease complications and therefore subclinical complications may have remained undiagnosed in some participants. However, the average disease duration of our participants at the time of enrollment was over 10 years making it unlikely for subclinical disease complications to have gone undiagnosed. Second, due to the cross-sectional nature of our study, we were unable to account for changes in the VAT volume overtime. In addition, we acknowledge that VAT volume may also be affected by CD medications including steroids and anti-TNF therapy, and therefore our observed association could represent a proxy effect. However, we showed that among individuals with multiple CT scans, there was no statistically significant difference between VAT volume before and after disease complications. In addition, in our sensitivity analysis, limiting our study population to individuals with VAT volume measured prior to development of complications, yielded similar results. Finally, in our exploratory analyses, we did not observe an effect modification by disease duration on the association between VAT and CD complications. Additionally, we adjusted for anti-TNF therapy and ever use of steroids in our multivariable analyses minimizing the possibility that our observed association may be explained by use of CD medications. Third, although this is the first study evaluating the interaction between visceral adiposity and genetic predisposition, we acknowledge that robust analysis of gene-environment interaction requires a far greater number of participants. Fourth, we acknowledge that our study is observational and therefore we cannot exclude the possibility of residual confounding particularly by factors such as family history of CD and NSAID’s that were not available in our cohort. Lastly, our study is based on an experience of a single tertiary referral center with a predominantly white population, therefore it is possible that our findings may not be applicable to other populations. However, the average age of diagnosis, rates of complications, and the distribution of life style factors such as smoking and BMI in our study population are similar to other large U.S. population cohorts.1,42
In conclusion, we show that visceral adiposity measured by VAT volume is associated with an increased risk of surgery and penetrating disease but not stricturing or perianal disease among CD patients. Taken together with previous findings that BMI is not associated with risk of CD complications, our data suggest that visceral rather than total adiposity, may negatively influence the long-term risk of progression of CD, regardless of genetic predisposition.
Grant Support: Dr. Khalili is supported by a career development award from the American Gastroenterological Association (AGA) and by National Institute of Diabetes and Digestive and Kidney Diseases (K23 DK099681).
Financial Disclosures: Dr. Ananthakrishnan is a member of the scientific advisory board for Exact Sciences, AbbVie, and Cubist pharmaceuticals. Dr. Khalili has received consultant fee from AbbVie. Dr. Yajnik has received consulting fees from NPS, Janssen Pharmaceuticals, and UCB.
Ethical Approval: The institutional review board at the Massachusetts General Hospital approved this study.
Author Contributions:KWJV - Acquisition of data; statistical analysis; interpretation of data; drafting of the manuscript.
ADJ - Acquisition of data; statistical analysis; interpretation of data.
DRB - Acquisition of data; interpretation of data; critical revision of the manuscript.
KKG - Acquisition of data; critical revision of the manuscript.
KOS - Acquisition of data; critical revision of the manuscript.
PL - Acquisition of data; critical revision of the manuscript.
JJG - Acquisition of data; critical revision of the manuscript.
CG- Acquisition of data; critical revision of the manuscript.
VY - Acquisition of data; critical revision of the manuscript.
ANA- Acquisition of data; critical revision of the manuscript.
BZA - Interpretation of data; critical revision of the manuscript.
RJX - Acquisition of data; critical revision of the manuscript.
HK – Study concept and design; acquisition of data; statistical analysis; interpretation of data; drafting of the manuscript.