Abstract
Background & Aims
Obesity has been associated with an increased risk of colonic diverticulosis. Evidence for this association is limited. We assessed whether anthropometric measures of obesity were associated with colonic diverticulosis.
Methods
We analyzed data from a prospective study of 623 patients undergoing screening colonoscopies from 2013 through 2015; colonoscopies included examinations for diverticulosis. Body measurements were made the day of the procedure. Multivariate analyses were performed using modified Poisson regression to estimate prevalence ratios (PRs) and 95% CIs while adjusting for confounding variables. All analyses were stratified by sex.
Results
Among men, there was no association between any measure of obesity and diverticulosis. After adjustment, women with an obese body mass index (BMI ≥ 30) had an increased risk of any diverticulosis (PR, 1.48; 95% CI, 1.08–2.04) compared to women with a normal body mass index (BMI 18.5–24.9). The strength of this association was greater for more than 5 diverticula (PR, 2.05; 95% CI, 1.23–3.40). There was no significant association between measures of central obesity and diverticulosis in women. Stratified by sex, colonic diverticulosis was significantly less prevalent in women compared to men before the age of 51 years (29% vs 45%, P=.06). The prevalence of diverticulosis did not differ by sex in older age groups.
Conclusions
In an analysis of data from 623 patients undergoing screening colonoscopies, we found that obesity (BMI ≥30) significantly increased the risk of colonic diverticulosis in women but not men. Colonic diverticulosis was less prevalent in premenopausal-age women compared with similar-age men. These findings suggest that sex hormones may influence the development of diverticulosis.
Keywords: diverticular disease, adiposity, male, female
Introduction
Colonic diverticula are commonly acquired defects in the colon wall.1 Diverticula can hemorrhage, become inflamed or infected, or rupture. These morbid complications are termed diverticular disease.2 Diverticular disease accounts for 1,920,970 ambulatory visits, 208,015 hospital admissions and 4,602 deaths every year in the United States.3 Health care expenditures attributable to colonic diverticulosis and diverticular disease total $5.5 billion dollars annually.3
Obesity has been associated with an increased risk of colonic diverticulosis,4–10 but evidence for this association is limited. Prior studies used body mass index but did not include measures of abdominal obesity, which is assessed by waist circumference, waist to hip ratio and waist to height ratio. Furthermore, prior studies did not assess for heterogeneity of these associations by sex, even though obesity in men and women differs significantly.
Understanding modifiable risk factors for colonic diverticulosis has the potential to improve prevention efforts for diverticular disease and improve our understanding of diverticulosis biology. In the present study, we assessed the association between anthropometric measures and the risk of colonic diverticulosis using data from a prospective study of patients undergoing screening colonoscopy and detailed examination for diverticulosis.
Methods
Using data from a colonoscopy-based study of colonic diverticulosis (NIH R01DK094738), we assessed whether anthropometric measures were associated with diverticulosis. The study recruited outpatients 30 years of age and older presenting for a screening colonoscopy at the University of North Carolina Hospital in Chapel Hill, North Carolina between 2013–2015. The study included patients with a complete examination to the cecum and at least a satisfactory preparation for a screening colonoscopy. The study excluded any patient with: 1) prior colonoscopy, 2) indication other than screening, 3) familial polyposis syndrome defined as greater than 100 polyps or FAP gene test positive, 4) evidence of colitis, either ulcerative, Crohn’s, radiation or infectious colitis, 5) previous colonic resection based on history and colonoscopy, or 6) previous colon cancer or adenomas. For this analysis, we excluded those with a history of diverticulitis (n=1). Informed consent was obtained from all participants. The University of North Carolina School of Medicine Institutional Review Board approved this study.
Anthropometric measures and covariates
Interviews were conducted over the telephone prior to the colonoscopy using a computer assisted telephone interview. The telephone interview used a structured questionnaire designed to obtain information about dietary and lifestyle exposures including smoking and alcohol use. Age and sex were extracted from the medical record. Race was self-reported. Diet was measured using a food frequency questionnaire developed by the National Cancer Institute.11 The instrument has been shown to reasonably estimate absolute intakes when validated against 24-hour recalls. Participants were asked to estimate consumption of dietary items during the year prior to colonoscopy. Physical activity was measured using the International Physical Activity Questionnaire.12 A research assistant followed a standard protocol to measure height, weight, hip and waist the day of the colonoscopy. The waist and hip were measured with a tape measure in centimeters. The waist measurement was taken at the midpoint between the lowest rib and the iliac crest. The participant was asked to breathe out gently at the time of the measurement for a more accurate reading. The hip measurement was taken at the fullest part of the hips over the buttocks. Two measurements were taken. If they differed by more than 1 cm, a third measurement was taken.
Colonoscopy and assessment of diverticulosis
All colonoscopies were performed by experienced board-certified attending gastroenterologists. The colonoscopy was complete to the cecum and the preparation judged to be at least adequate (excellent, good or fair) for a screening exam. Endoscopists were instructed to carefully examine the colon for diverticula and a research assistant was present during the entire procedure to enumerate the diverticula on a standard data collection form. The form included information on the exact number and location of diverticula.
Statistical analysis
Means and standard deviations were calculated for continuous variables and proportions for categorical data. We included four measures of obesity: body mass index, waist circumference, waist to height ratio and waist to hip ratio. Body mass index is a measure of general adiposity. The remaining three are measures of central obesity. For analysis in the models, waist circumference, waist to hip ratio and waist to height ratio were converted into categories (quartiles). The 10% change-in estimate approach was used to identify confounding variables. Several potential confounding variables were considered including age, race, education, NSAID and aspirin use, dietary fiber, red meat and fat intake, alcohol use, physical activity and total energy intake. In the final models, central measures of obesity were adjusted for the confounding variables identified using 10% approach. Because all measures of central obesity are co-linear with body mass index, we performed modified Poisson regression (Poisson regression with a robust variance estimator)13 and also Bayesian log-binominal models with moderately informative null-priors. The models produced similar results, so we only report results from modified Poisson analyses. Using likelihood ratio tests, there was significant heterogeneity of the association with exposures by sex. Therefore, all analyses were stratified by sex. Multivariate analyses were performed using modified Poisson regression to estimate prevalence ratios and 95% confidence intervals while adjusting for confounding variables. We performed analyses with all diverticulosis cases and cases stratified by number of diverticula (1–5, and >5). The analysis was performed using SAS 9.4 (SAS, Cary, North Carolina).
Results
Our analysis included 623 participants; 56% were women. The prevalence of diverticulosis increased with age (Figure 1). Before the age of 51, colonic diverticulosis was less prevalent in women (29%) compared to men (45%) (p=0.06; Figure 1). In the group 51–60 years old, the prevalence of colonic diverticulosis in women was 37% and in men 43% (p=0.31). After the age of 60, colonic diverticulosis was equally prevalent in women (55%) and men (57%) (p=0.90). Women with diverticulosis were older and more likely to have an obese body mass index compared to women without diverticulosis (Table 1). Among women with diverticulosis, 7% had proximal diverticulosis, 61% distal, and 31% both proximal and distal. The average number of diverticula was 14 (standard deviation 19). Men with diverticulosis were not more likely to be obese. Among men with diverticulosis, 8% had proximal diverticulosis, 61% distal, and 30% both proximal and distal. The average number of diverticula was 15 (standard deviation 21).
Figure 1.
Colonic diverticulosis on colonoscopy in screening only persons by age and sex
Table 1.
Study characteristics
| Women | Men | |||||
|---|---|---|---|---|---|---|
| Diverticulosis | No diverticulosis | p-value | Diverticulosis | No diverticulosis | p-value | |
| n = 135 | n = 214 | n = 124 | n = 150 | |||
| % or mean ± standard deviation | ||||||
| Age | 56.3 ± 6.8 | 53.2 ± 6.5 | <0.001 | 54.7 ± 7.4 | 53.5 ± 6.5 | 0.16 |
| Race | 0.11 | 0.44 | ||||
| White | 77.5% | 76.1% | 81.0% | 75.9% | ||
| Black | 22.5% | 20.7% | 18.2% | 21.4% | ||
| Other | 0.0% | 3.3% | 0.8% | 2.8% | ||
| Education | 0.17 | .46 | ||||
| Less than high school | 1.0% | 2.5% | 3.2% | 3.9% | ||
| High school | 15.8% | 7.5% | 14.7% | 7.8% | ||
| Some college | 23.8% | 26.1% | 20.0% | 18.6% | ||
| College graduate | 59.4% | 64.0% | 62.1% | 69.6% | ||
| Smoking status | 0.49 | .05 | ||||
| Never | 63.0% | 59.0% | 43.2% | 60.8% | ||
| Former | 24.0% | 30.0% | 39.0% | 26.5% | ||
| Current | 13.0% | 10.6% | 17.9% | 12.8% | ||
| Body mass index, kg/m2 | 0.04 | 0.36 | ||||
| Underweight (<18.5) | 2.2% | 3.3% | 2.4% | 0.0% | ||
| Normal (18.5–25) | 26.7% | 40.1% | 25.0% | 26.4% | ||
| Overweight (25–30) | 24.4% | 22.6% | 41.1% | 42.6% | ||
| Obese (>30) | 46.7% | 34.0% | 31.5% | 31.1% | ||
| Waist circumference, centimeters | 94.9 ± 18.4 | 88.1 ± 16.2 | 0.001 | 100.4 ± 15.0 | 99.6 ± 12.7 | 0.65 |
| Waist to hip ratio | 0.9 ± 0.1 | 0.9 ± 0.1 | 0.02 | 1.0 ± 0.1 | 1.0 ± 0.1 | 0.58 |
| Waist to height ratio | 0.6 ± 0.1 | 0.5 ± 0.1 | 0.002 | 0.6 ± 0.1 | 0.6 ± 0.1 | 0.92 |
| Physical activity per week, metabolic equivalent task | 2,814.2 ± 3,264.2 | 3,375.8 ± 4,248.8 | 0.23 | 4,306.8 ± 5800 | 3,256.5 ± 3970 | 0.14 |
| Daily dietary intake | ||||||
| Total energy intake, kilocalories | 1890.9 ± 712.6 | 1953.6 ± 646.9 | 0.47 | 2,324.0 ± 828 | 2,183.0 ± 923 | 0.27 |
| Total fiber, grams | 19.8 ± 9.3 | 19.8 ± 8.7 | 0.99 | 20.7 ± 8.1 | 21.8 ± 12.0 | 0.45 |
| Red meat, ounces | 1.2 ± 0.8 | 1.3 ± 1.0 | 0.42 | 2.0 ± 1.5 | 1.6 ± 1.2 | 0.06 |
| Fat, grams | 78.6 ± 31.7 | 85.1 ± 35.3 | 0.13 | 97.6 ± 40.4 | 93.3 ± 43.3 | 0.48 |
| Nonsteroidal antiinflammatory drug use per month | 0.09 | 0.43 | ||||
| Never | 48.0% | 56.3% | 55.9% | 64.4% | ||
| 1–4 times | 30.0% | 18.1% | 24.7% | 17.8% | ||
| >4 times | 22.0% | 25.6% | 19.4% | 17.8% | ||
| Aspirin use per month | 0.40 | 0.39 | ||||
| Never | 56.4% | 64.6% | 51.6% | 57.8% | ||
| 1–4 times | 32.7% | 27.3% | 45.2% | 36.3% | ||
| >4 times | 10.9% | 8.1% | 3.2% | 5.9% | ||
| Daily alcohol use, drinks | 0.8 ± 1.6 | 0.6 ± 1.2 | 0.49 | 0.9 ± 1.4 | 0.9 ± 2.0 | 0.88 |
Body mass index
After adjusting for confounding variables, women with an obese body mass index (BMI ≥ 30) had an increased risk of any diverticulosis (PR 1.48; 95% CI 1.08–2.04) compared to women with a normal body mass index (BMI 18.5–24.9) (Table 2). When we categorize obesity by severity, the strength of the association was greatest (PR 1.92; 95% CI 1.29–2.84) for women with the most severe class of obesity (Class III [BMI >40]). The strength of the association with an obese body mass index (BMI ≥ 30) was greater (PR 2.05; 95% CI 1.23–3.40) when considering women with >5 diverticula (versus 1–5 diverticula) (Table 4). Stratified by age category, women ≤ 50 (PR 1.99; 95% CI 0.79–5.05) and 51–60 years old (PR 1.63; 95% CI 1.07–2.47) with an obese body mass index (≥ 30) had an increased risk of diverticulosis compared to women with a normal body mass index (18.5–24.9) (Supplemental table). There was no association between body mass index and risk of diverticulosis among women >60 years old (PR 1.00; 95% CI 0.57–1.74).
Table 2.
Anthropometric measures and risk of colonic diverticulosis in women
| Body mass index, kg/m2 | |||||
| 18–25 | 25–30 | >30 | P for trend | ||
| Cases (n) | 36 | 33 | 63 | ||
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 1.37 (0.94, 2.00) | 1.57 (1.13, 2.18) | 0.006 | |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 1.27 (0.88, 1.85) | 1.48 (1.08, 2.04) | 0.01 | |
| Waist circumference, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 21 | 18 | 26 | 48 | |
| Range (cm) | 40, 76 | 77, 84 | 85, 98 | 99, 140 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 0.98 (0.58, 1.64) | 1.17 (0.73, 1.86) | 1.71 (1.15, 2.56) | 0.003 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 1.05 (0.63, 1.77) | 1.18 (0.75, 1.87) | 1.67 (1.12, 2.48) | 0.007 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.99 (0.57, 1.74) | 1.09 (0.57, 2.09) | 1.56 (0.74, 3.27) | 0.21 |
| Waist to hip ratio, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 25 | 21 | 25 | 42 | |
| Range | 0.70, 0.81 | 0.81, 0.86 | 0.86, 0.90 | 0.91, 1.14 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 0.91 (0.57, 1.45) | 0.99 (0.63, 1.53) | 1.39 (0.96, 2.03) | 0.06 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 0.95 (0.60, 1.50) | 1.00 (0.65, 1.54) | 1.35 (0.92, 1.99) | 0.10 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.87 (0.55, 1.39) | 0.88 (0.57, 1.38) | 1.13 (0.74, 1.71) | 0.49 |
| Waist to height ratio, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 19 | 19 | 30 | 45 | |
| Range | 0.22, 0.46 | 0.46, 0.52 | 0.52, 0.60 | 0.61, 0.85 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 1.00 (0.59, 1.69) | 1.34 (0.84, 2.12) | 1.66 (1.09, 2.54) | 0.005 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 1.05 (0.62, 1.77) | 1.32 (0.84, 2.07) | 1.60 (1.05, 2.43) | 0.01 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.99 (0.57, 1.71) | 1.14 (0.60, 2.18) | 1.27 (0.61, 2.65) | 0.50 |
BMI, body mass index; CI confidence intervals
adjusted for age
adjusted for age and body mass index
Table 4.
Anthropometric measures and risk by number of diverticula
| Body mass index, kg/m2 | ||||
| Women | Men | |||
| Adjusted prevalence ratios (95% CI)a | 18–25 | >30 | 18–25 | >30 |
| Number of diverticula | ||||
| 1–5 | 1.0 (reference) | 1.15 (0.67, 1.97) | 1.0 (reference) | 0.73 (0.36, 1.46) |
| >5 | 1.0 (reference) | 2.05 (1.23, 3.40) | 1.0 (reference) | 1.35 (0.82, 2.21) |
| Waist circumference, quartiles | ||||
| Women | Men | |||
| Adjusted prevalence ratios (95% CI)b | 1 | 4 | 1 | 4 |
| Number of diverticula | ||||
| 1–5 | 1.0 (reference) | 2.42 (0.65, 9.03) | 1.0 (reference) | 0.93 (0.24, 3.49) |
| >5 | 1.0 (reference) | 1.90 (0.63, 5.79) | 1.0 (reference) | 1.51 (0.62, 3.65) |
| Waist to hip ratio, quartiles | ||||
| Women | Men | |||
| Adjusted prevalence ratios (95% CI)b | 1 | 4 | 1 | 4 |
| Number of diverticula | ||||
| 1–5 | 1.0 (reference) | 1.40 (0.68, 2.89) | 1.0 (reference) | 0.80 (0.24, 2.65) |
| >5 | 1.0 (reference) | 1.09 (0.61, 1.94) | 1.0 (reference) | 1.00 (0.52, 1.94) |
| Waist to height ratio, quartiles | ||||
| Women | Men | |||
| Adjusted prevalence ratios (95% CI)b | 1 | 4 | 1 | 4 |
| Number of diverticula | ||||
| 1–5 | 1.0 (reference) | 1.44 (0.39, 5.36) | 1.0 (reference) | 0.59 (0.17, 2.09) |
| >5 | 1.0 (reference) | 1.74 (0.58, 5.22) | 1.0 (reference) | 1.08 (0.41, 2.85) |
CI confidence intervals
adjusted for age
adjusted for age and body mass index
Among men, there was no association between body mass index and risk of any diverticulosis. After adjusting for confounding variables, the risk for diverticulosis was similar for men with an obese body mass index (≥ 30) compared to men with a normal body mass index (18.5–24.9), (PR 1.07; 95% CI 0.75–1.51) (Table 3). There was also no association (PR 1.35; 95% CI 0.82–2.21) with risk of >5 diverticula (Table 4).
Table 3.
Anthropometric measures and risk of colonic diverticulosis in men
| Body mass index, kg/m2 | |||||
| 18–25 | 25–30 | >30 | P for trend | ||
| Cases (n) | 31 | 51 | 39 | ||
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 1.01 (0.72, 1.41) | 1.04 (0.73, 1.47) | 0.84 | |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 1.02 (0.73, 1.42) | 1.07 (0.75, 1.51) | 0.72 | |
| Waist circumference, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 31 | 15 | 29 | 30 | |
| Range (cm) | 65, 91 | 92, 97 | 98, 107 | 108, 147 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 0.78 (0.48, 1.27) | 1.04 (0.72, 1.51) | 1.06 (0.74, 1.53) | 0.55 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 0.78 (0.48, 1.26) | 1.03 (0.71, 1.49) | 1.06 (0.74, 1.53) | 0.57 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.84 (0.50, 1.41) | 1.16 (0.72, 1.89) | 1.20 (0.65, 2.21) | 0.50 |
| Waist to hip ratio, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 25 | 22 | 38 | 20 | |
| Range | 0.73, 0.92 | 0.92, 0.96 | 0.96, 1.00 | 1.01, 1.10 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 0.93 (0.60, 1.44) | 1.14 (0.79, 1.65) | 0.99 (0.64, 1.54) | 0.71 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 0.92 (0.60, 1.43) | 1.12 (0.78, 1.63) | 0.97 (0.63, 1.51) | 0.79 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.98 (0.61, 1.58) | 1.20 (0.80, 1.81) | 1.01 (0.60, 1.68) | 0.62 |
| Waist to height ratio, quartiles | |||||
| 1 | 2 | 3 | 4 | P for trend | |
| Cases (n) | 30 | 15 | 34 | 26 | |
| Range | 0.36, 0.51 | 0.51, 0.55 | 0.55, 0.60 | 0.60, 0.85 | |
| Crude prevalence ratios (95% CI) | 1.0 (reference) | 0.66 (0.40, 1.08) | 1.05 (0.74, 1.49) | 0.94 (0.64, 1.38) | 0.78 |
| Adjusted prevalence ratios (95% CI)a | 1.0 (reference) | 0.62 (0.38, 1.01) | 1.03 (0.72, 1.46) | 0.91 (0.62, 1.34) | 0.83 |
| Adjusted prevalence ratios (95% CI)b | 1.0 (reference) | 0.65 (0.37, 1.14) | 1.06 (0.63, 1.78) | 0.89 (0.47, 1.71) | 0.92 |
BMI, body mass index; CI confidence intervals
adjusted for age
adjusted for age and body mass index
Waist circumference
After adjusting for confounding variables, there was no significant association between women in the highest quartile of waist circumference, compared with women in the lowest quartile (PR 1.56; 95% CI, 0.74–3.27) (Table 2). There was no association between men in the highest quartile of waist circumference, compared with men in the lowest quartile (PR 1.20; 95% CI, 0.65–2.21) for any diverticulosis (Table 3). There was no association when limited to women or men with >5 diverticula (Table 4).
Waist-to-hip
There was no significant association between waist to hip ratio and any diverticulosis when comparing the highest quartile of this ratio with the lowest among women (PR 1.13; 95% CI 0.74–1.71) or men (PR 1.01; 95% CI 0.60–1.68) (Tables 2 and 3). There was no association when limited to >5 diverticula (Table 4).
Waist-to-height
After adjusting for confounding variables, there was no significant association between women in the highest quartile of waist-to-height ratio, compared with women in the lowest quartile (PR 1.27; 95% CI, 0.61–2.65). Among men, waist-to-height ratio was not associated with the risk of any diverticulosis when the highest and lowest quartiles were compared (PR 0.89; 95% CI, 0.47–1.71). There was no association when limited to women or men with >5 diverticula (Table 4).
Discussion
In this prospective colonoscopy-based study, general obesity was significantly associated with the risk of colonic diverticulosis among women, but not among men. These associations were dose-dependent and stronger when considering >5 diverticula. We also found that premenopausal age women appear to be protected from diverticulosis compared with age-comparable men. There was no difference in the prevalence of diverticulosis by sex after the age of 50.
Previous studies in Western populations have found that obesity, measured by body mass index only, is a risk factor for diverticulosis.4–10 In contrast with prior work, the present study determined whether participants had colonic diverticulosis in a standard manner by colonoscopy, examined the association between multiple measures of obesity and diverticulosis, and accounted for effect measure modification between measures of obesity and sex. Similar to our work, a genetic study of colonic diverticulosis in Germany and Lithuania also found that obesity was associated with an increased risk of diverticulosis in women but not men.14
General obesity, not central obesity, significantly increased the risk of colonic diverticulosis among women but not men. This elevated risk was limited to women before the age of 60. Obesity in premenopauseal women is associated with higher concentrations of free testosterone and lower concentrations of sex hormone-binding globulin, and total estrogen.15 We hypothesize that premenopausal ovarian steroid hormones may protect women from developing diverticulosis. Premenopauseal obesity could increase the risk by decreasing circulating estrogen and increasing free testosterone. Obesity is associated with alterations in the gut microbiome, however in a prior study we found little association between the mucosal adherent microbial community composition and colonic diverticulosis.16
The lower prevalence of diverticulosis among women compared with men before the age of 51 is a novel finding and suggests that ovarian steroid hormones in premenopausal women may reduce the risk of diverticulosis. The lower prevalence of diverticulosis in women is congruent with a population-based study that found that diverticulitis was more common in men compared with women before the age of 49. After the age of 60, diverticulitis was more common in women.17 Genome-wide association analyses suggest that some of the risk of developing diverticulosis is attributable to impaired intestinal neuromuscular function, mesenteric vascular smooth muscle and connective tissue.14, 18, 19 Sex hormones alter elastin and collagen synthesis,20, 21 and endogenous ovarian steroid hormones may therefore reduce the risk of diverticulosis in premenopausal women because of beneficial effects of steroid hormones on collagen and/or elastin.
The lower prevalence of diverticulosis among women compared with men before the age of 51 is another argument against the hypothesis that constipation is the root cause of diverticulosis. Compared to women, men more commonly report looser and more frequent bowel movements.22 If the hypothesis that colonic diverticulosis results from constipation and higher intraluminal pressures were true, then we would expect a higher prevalence of diverticulosis in women compared with men. 6, 23 Instead, we found the opposite.
Central obesity, defined by waist circumference and waist to hip ratio, is a risk factor for diverticulitis and diverticular hemorrhage.24–26 Menopausal hormone therapy may also be associated with an increased risk of diverticulitis.27 Because diverticulitis and diverticular hemorrhage are distinct pathologies, risk factors for developing colonic diverticulosis must be considered separately from risk factors for diverticular diseases.
This work has limitations. A research assistant measured height, weight, hip and waist the day of the colonoscopy. Measurements were taken after the patients took the preparation for colonoscopy. Because every participant was prepped for the procedure, we would expect uniform changes in waist measurements. This would be a non-differential measurement bias towards the null. We have no data on anthropometric measurements prior to the colonoscopy and could not assess whether early life obesity or weight gain over time is associated with diverticulosis. Furthermore, this is a cross-sectional study that assessed associations and cannot be used to determine causality. Despite a detailed colonoscopy performed by an experienced gastroenterologist with a trained research assistant present to record data, colonic diverticula may have been missed. If diverticula were missed, this non-differential misclassification would bias our results towards the null. Because diverticulosis limited to the proximal colon was rare, we could not assess whether there was a differential association between distal and proximal diverticulosis. Finally, although the sex/age stratified results in the supplemental table are intriguing, we stress that these stratified results are based on small numbers and should be interpreted cautiously.
In conclusion, general obesity significantly increased the risk of colonic diverticulosis among women. There was no association between any measure of obesity and diverticulosis among men. Diverticulosis is less prevalent in premenopausal age women compared with age-comparable men. These sex differences may have implications that lead to a better understanding of diverticulosis pathogenesis.
Supplementary Material
Need to Know.
Background
We assessed whether anthropometric measures of obesity were associated with colonic diverticulosis.
Findings
In an analysis of data from 623 patients undergoing screening colonoscopies, we found that obesity significantly increased the risk of colonic diverticulosis in women but not men. Colonic diverticulosis was less prevalent in premenopausal-age women compared with similar-age men.
Implications for Patient Care
In women, obesity might contribute to development of diverticulosis.
Acknowledgments
Grant Support: This research was supported in part by grants from the National Institutes of Health K23DK113225, R01DK09473, and P30 DK034987
Footnotes
Conflicts of Interest/Disclosures: None
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Everhart JE, Ruhl CE. Burden of digestive diseases in the United States part I: overall and upper gastrointestinal diseases. Gastroenterology. 2009;136(2):376–86. [DOI] [PubMed] [Google Scholar]
- 2.Strate LL, Morris AM. Epidemiology, Pathophysiology, and Treatment of Diverticulitis. Gastroenterology. 2019. Epub 2019/0½1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Peery AF, Crockett SD, Murphy CC, et al. Burden and Cost of Gastrointestinal, Liver, and Pancreatic Diseases in the United States: Update 2018. Gastroenterology. 2019;156(1):254–72 e11 Epub 2018/10/14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mashayekhi R, Bellavance DR, Chin SM, et al. Obesity, but Not Physical Activity, Is Associated With Higher Prevalence of Asymptomatic Diverticulosis. Clin Gastroenterol Hepatol. 2018;16(4):586–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wijarnpreecha K, Ahuja W, Chesdachai S, et al. Obesity and the Risk of Colonic Diverticulosis: A Meta-analysis. Dis Colon Rectum. 2018;61(4):476–83. [DOI] [PubMed] [Google Scholar]
- 6.Peery AF, Sandler RS, Ahnen DJ, et al. Constipation and a low-fiber diet are not associated with diverticulosis. Clin Gastroenterol Hepatol. 2013;11(12):1622–7. Epub 2013/07/31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Dore MP, Pes GM, Marras G, et al. Risk factors associated with colonic diverticulosis among patients from a defined geographic area. Tech Coloproctol. 2016;20(3):177–83. [DOI] [PubMed] [Google Scholar]
- 8.Comstock SS, Lewis MM, Pathak DR, et al. Cross-sectional analysis of obesity and serum analytes in males identifies sRAGE as a novel biomarker inversely associated with diverticulosis. PLoS One. 2014;9(4):e95232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Teixeira C, Trabulo D, Ribeiro S, et al. Colonic diverticulosis and the metabolic syndrome: an association? Rev Esp Enferm Dig. 2017;109(11):768–71. [DOI] [PubMed] [Google Scholar]
- 10.Peery AF, Keku TO, Martin CF, et al. Distribution and Characteristics of Colonic Diverticula in a United States Screening Population. Clin Gastroenterol Hepatol. 2016;14(7):980–5 e1 Epub 2016/02/14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Subar AF, Thompson FE, Kipnis V, et al. Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires : the Eating at America’s Table Study. Am J Epidemiol. 2001;154(12):1089–99. [DOI] [PubMed] [Google Scholar]
- 12.Craig CL, Marshall AL, Sjostrom M, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381–95. [DOI] [PubMed] [Google Scholar]
- 13.Zou G A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–6. Epub 2004/03/23. [DOI] [PubMed] [Google Scholar]
- 14.Reichert MC, Kupcinskas J, Krawczyk M, et al. A Variant of COL3A1 (rs3134646) Is Associated With Risk of Developing Diverticulosis in White Men. Dis Colon Rectum. 2018;61(5):604–11. Epub 2018/03/14. [DOI] [PubMed] [Google Scholar]
- 15.Tworoger SS, Eliassen AH, Missmer SA, et al. Birthweight and body size throughout life in relation to sex hormones and prolactin concentrations in premenopausal women. Cancer Epidemiol Biomarkers Prev. 2006;15(12):2494–501. Epub 2006/12/14. [DOI] [PubMed] [Google Scholar]
- 16.Jones RB, Fodor AA, Peery AF, et al. An Aberrant Microbiota is not Strongly Associated with Incidental Colonic Diverticulosis. Sci Rep. 2018;8(1):4951 Epub 2018/03/23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bharucha AE, Parthasarathy G, Ditah I, et al. Temporal Trends in the Incidence and Natural History of Diverticulitis: A Population-Based Study. Am J Gastroenterol. 2015;110(11):1589–96. Epub 2015/09/30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Schafmayer C, Harrison JW, Buch S, et al. Genome-wide association analysis of diverticular disease points towards neuromuscular, connective tissue and epithelial pathomechanisms. Gut. 2019. Epub 2019/0½1. [DOI] [PubMed] [Google Scholar]
- 19.Maguire LH, Handelman SK, Du X, et al. Genome-wide association analyses identify 39 new susceptibility loci for diverticular disease. Nat Genet. 2018;50(10):1359–65. Epub 2018/09/05. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kanda N, Watanabe S. Regulatory roles of sex hormones in cutaneous biology and immunology. J Dermatol Sci. 2005;38(1):1–7. Epub 2005/03/30. [DOI] [PubMed] [Google Scholar]
- 21.Leblanc DR, Schneider M, Angele P, et al. The effect of estrogen on tendon and ligament metabolism and function. J Steroid Biochem Mol Biol. 2017;172:106–16. Epub 2017/06/21. [DOI] [PubMed] [Google Scholar]
- 22.Mitsuhashi S, Ballou S, Jiang ZG, et al. Characterizing Normal Bowel Frequency and Consistency in a Representative Sample of Adults in the United States (NHANES). Am J Gastroenterol. 2018;113(1):115–23. Epub 2017/08/02. [DOI] [PubMed] [Google Scholar]
- 23.Peery AF, Barrett PR, Park D, et al. A high-fiber diet does not protect against asymptomatic diverticulosis. Gastroenterology. 2012;142(2):266–72 e1 Epub 2011/11/09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Strate LL, Liu YL, Aldoori WH, et al. Obesity increases the risks of diverticulitis and diverticular bleeding. Gastroenterology. 2009;136(1):115–22 e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ma W, Jovani M, Liu PH, et al. Association Between Obesity and Weight Change and Risk of Diverticulitis in Women. Gastroenterology. 2018;155(1):58–66 e4 Epub 2018/04/04. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hjern F, Wolk A, Hakansson N. Obesity, physical inactivity, and colonic diverticular disease requiring hospitalization in women: a prospective cohort study. Am J Gastroenterol. 2012;107(2):296–302. [DOI] [PubMed] [Google Scholar]
- 27.Jovani M, Ma W, Joshi AD, et al. Menopausal Hormone Therapy and Risk of Diverticulitis. Am J Gastroenterol. 2019;114(2):315–21. Epub 2019/02/08. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.