Obesity and colorectal adenomatous polyps: A cross-sectional study in Korean adults
Ethical approval was obtained from the IRB of National Cancer Center.
Author contributions: JY Lee, SK Myung, and SH Jee are responsible for the initial plan, study design, and statistical analysis and for conducting the study. JY Lee, SM Kwak, SK Myung, and SH Jee are responsible for data collection, data extraction, data interpretation, and manuscript drafting. SK Myung and SH Jee are the guarantor for this paper and has full responsibility for this study.
Abstract
Objective
This study aimed to examine the association between obesity and the risk of colorectal adenomatous polyps.
Design and Methods
The design of this study is a cross-sectional study. A total of 1,574 participants (818 males and 756 females), who had undergone colonoscopy for health examinations at the National Cancer Center hospital in South Korea from February to September 2009, were included in the analysis.
Results
The prevalence of colorectal adenomatous polyps was 39.5% in males and 22.6% in females. In multiple logistic regression analysis, obesity (odds ratio [OR] = 1.69, 95% CI = 1.16-2.48) and abdominal obesity (OR = 1.59, 95% CI = 1.13-2.23) were significantly associated with an increased risk of colorectal adenomatous polyps in males, but not in females. In the subgroup analyses by age, obesity (OR = 2.27, 95% CI = 1.32-3.89), and abdominal obesity (OR = 1.76, 95% CI = 1.12-2.76) significantly increased a risk of colorectal adenomatous polyps in males aged <50 years, but not in males aged ≥50 years. However, interaction terms were not significant between obesity and gender (P = 0.397) and between obesity and age (P = 0.531) in the overall analyses.
Conclusion
Our findings should be explored in further research.
Introduction
According to the World Health Organization's report, colorectal cancer (CRC) is one of the most common fatal cancers in the world, the second cause of cancer mortality in the regions of the Americas, and the fourth in the Western Pacific areas (1). Additionally, the incidence rate of CRC has continued to increase in the economically transitioning countries, including Eastern European countries, most parts of Asia, and some countries of South America (2). The increasing rate of CRC in developing countries may be related to the adoption of western lifestyles and behaviors, such as the high consumption of saturated animal fat, low consumption of vegetables or fiber, physical inactivity, and smoking (2). Meanwhile, about 80% of CRCs are sporadic and caused by the interaction between genetic and various environmental factors. On the basis of the adenoma-carcinoma sequence, adenomatous polyps are known as main precursors for CRCs and precancerous lesions. Therefore, it is imperative that we identify the etiology of adenomatous polyps and elucidate the risk factors associated with adenomatous polyps for the prevention of CRCs (5).
Especially, dietary habits and nutritional status are known as important modifiable life-style related risk factors for adenomatous polyps or CRCs (6, 7). A large prospective cohort study showed that a high consumption of fruits and vegetables is associated with a reduced risk of CRCs (6). Moreover, the intake of soluble fiber may prevent the risk of distal colon adenomatous polyps (7).
Many studies suggested that there is an association between CRC and obesity, as indicated by high BMI or waist circumference (WC) (8). As for the association between obesity and colorectal adenomatous polyps, even though most epidemiological studies including cross-sectional studies, case-control studies, and cohort studies (11) reported a positive association, several studies (20) showed no significant association. Moreover, the findings on the relationship between obesity and the risk of colorectal polyps were inconsistent by gender (12, 14, 15, 19) or types of body measures for obesity such as BMI and WC (12, 19).
Also, several studies reported that the association between obesity and the risk of CRC or adenomatous polyps among women was confined to specific subgroups related to age or estrogen status (15, 21, 26, 27). Terry et al. (26) found that there was an increased risk of CRCs among younger women (e.g., <55 years), but not in older women. Hou et al. (27) reported that obesity was significantly associated with CRCs risk among pre-menopausal women, but not in post-menopausal women. Consequently, the changes in sex hormone levels such as estrogen, progesterone, or testosterone in middle-aged men and women cause andropause or menopause, which may subsequently alter the relation between obesity and the formation of adenomatous polyps as a confounding factor. Hu et al. (20) showed that old age (≥50 years) was associated with increasing risk of adenomatous polyps in a Taiwanese population.
We conducted a cross-sectional study to further evaluate the association between obesity or abdominal obesity and the risk of colorectal adenomatous polyps among Korean adults by gender and age, especially two groups divided by the age of 50.
Methods
Study population
We included 3,653 participants (≥19 years old) who visited the Center for Cancer Prevention and Detection at the National Cancer Center, South Korea, from February to September 2009 for routine health examinations, which are not covered by medical insurance, and underwent colonoscopy. Participants with histologically confirmed colorectal adenomatous polyps such as tubular, villous, or dysplastic adenoma were defined as an adenoma group, and the remaining participants who showed normal findings or other polyps including hyperplastic polyps or nonspecific changes were defined as a nonadenoma group. We excluded a total of 2,079 participants because they had undergone colonoscopy in the past (n = 1,901), did not provide any information on whether they underwent it or not (n = 154), were histologically confirmed as malignant colorectal neoplasm (n = 7), and had insufficient data on obesity (n = 17). We excluded participants with a history of colonoscopy because it could affect those lifestyles and other behaviors (especially those with positive findings on colonoscopy) and eventually affect the development of colorectal adenomatous polyps. An individual with multiple adenomas was analyzed based on the most malignant one.
Data collection on baseline sociodemographic characteristics and medical history
Each participant answered a self-administered questionnaire including detailed personal information for sociodemographic factors, personal or familial medical history about colonic polyps, cancer, hypertension (HTN), diabetes mellitus (DM), current medication, and lifestyle behaviors such as smoking, alcohol consumption, and exercise. Also, all the participants took blood biochemical tests and physical examinations including anthropometric measurements.
Anthropometric measurements and definition of obesity
In anthropometric measurement, height was measured to the nearest 0.1 cm and weight to the nearest 0.1 kg using X-Scan Plus II Body Composition Analyzer (Jawon Medical Co. Ltd., 2006, Korea). BMI was calculated as the individual's body weight (kg) divided by the square of his or her height (m2), and WC was measured to the nearest 0.1 cm, at the narrowest part of abdomen between the lowest rib cage and the hip. The BMI categorization was based on the WHO Asia-Pacific classification: “normal” (BMI < 23.0), “overweight (23 ≤ BMI <25), and “obese” (BMI ≥ 25) (28). Abdominal obesity was defined as a WC of ≥90 cm for men and ≥85 cm for women, based on the recommendations of the Korean Society for the Study of Obesity (KSSO) (29).
Colonoscopy procedure
After an adequate bowel preparation with a 4 L of polyethylene glycol-electrolyte oral lavage solution (Tae-Jun Pharma, Korea), colonoscopy was performed for each participant from the rectum to the cecum by well- experienced gastroenterologists using an EVIS-260 colonoscope (Olympus, Tokyo, Japan). All visualized lesions were biopsied and diagnosed with histological descriptions by pathologists.
Statistical analysis
Study participants were categorized according to age (<40, 40-49, 50-59, and ≥60 years), smoking status (none, former smoker, and current smoker), drinking status (none vs. current drinker), blood type (A, B, AB, and O), exercise (none vs. regular), and medical history of HTN, DM, or cancer. The χ2 test or student T-test was used to compare differences in categorical or continuous variables between the adenoma and nonadenoma groups.
Also, we separately compared differences in adenoma-related variables between the two groups by gender. Then, we performed multiple logistic regression analysis with adjustment for variables that showed a significant association with the outcome variables (P < 0.05) or had P-values of at least 0.2 in the univariable analyses. The odds ratios (ORs) and 95% confidence intervals (CIs) were presented.
We subsequently divided each gender group into two subgroups based on age (aged <50 yr vs. aged ≥50 yr) and separately analyzed the association between WC or BMI and the risk of adenoma by age group (30). Also, we tested the effect of obesity on the risk of adenoma in the overall logistic regression analyses by using interaction terms between obesity (WC or BMI) and gender (female vs. male) and between obesity and age (aged <50 yr vs. aged ≥50 yr). All the analyses were performed using SAS (version 9.1), and a P-value less than 0.05 was considered statistically significant.
Results
A total of 1,574 participants, who consisted of 494 (31.4%) in the adenoma group and 1,080 (68.6%) in the nonadenoma group, were included in the final analysis. Among them, 818 (52%) were males and 756 (48%) females. The mean age was 48.1 ± 10 years for males (range of 22-84 years) and 47.5 ± 9.8 for females (range of 21-77 years). Table 1 shows the frequency of colorectal adenomatous polyps by location. Frequent locations of colorectal adenomatous polyps were ascending colon (26.7%), sigmoid colon (23.3%), transverse colon (14.0%), and rectum (10.7%).
| Location | No. (%) |
|---|---|
| Cecum | 36 (7.3) |
| Ascending colon | 132 (26.7) |
| Hepatic flexure | 34 (6.9) |
| Transverse colon | 69 (14.0) |
| Splenic flexure | 6 (1.2) |
| Descending colon | 49 (9.9) |
| Sigmoid colon | 115 (23.3) |
| Rectum | 53 (10.7) |
| Total | 494 (100) |
Univariable analysis
Table 2 shows the differences in demographic and clinical characteristics between the adenoma and nonadenoma groups. Compared with the nonadenoma group, the adenoma group was more likely to be old, have higher BMI, larger WC, higher diastolic blood pressure, higher fasting blood glucose levels, and a history of HTN or DM, drink more, exercise regularly, and be a smoker (former or current). Similarly, in the subgroup analysis for males, the adenoma group was more likely to be old, have larger WC, higher fasting blood glucose levels, and a history of cancer, HTN, or DM. In the subgroup analysis for females, the adenoma group was more likely to be old, have higher BMI and higher fasting blood glucose levels, and exercise regularly.
| Variablea | Total (n = 1,574) | Male (n = 818) | Female (n = 756) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Nonadenoma group | Adenoma group | P | Non-adenoma group | Adenoma group | P | Non-adenoma group | Adenoma group | P | |
| No. of subjects (%) | 1,080 (68.6) | 494 (31.4) | 495 (60.5) | 323 (39.5) | 585 (77.4) | 171 (22.6) | |||
| Age, year | 46.1 ± 9.8 | 51.5 ± 9.2 | <0.001 | 46.0 ± 10.0 | 51.2 ± 9.2 | <0.001 | 46.2 ± 9.6 | 52.0 ± 9.1 | <0.001 |
| BMI, kg/m2 | 23.3 ± 3.2 | 24.1 ± 3.0 | <0.001 | 24.3 ± 2.9 | 24.6 ± 2.9 | 0.102 | 22.5 ± 3.3 | 23.2 ± 3.1 | 0.031 |
| <23 (%) | 519 (48.1) | 180 (36.4) | <0.001 | 170 (34.3) | 92 (28.5) | 0.104 | 349 (59.7) | 88 (51.5) | 0.125 |
| 23-25 (%) | 245 (22.7) | 126 (25.5) | 133 (26.9) | 83 (25.7) | 112 (19.1) | 43 (25.1) | |||
| ≥25 (%) | 316 (29.3) | 188 (39.1) | 192 (38.8) | 148 (45.8) | 124 (21.2) | 40 (23.4) | |||
| Waist Circumference, cm | 82.5 ± 9.2 | 85.1 ± 9.0 | <0.001 | 86.0 ± 7.7 | 87.3 ± 7.6 | 0.025 | 79.6 ± 9.3 | 81.0 ± 8.8 | 0.079 |
| Male <90, female <85 (%) | 775 (71.8) | 310 (62.8) | <0.001 | 348 (70.3) | 199 (61.6) | 0.010 | 427 (73.0) | 111 (64.9) | 0.040 |
| Male ≥90, female ≥85 (%) | 305 (28.2) | 184 (37.2) | 147 (29.7) | 124 (38.4) | 158 (27.0) | 60 (35.1) | |||
| Blood pressure | |||||||||
| Systolic, mmHg | 120.9 ± 15.0 | 122.2 ± 14.9 | 0.112 | 124.9 ± 14.7 | 123.9 ± 14.8 | 0.350 | 117.6 ± 14.4 | 119.0 ± 14.5 | 0.246 |
| Diastolic, mmHg | 73.9 ± 10.8 | 75.6 ± 10.6 | 0.003 | 77.0 ± 10.7 | 77.3 ± 10.1 | 0.708 | 71.2 ± 10.2 | 72.4 ± 10.9 | 0.174 |
| Fasting blood glucose, g/dL | 86.9 ± 19.7 | 92.7 ± 22.2 | <0.001 | 88.1 (19.2) | 94.1 ± 23.4 | <0.001 | 86.0 ± 20.1 | 90.1 ± 19.4 | 0.015 |
| Blood type (%) | 1050 | 479 | 0.339 | 484 | 310 | 0.299 | 566 | 169 | 0.913 |
| A | 371 (35.3) | 151 (31.5) | 168 (34.7) | 92 (29.7) | 203 (35.9) | 59 (34.9) | |||
| B | 294 (28.0) | 130 (27.1) | 138 (28.5) | 84 (27.1) | 156 (27.6) | 46 (27.2) | |||
| O | 276 (26.3) | 142 (29.6) | 126 (26.0) | 93 (30.0) | 150 (26.5) | 49 (29.0) | |||
| AB | 109 (10.4) | 56 (11.7) | 52 (10.7) | 41 (13.2) | 57 (10.1) | 15 (8.9) | |||
| Cancer history (%) | 1016 | 450 | 0.215 | 466 | 298 | 0.005 | 550 | 152 | 0.819 |
| − | 972 (69.7) | 423 (30.3) | 456 (62.0) | 279 (38.0) | 516 (78.2) | 144 (21.8) | |||
| + | 44 (62.0) | 27 (38.0) | 10 (34.5) | 19 (65.5) | 34 (81.0) | 8 (19.0) | |||
| Family history of cancer (%) | 985 | 439 | 0.525 | 466 | 294 | 0.503 | 519 | 145 | 0.503 |
| − | 492 (49.9) | 228 (51.9) | 228 (48.9) | 152 (51.7) | 264 (50.9) | 76 (52.4) | |||
| + | 493 (50.1) | 211 (48.1) | 238 (51.1) | 142 (48.3) | 255 (49.1) | 69 (47.6) | |||
| Hypertension (%) | 1080 | 494 | <0.001 | 495 | 323 | 0.004 | 585 | 171 | 0.103 |
| − | 935 (86.6) | 394 (79.8) | 431 (87.1) | 256 (79.3) | 504 (86.2) | 138 (80.7) | |||
| + | 145 (13.4) | 100 (20.2) | 64 (12.9) | 67 (20.7) | 81 (13.8) | 33 (19.3) | |||
| Diabetes mellitus (%) | 1080 | 494 | <0.001 | 495 | 323 | <0.001 | 585 | 171 | 0.098 |
| − | 1041 (96.4) | 447 (90.5) | 475 (96.0) | 256 (79.3) | 566 (96.8) | 160 (93.6) | |||
| + | 39 (3.6) | 47 (9.5) | 20 (40.0) | 67 (20.7) | 19 (96.8) | 11 (6.4) | |||
| Smoking status (%) | 1077 | 494 | <0.001 | 494 | 323 | 0.172 | 583 | 171 | 0.628 |
| None | 63.7 (59.1) | 209 (42.3) | 104 (21.1) | 51 (15.8) | 533 (91.4) | 158 (92.4) | |||
| Former | 214 (19.9) | 138 (27.9) | 192 (38.9) | 134 (41.5) | 22 (3.8) | 4 (2.3) | |||
| Current | 226 (21.0) | 147 (29.8) | 198 (40.1) | 138 (42.7) | 28 (4.8) | 9 (5.3) | |||
| Drinking status (%) | 1072 | 488 | 0.049 | 490 | 318 | 0.809 | 582 | 170 | 0.165 |
| − | 458 (42.7) | 182 (37.3) | 97 (19.8) | 66 (20.8) | 361 (62.0) | 116 (68.2) | |||
| + | 614 (57.3) | 306 (62.7) | 393 (80.2) | 252 (79.2) | 221 (38.0) | 54 (31.8) | |||
| Regular exercise (%) | 970 | 426 | 0.003 | 449 | 289 | 0.149 | 521 | 137 | 0.036 |
| − | 521 (53.7) | 191 (44.8) | 223 (49.7) | 127 (79.2) | 298 (57.2) | 64 (46.7) | |||
| + | 449 (46.3) | 235 (55.2) | 226 (50.3) | 162 (56.1) | 223 (42.8) | 73 (53.3) | |||
- a Values are means ± SD unless otherwise stated
Multivariable analysis
In the multiple logistic regression analysis, a significant increased risk of colorectal adenomatous polyps was observed in males (OR = 1.86, 95% CI = 1.26-2.75), older participants (OR = 2.48, 95% CI = 1.67-3.67 for aged 40-49 years; OR = 4.05, 95% CI = 2.66-6.16 for aged 50-59 years; OR = 5.11, 95% CI = 3.08-8.46 for aged ≥60 years), patients with DM (OR = 1.87, 95% CI = 1.11-3.17), participants with abdominal obesity (OR = 1.42, 95% CI = 1.06-1.89), and participants with obesity (OR = 1.41, 95% CI = 1.05-1.89) (Table 3).
| Variable | Total (N = 1,574)a | P | Male (n = 818)b | P | Female (n = 756)c | p |
|---|---|---|---|---|---|---|
| Gender | ||||||
| Female | 1 | n.a. | n.a. | |||
| Male | 1.86 (1.26-2.75) | 0.001 | n.a. | n.a. | ||
| Age, year | ||||||
| <40 | 1 | 1 | 1 | |||
| 40-49 | 2.48 (1.67-3.67) | <0.001 | 2.77 (1.68-4.59) | <0.001 | 2.02 (1.10-3.71) | 0.023 |
| 50-59 | 4.05 (2.66-6.16) | <0.001 | 5.05 (2.95-8.66) | <0.001 | 2.71 (1.38-5.33) | 0.004 |
| ≥60 | 5.11 (3.08-8.46) | <0.001 | 4.74 (2.48-9.05) | <0.001 | 5.41 (2.40-12.20) | <0.001 |
| Hypertension (%) | ||||||
| − | 1 | 1 | 1 | |||
| + | 1.03 (0.73-1.45) | 0.830 | 1.04 (0.66-1.63) | 0.864 | 1.01 (0.58-1.76) | 0.981 |
| Diabetes mellitus (%) | ||||||
| − | 1 | 1 | 1 | |||
| + | 1.87 (1.11-3.17) | 0.013 | 2.09 (1.07-4.07) | 0.031 | 1.60 (0.65-3.87) | 0.300 |
| Smoking status (%) | ||||||
| None | 1 | 1 | n.a | n.a | ||
| Former | 1.03 (0.69-1.55) | 0.828 | 1.18 (0.75-1.86) | 0.484 | n.a | n.a |
| Current | 1.45 (0.97-2.16) | 0.047 | 1.66 (1.05-2.62) | 0.032 | n.a | n.a |
| Drinking status (%) | ||||||
| − | 1 | 1 | 1 | |||
| + | 1.08 (0.81-1.46) | 0.632 | 0.98 (0.65-1.48) | 0.918 | 1.12 (0.72-1.73) | 0.615 |
| Regular exercise (%) | ||||||
| − | 1 | 1 | 1 | |||
| + | 1.15 (0.89-1.45) | 0.229 | 1.10 (0.80-1.54) | 0.530 | 1.20 (0.80-1.78) | 0.399 |
| Waist circumference, cm | ||||||
| Male <90, female <85 | 1 | 1 | 1 | |||
| Male ≥90, female ≥85 (abdominal obesity) | 1.42 (1.06-1.89) | 0.022 | 1.59 (1.13-2.23) | 0.008 | 1.03 (0.65-1.63) | 0.904 |
| Waist circumference, cm (continuous variable) | 1.02 (1.00-1.03) | 0.035 | 1.03 (1.01-1.06) | 0.004 | 1.00 (0.98-1.02) | 0.935 |
| BMI, kg/m2 | ||||||
| <23 | 1 | 1 | 1 | |||
| 23-25 (overweight) | 1.25 (0.91-1.70) | 0.168 | 1.45 (0.96-2.20) | 0.081 | 1.02 (0.62-1.68) | 0.936 |
| ≥25 (obesity) | 1.41 (1.05-1.89) | 0.024 | 1.69 (1.16-2.48) | 0.007 | 1.04 (0.61-1.74) | 0.897 |
| BMI, kg/m2 (continuous variable) | 1.05 (1.01-1.09) | 0.025 | 1.08 (1.02-1.14) | 0.009 | 1.01 (0.95-1.08) | 0.693 |
- n.a., not available; BMI, body mass index.
- a Adjusted for gender, age, smoking, drinking, regular exercise, history of hypertension, diabetes mellitus, and waist circumference.
- b Adjusted for age, smoking, drinking, regular exercise, history of hypertension, diabetes mellitus, and waist circumference.
- c Adjusted for age, drinking, regular exercise, history of hypertension, diabetes mellitus, and waist circumference. Smoking status was excluded for female because of its extremely low smoking prevalence (about 5%).
- Statistically significant: P < 0.05.
Similarly, in the subgroup analysis for males, a significant increased risk of colorectal adenomatous polyps was observed in older participants, patients with DM, current smokers, participants with abdominal obesity, and participants with obesity. In addition, WC (OR = 1.03, 95% CI = 1.01-1.06) and BMI (OR = 1.08, 95% CI = 1.02-1.14) as continuous variables were significantly associated with the increased risk of colorectal adenomatous polyps (Table 3).
In the subgroup analysis for females, a significant positive association with the risk of colorectal adenomatous polyps was observed only in older participants (Table 3).
In the multiple logistic regression analysis by age group, obesity (OR = 2.27, 95% CI = 1.32-3.89) and abdominal obesity (OR = 1.76, 95% CI = 1.12-2.76) showed an increased risk of colorectal adenomatous polyps in males aged <50 years, while there was no association in males aged ≥50 years (Table 4). The differences according to age and gender also were similarly observed for WC and BMI as a continuous variable. WC (OR = 1.04, 95% CI = 1.01-1.07) and BMI (OR = 1.09, 95% CI = 1.02-1.18) as continuous variables were significantly associated with the increased risk of colorectal adenomatous polyps in males aged <50 years, while there was no association in males aged ≥50 years or in females regardless of age groups.
| No. of subjects (%) | OR (95% CI)a | p | ||
|---|---|---|---|---|
| Non-adenoma group | Adenoma group | |||
| Male | ||||
| Aged <50 yrs (n = 483) | 332 | 151 | ||
| Waist circumference, cm | ||||
| <90 | 231 (69.6) | 88 (58.3) | 1 | |
| ≥90 | 101 (30.4) | 63 (41.7) | 1.76 (1.12-2.76) | 0.015 |
| Waist, cm (continuous variable) | - | - | 1.04 (1.01-1.07) | 0.007 |
| BMI | ||||
| <23 | 109 (32.8) | 36 (23.8) | 1 | |
| 23-25 | 93 (28.0) | 39 (25.8) | 1.66 (0.93-2.96) | 0.089 |
| ≥25 | 130 (39.2) | 76 (50.3) | 2.27 (1.32-3.89) | 0.003 |
| BMI (continuous variable) | - | - | 1.09 (1.02-1.18) | 0.016 |
| Aged ≥50 yrs (n = 335) | 163 | 172 | ||
| Waist circumference, cm | ||||
| <90 | 117 (71.8) | 111 (64.5) | 1 | |
| ≥90 | 46 (28.2) | 61 (35.5) | 1.36 (0.80-2.30) | 0.254 |
| Waist, cm (continuous variable) | - | - | 1.02 (0.98-1.05) | 0.334 |
| BMI | ||||
| <23 | 61 (37.4) | 56 (32.6) | 1 | |
| 23-25 | 40 (24.5) | 44 (25.6) | 1.24 (0.66-2.33) | 0.509 |
| ≥25 | 62 (38.0) | 72 (41.9) | 1.18 (0.68-2.06) | 0.551 |
| BMI (continuous variable) | - | - | 1.05 (0.95-1.15) | 0.346 |
| Female | ||||
| Aged <50 yrs (n = 461) | 385 | 76 | ||
| Waist circumference, cm | ||||
| <85 | 313 (81.3) | 63 (82.9) | 1 | |
| ≥85 | 72 (18.7) | 13 (17.1) | 0.94 (0.46-1.90) | 0.855 |
| Waist, cm (continuous variable) | - | - | 0.99 (0.96-1.02) | 0.506 |
| BMI | ||||
| <23 | 266 (69.1) | 49 (64.5) | 1 | |
| 23-25 | 56 (14.5) | 19 (25.0) | 1.41 (0.72-2.76) | 0.313 |
| ≥25 | 63 (16.4) | 8 (10.5) | 0.64 (0.27-1.53) | 0.310 |
| BMI (continuous variable) | - | - | 0.97 (0.88-1.06) | 0.481 |
| Aged ≥50 yrs (n = 295) | 200 | 95 | ||
| Waist circumference, cm | ||||
| <85 | 114 (57.0) | 48 (50.5) | 1 | |
| ≥85 | 86 (43.0) | 47 (49.5) | 1.06 (0.58-1.95) | 0.845 |
| Waist, cm (continuous variable) | - | - | 1.01 (0.97-1.04) | 0.647 |
| BMI | ||||
| <23 | 83 (41.5) | 39 (41.1) | 1 | |
| 23-25 | 56 (28.0) | 24 (25.3) | 0.73 (0.35-1.55) | 0.416 |
| ≥25 | 61 (30.5) | 32 (33.7) | 1.32 (0.66-2.64) | 0.441 |
| BMI (continuous variable) | - | - | 1.06 (0.97-1.17) | 0.218 |
- OR, odds ratio; 95% CI, 95% confidence interval.
- a Adjusted for age, smoking, drinking, regular exercise, history of hypertension, and diabetes mellitus. Smoking status was excluded for female because of its extremely low smoking prevalence (about 5%).
- Statistically significant: P < 0.05.
However, when we tested interaction terms, there was no significant association both between obesity and gender (P = 0.288 for WC and P = 0.397 for BMI) and between obesity and age (P = 0.796 for WC and P = 0.531 for BMI) in the overall analyses was insignificant (data not shown in Table).
Discussion
This cross-sectional study found that factors associated with an increased risk of colorectal adenomatous polyps were male gender, old age, obesity, abdominal obesity, smoking, and a history of DM. Also, obesity and abdominal obesity were significantly associated with an increased risk of colorectal adenomatous polyps only in males aged <50 years, but not in both males aged ≥50 years and females of both age group in the subgroup analyses of multiple logistic regression. However, interaction terms between obesity and gender or age in the overall analyses were not significant.
The important finding was that obesity and abdominal obesity were significantly associated with an increased risk of adenoma only in males aged <50 years, not in males aged ≥50 years. There have been few studies that investigated an association between obesity and colorectal adenoma according to age except for Kim et al.'s study (31), which indicated that obesity increased the risk of colonic adenoma in relatively young men. We are unable to suggest a definite reason for this finding because of a paucity of data available. However, it could be a potential mechanism that as IGF-1 decreases with age after puberty (32), older obese people might have a lower chance to be affected by IGF-1 than younger obese people. That is, obesity in older people could not increase the IGF-1 levels enough to proliferate colonic mucosal cells.
Another finding that obesity or abdominal obesity was not associated with an increased risk of colonic adenoma in women is compatible with that of the previous studies (18, 19). The pattern of body fat distribution in females could be a plausible mechanism for that finding. It has been indicated that men had a stronger association between obesity and the risk of colon cancer than women because men have a tendency toward having visceral adiposity and higher insulin levels (11). Also, it has been reported that the risk of CRC related with obesity was influenced by the woman's hormonal status (12). Estrogen may exert a protective effect through the insulin/IGF axis by reducing serum IGF levels (21) and alter the bile acid pool, which is believed to play an important role in colon carcinogenesis (35). Additionally, Yamaji et al. (36) identified that the insulin/insulin-like growth factor-1(IGF) axis acts differently by gender in colorectal carcinogenesis, at least in its early stage. Besides, several studies reported that the association between obesity and CRC risk among women would differ according to the levels of their estrogen status (premenopausal vs. postmenopausal) (15, 26, 27). However, when we separately analyzed for female groups divided by the age of 50, which was known to be the median age of menopause among Korean women (30), there were no significant associations between obesity and the risk of CRC.
The findings of the present study were similar to those of the previous ones, which reported that the prevalence of adenomatous polyps was found to be greater in obese male subjects than in controls (11, 13, 16). Obese males have a tendency toward abdominal distribution of fat, which could have a negative influence on insulin sensitivity and the levels of circulating IGF-1 (11). The increased levels of insulin and free IGF-1 promote proliferation of colonic mucosal cells and decrease those deaths, acting as contributors to colorectal carcinogenesis (11, 37). Also, several studies revealed that elevated levels of insulin and glucose were associated with an increased risk of adenoma and a decreased apoptosis in normal rectal mucosa (38, 39). Another possible explanation for the link between obesity and colorectal adenomatous polyp could be adiponection, the protein hormone exclusively produced by adipocytes. It modulates a number of metabolic processes including glucose regulation, fatty acid catabolism, and energy metabolism. Recent studies have demonstrated that patients with CRC had lower plasma adiponectin levels compared with the control group, as well as the negative correlation of adiponectin with colorectal adenomatous polyps (40).
However, our study was unable to determine that the effect of obesity on the risk of adenomatous polyps is different according to age or gender because of the nonsignificant results of interaction terms in the overall analyses, even though the findings in subgroup analyses indicated some possibilities. Therefore, considering that those interaction terms in the overall analyses were not significant, results in the subset analyses by gender and age should be interpreted with caution.
This study had several limitations. First, this study was a cross-sectional design. Therefore, we are unable to determine that obesity causes CRC. Second, there might be selection bias because we included the participants who voluntarily visited the hospital for routine health examinations, which are not covered with medical insurance and cost about over $1,000. That is, the study population included in this analysis might be a self-selected, higher socio-economic one than the general population in Korea. Therefore, they were more likely to be of a higher socioeconomic status than the general population. Last, we were unable to include nutritional variables such as meat consumption or fruit and vegetable intakes, which are known to be associated with an increased risk of CRCs or colorectal adenomatous polyps.
In conclusion, the current cross-sectional study showed that the prevalence of colorectal adenomatous polyp was significantly higher in males than in females, and obesity and abdominal obesity were associated with the increased risk of colorectal adenomatous polyp only in males aged <50 years, but not in both males aged 50 or older and females. However, these finding should be interpreted with caution because the interaction terms in the overall analyses were not statistically significant. Further large cohort studies are required to investigate the effect of obesity on the development of colorectal adenomatous polyps according to age and gender.
