Abstract
Bone mineral density (BMD) is a biomarker for cumulative exposure to multiple factors including estrogen, calcium, vitamin D and physical activity, which have all been independently associated with colorectal cancer. Furthermore, higher levels of BMD have been inversely associated with colorectal cancer risk, particularly in postmenopausal women. However, no prior studies have examined the potential association between BMD and colorectal adenomas, which are precursor lesions to most colorectal cancers. Therefore, we evaluated the association between BMD, which was measured using a whole body, dual-energy X-ray absorptiometry scan, and colorectal adenomas in 167 patients who underwent colonoscopy screening. We found that patients in the highest tertile of total body BMD (>1.294 g/cm2) and in the middle tertile (≥1.167 to ≤1.294 g/cm2) compared to those with a total body BMD in the lowest tertile (<1.167 g/cm2) had a lower risk of colorectal adenomas (highest vs. lowest tertile: OR=0.29 (0.10–0.84); middle vs. lowest tertile: OR=0.26 (0.08–0.80); p-trend=0.02). Stratification by gender revealed that this association was more pronounced in women (highest (>1.280 g/cm2) vs. lowest (<1.130 g/cm2) tertile: OR=0.08 (0.01–0.70); middle (≥1.130 to ≤1.280 g/cm2) vs. lowest tertile: OR=0.15 (0.04–0.94); p-trend=0.02) even after excluding hormone replacement therapy users (highest (>1.295 g/cm2) and middle (≥1.132 to ≤1.295 g/cm2) vs. lowest (<1.132 g/cm2) tertile: OR=0.17 (0.03–0.97); p-trend=0.04). Our results show, for the first time, that BMD is inversely associated with colorectal adenomas, particularly in women. Although additional larger, prospective studies are needed, our results suggest that BMD may be a biomarker for colorectal cancer precursor lesions.
Keywords: Bone mineral density, colorectal adenomas, biomarkers
Introduction
Colorectal cancer is the third most common non-skin cancer and the third leading cause of cancer death among men and women in the U.S. (1). The majority of colorectal cancers are believed to originate through the adenoma-(progression)-to-carcinoma paradigm (2); and, removing adenomas through colonoscopy has been estimated to decrease future colorectal cancer by 76% to 90% (3). Nevertheless, the persistence of colorectal cancer, which continues to affect over 150,000 individuals and account for over 50,000 deaths annually in the U.S. despite increased colonoscopy screening and subsequent adenomatous polyp removal (4, 5), suggests there is a need for identifying additional early detection markers.
Colorectal adenomas may arise through genetic syndromes such as familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC); however, the majority of sporadic adenomas appear to have a strong environmental component with increasing age, smoking and alcohol being associated with increased risk and, physical activity and non-steroidal anti-inflammatory drug (NSAID) being associated with a decreased risk (6–8).
Higher dietary intakes of calcium and circulating levels of vitamin D metabolites have been inversely associated with the risk of colorectal cancer (9–11) and colorectal adenomas (12–14). Calcium and vitamin D may act synergistically to decrease adenoma recurrence as evidenced, most notably, by the reduction in recurrence only among those patients who received calcium supplements and had higher serum levels of 25-hydroxy-vitamin D (25(OH)D) in the Calcium Polyp Prevention Study (15). Higher calcium levels have been shown to exert an anti-neoplastic effect in the colon epithelium by inducing higher levels of apoptosis (16, 17). Vitamin D also helps to maintain calcium homeostasis and may act directly on the colon epithelium by regulating apoptosis and cellular differentiation and by modulating growth factor and cytokine levels (18).
Exogenous hormone use among older and primarily postmenopausal women with lower endogenous production of estrogen has been shown to decrease the risk of colorectal cancer (19, 20, 20) as well as distal (21) and larger (≥ 1 cm) (22) colon adenomas. Although the exact mechanism driving this putative preventive effect is not known, estrogen’s ability to reverse age related decreases in calcium absorption and to increase serum levels of 25(OH)D3 and 1α-25(OH)2D3, the active form of vitamin D, which also affects bone metabolism, may play a role (23).
Bone mass as measured by bone mineral density (BMD) has been proposed as a marker of the cumulative exposure to endogenous and exogenous estrogens (24–26). However, BMD also reflects an individual’s lifetime exposure to calcium and the intake of other vitamins and minerals that affect calcium absorption and deposition, such as vitamin D, as well as exercise, particularly weight bearing exercise (27, 28). In addition, there is some evidence that smoking and alcohol use may affect BMD levels (29, 30). Thus, BMD may serve as a composite marker for an individual’s long term, synergistic exposure to several factors which, interestingly, have all been shown to be independently associated with the risk of colorectal cancer and adenomas including calcium, vitamin D, estrogen, smoking, and alcohol, as discussed above, as well as physical activity (31–34). Furthermore, utilizing BMD as an objective marker for long-term exposure to all of these factors combined may be particularly appealing due to the measurement error, bias and other problems associated with assessing lifetime lifestyle patterns using self-reported questionnaires and snap-shot serum biomarker data (35–37).
Higher BMD levels have been inversely associated with colorectal cancer (38) and colon cancer in postmenopausal women (39, 40). However, no prior studies have examined the potential association between BMD and colorectal adenomas. Therefore, we evaluated the potential association between BMD using a whole body dual-energy X-ray absorptiometry (DXA) scan and colon adenomas in 167 patients who underwent colonoscopy screening at University Hospitals in Cleveland, Ohio. We hypothesized that higher bone mineral density would be inversely associated with colorectal adenoma risk.
Materials and Methods
Study Population
The study population consisted of a subset of patients from a larger study conducted in the greater Cleveland, Ohio region involving patients undergoing routine colonoscopy screening at the University Hospitals Health System (UHHS) from January, 2006 through August, 2009. Patients were required to complete an initial screening questionnaire over the telephone to determine if they were eligible to participate in the study. Patients who reported having a personal history of any cancer, ulcerative colitis or Crohn’s disease, prior colorectal adenomatous polyps, a family history of HNPCC or FAP, or that were younger than 30 years old were excluded from further consideration.
Eligible patients having histologically confirmed adenomatous polyps in their colon and rectum, including tubular, sessile serrated or tubulovillous subtypes, were defined as cases. Patients with histologically confirmed colorectal cancer found during the colonoscopy were excluded. In the larger study, less than 1% of the patients were excluded for a diagnosis of colorectal cancer and no patients in this sub-study were excluded for a diagnosis of colorectal cancer. Patients who had hyperplastic polyps or no observed adenomas (negative colonoscopy) were included as controls in this study.
The recruitment rate among eligible participants was approximately 64.9 percent for this study. Subjects who declined to participate were demographically similar to those who agreed to participate. Specifically, patients who declined to participate compared to those who enrolled were of similar age (55.2 vs. 56.3 years), gender (59.6% vs. 52.7% female) and ethnic composition (60.4 vs. 67.1% Caucasian; 37.1 vs. 32.9% African-American). The study protocol was approved by the Institutional Review Boards of University Hospitals Case Medical Center and the Cleveland Clinic.
BMD and Other Risk Factor Measures
Eligible patients were required to complete a computer-aided personal interview (CAPI), prior to their colonoscopy, to assess lifestyle and behavioral risk factors. The CAPI was based on a risk factor questionnaire developed by the National Cancer Institute Colon Cancer Familial Cancer Registry (http://epi.grantss.cacner.gov/documents/CFR/center_questionnaires/Colon/LA/ColonRiskFactor_USC.pdf). A positive family history of colorectal cancer was defined as having one or more first degree relatives with colon or rectal cancer. Smokers were defined as patients who reported ever smoking cigarettes for 6 months or longer. Alcohol users were defined as those subjects who reported regular intake of alcohol defined as two or more drinks/week for six months or longer. Non-steroidal anti-inflammatory drug users were defined as subjects who reported using aspirin or ibuprofen at least twice a week for one month or longer.
Patients were also required to complete a validated semi-quantitative food frequency questionnaire (FFQ) (41) and a validated physical activity questionnaire (PAQ) (42) prior to their colonoscopy. Total daily intake values for calcium and vitamin D were determined by combining the values from the FFQ (41) for daily food and supplement intake reported in the year prior to colonoscopy. Physical activity was quantified from the PAQ (42) using the total frequency, duration and intensity reported for leisure time and recreational activities in the year prior to colonoscopy.
Subjects enrolled in the sub-study were also required to complete body composition testing, which included measurement of their height and weight (without shoes and in a gown) and completion of a whole body dual-energy X-ray absorptiometry (DXA) scan (Lunar iDXA™, GE Healthcare, Madison, WI) that was conducted at the Cleveland Clinic. Body mass index (BMI) was calculated as the ratio of measured weight in kilograms divided by height in meters squared. Total body and regional site (pelvis, spine) BMD were obtained by dividing the bone mineral content (BMC) by the projected area scanned using the Lunar iDXA™ software (GE Healthcare, Madison, WI).
Menopause was confirmed by the absence of menses for at least 12 months (as self-reported in the CAPI) and by high serum levels of follicle-stimulating hormone (FSH >40 mIU/mL) and/or low serum levels of estradiol (E2 <20 pg/mL). A 43 year old female who reported an absence of menses for at least 12 months and no HRT use was observed to have a FSH level of 4.18 mIU/mL and E2 level of 144 pg/mL; and, since we could not fully resolve her post-menopausal status, she was excluded from all analyses involving the menopause variable. Venous blood samples were collected just prior to colonoscopy and placed in polystyrene tubes containing sodium EDTA (1 mg/mL). The EDTA-containing tubes were chilled promptly in an ice bath and serum was separated by centrifugation at 1000g at room temperature for 15 minutes. Serum samples were stored at −80°C until assays were conducted. The FSH and E2 assays were run using commercial Coat-a-Count immunoradiometric assay (IRMA) and radioimmunoassay (RIA) kits, respectively (Siemans Medical Solutions Diagnostics, Los Angeles, CA). All samples were run in duplicate. The intra-assay coefficient of variation was 3.8% and 6.9% for FSH and E2, respectively. The inter-assay coefficient of variation was 3.9% and 5.5% for FSH and E2, respectively.
Statistical Analysis
We used unconditional logistic regression modeling to evaluate the potential associations between BMD and colorectal adenoma risk. We defined BMD tertiles using cut-point values observed in control subjects for each region (total body, pelvis, spine), separately; and, for male and female controls in each region separately for analyses stratified by gender. We employed a modeling strategy that evaluated minimally adjusted and a more fully adjusted model since the risk factors included in prior studies reporting an association between BMD and colorectal cancer varied. Following Nelson et al. (38), we evaluated a “minimally adjusted” model that included BMD, age, gender, race and BMI. In addition, we evaluated a “full” model (Model 1) similar to that of Zhang et al. (40) that included age, gender, race and BMI as well as other known factors for colorectal adenoma risk including family history of colon cancer, NSAID use, smoking, alcohol, physical activity, total calcium intake, total vitamin D intake and total energy intake.
We stratified our analyses by gender and menopausal status. Due to the small number of premenopausal women in our study population, we were unable to investigate this subgroup. Stratified analyses conducted in females were additionally adjusted for pre- vs. post-menopausal status, age at menopause and years of hormone replacement therapy (HRT) use. Because we were unable to fully characterize the type of HRT use, we also performed the analyses with HRT users excluded (Model 2).
All p-values reported are from two-sided tests. All analyses were undertaken with SAS (Version 9.1.3, SAS Institute Inc., Cary, NC).
Results
The characteristics of the colorectal adenoma screening study population are summarized in Table 1. In general, cases were similar to controls. However, cases were more likely to be male (61.2% vs. 38.8%; p<0.01) and older (58.6 ± 7.9 (yrs.) vs. 54.8 ± 7.4 (yrs); p<0.01) compared to controls. Cases also had a marginally significantly lower mean total daily calcium intake (1129.72 ± 615.38 (g/day) vs. 1313.39 ± 650.51 (g/day); p=0.07) and female cases had an earlier age at menopause (42.5 ± 8.9 vs. 46.5 ± 7.0 (yrs); p=0.08) compared to controls.
Table 1.
Characteristics of the Colorectal Adenomas Study Population
| Characteristic | Cases | Controls |
|---|---|---|
| Sample Size | 67 | 100 |
| Age (years) | 58.6 (7.9) | 54.8 (7.4) * |
| Gender | ||
| Males | 41 (61.2%) | 38 (38.0%) |
| Females | 26 (38.8%) | 62 (62.0%) * |
| Race | ||
| Caucasians | 42 (62.7%) | 70 (70.0%) |
| African-Americans | 25 (37.3%) | 30 (30.0%) |
| BMI (kg/m2) | 30.3 (7.2) | 29.4 (7.1) |
| Smoked | 41 (61.2%) | 48 (48.0%) † |
| Used Alcohol | 45 (67.2%) | 71 (71.0%) |
| Family History of Colon Cancer | 14 (20.9%) | 29 (29.0%) |
| NSAID** | 6 (9.2%) | 17 (17.0%) |
| Total Energy Intake (kcal/day) | 2135.3 (1626.7) | 2032.8 (890.1) |
| Calcium Dietary Intake (g/day) | 952.5 (603.3) | 1070.8 (542.2) |
| Vitamin D Dietary Intake (IU/day) | 143.0 (117.5) | 171.8 (129.6) |
| Calcium Supplements (g/day) | 177.2 (294.4) | 242.5 (396.3) |
| Vitamin D Supplements (IU/day) | 171.5 (236.4) | 195.0 (251.2) |
| Total Calcium Intake (g/day) | 1129.7 (615.4) | 1313.4 (650.5) † |
| Total Vitamin D Intake (IU/day) | 314.5 (272.1) | 366.8 (297.3) |
| Physical Activity (kJ/day) | 2521.7 (1046.0) | 2542.6 (1150.0) |
| Post-Menopausal Women | 18 (75.0%) | 39 (62.9%) |
| Age at Menopause (yrs) | 42.5 (8.8) | 46.7 (7.0) † |
| HRT Use in Post-Menopausal Women‡ | 8 (33.3%) | 13 (32.5%) |
| Years HRT Use In Post-Menopausal Women‡ | 10.9 (6.3) | 11.1 (8.0) |
| Follicle-Stimulating Hormone (FSH; mIU/mL) in Postmenopausal Women | 70.1 (30.5) | 68.1 (29.4) |
| Estrodial (E2; pg/mL) in Postmenopausal Women | 22.1 (20.0) | 18.1 (9.8) |
p≤0.05 for Chi-square test or t-test between cases and controls
0.05<p ≤ 0.10 Chi-square test or t-test between cases and controls
Non-steroidal anti-inflammatory drug (NSAID) use
Hormone replacement therapy (HRT) use was only found in post-menopausal women
We found an inverse association between higher total body BMD and colorectal adenomas in a minimally adjusted model that included BMD, age, gender, race and BMI (not shown). Specifically, subjects in the highest tertile (≥1.294 g/cm2) compared to those in the lowest tertile (<1.167 g/cm2 (lowest tertile) had a significantly lower risk of colon adenomas (OR=0.34; 95% C.I.: 0.13–0.93; p=0.03) (not shown). A marginal inverse association was observed in subjects in the middle tertile (≥1.167 to <1.294 g/cm2) compared to the lowest tertile (OR=0.42; 95% C.I.: 0.17–1.05; p=0.07) and the p-value for trend was 0.03. Adjustment for additional risk factors (Model 1, Table 2) improved precision and decreased effect estimates slightly but did not meaningfully change results (highest tertile vs. lowest tertile: OR=0.26; 95% C.I.: 0.08–0.80; p=0.02; middle vs. lowest tertile: OR=0.29; 95% C.I.: 0.10–0.84; p=0.02; p-trend=0.02). When evaluating BMD tertiles in the pelvic region, we found that subjects with a BMD exceeding 1.162 g/cm2 (highest tertile) compared to those with a BMD less than 1.001 g/cm2 (lowest tertile) had a decreased risk (OR=0.39; 95% C.I.: 0.15–1.00; p=0.05; p-trend=0.05). When examining the spinal region, subjects in the highest BMD tertile (≥1.220 g/cm2) compared to those in the lowest tertile (<1.050 g/cm2) had a marginally reduced risk of colorectal adenomas (OR=0.42; 95% C.I.: 0.16–1.13; p=0.09). Results for more fully adjusted models (Model 1, Table 2) in pelvic and spinal regions were similar to their respective minimally adjusted models.
Table 2.
Associations Between Bone Mineral Density (BMD) and Colorectal Adenomas
| Skeletal Site BMD Tertile (T) (g/cm2) | Model 1* |
Skeletal Site BMD Tertile (T) (g/cm2) | Model 2† |
||||
|---|---|---|---|---|---|---|---|
| N | OR (95% CI) ** | P | N | OR (95% CI) ** | p | ||
| Total Body | Total Body | ||||||
| T1: <1.167 | 24/33 | 1.00 (Referent) | - | T1: <1.169 | 20/29 | 1.00 (Referent) | - |
| T2:1.167–1.294 | 23/33 | 0.29 (0.10–0.84) | 0.02 | T2: 1.169–1.301 | 20/29 | 0.39 (0.13–1.03) | 0.07 |
| T3: >1.294 | 20/34 | 0.26 (0.08–0.80) | 0.02 0.02‡ |
T3: >1.301 | 19/29 | 0.38 (0.12–1.11) | 0.09 0.09‡ |
| T3 & T2 vs. T1 | 43/67 | 0.28 (0.10–0.74) | 0.01 | T3 & T2 vs. T1 | 39/58 | 0.38 (0.14–0.99) | 0.05 |
| Pelvis | Pelvis | ||||||
| T1: <1.001 | 26/33 | 1.00 (Referent) | - | T1: <0.998 | 21/29 | 1.00 (Referent) | - |
| T2: 1.001–1.162 | 24/33 | 0.61 (0.25–1.53) | 0.29 | T2: 0.998–1.167 | 21/29 | 0.67 (0.26–1.64) | 0.42 |
| T3: >1.162 | 17/34 | 0.38 (0.13–1.05) | 0.06 0.06‡ |
T3: >1.167 | 17/29 | 0.42 (0.15–1.12) | 0.10 0.10‡ |
| T3 vs. T2 & T1 | 17/67 | 0.51 (0.22–1.20) | 0.12 | T3 vs. T2 & T1 | 38/58 | 0.56 (0.23–1.27) | 0.20 |
| Spine | Spine | ||||||
| T1: <1.050 | 22/33 | 1.00 (Referent) | - | T1: <1.068 | 20/29 | 1.00 (Referent) | - |
| T2: 1.050–1.220 | 24/33 | 0.52 (0.19–1.36) | 0.18 | T2: 1.068–1.223 | 19/29 | 0.40 (0.14–1.05) | 0.08 |
| T3: >1.220 | 21/34 | 0.40 (0.13–1.10) | 0.09 0.11‡ |
T3: >1.223 | 19/29 | 0.38 (0.12–1.10) | 0.09 0.08‡ |
| T2 & T3 vs. T1 | 45/67 | 0.47 (0.19–1.18) | 0.11 | T2 & T3 vs. T1 | 38/58 | 0.39 (0.14–1.02) | 0.06 |
Adjusted for age, race, gender, BMI, smoking, alcohol, family history of colon cancer, NSAIDs, physical activity, total calcium, total Vitamin D, total energy intake
Excludes HRT users and is adjusted for all variables listed above in *.
Odds Ratio (OR) and 95% Confidence Interval (CI) of OR
p-value for trend
Because we were not able to fully characterize the type of hormone replacement therapy (HRT) use in females, we also performed the analyses with HRT users excluded. As shown in Model 2 (Table 2), we found a significant inverse association between total body BMD and colorectal adenomas when comparing the highest (>1.301 g/cm2) and middle (≥1.169–≤1.301 g/cm2) with the lowest (<1.169 g/cm2) tertile (OR=0.38; 95% C.I.: 0.14–0.99; p=0.05). We observed a marginally statistically significant association between spinal region BMD and adenomas when comparing the highest (>1.223 g/cm2) and middle (≥1.068–≤1.223 g/cm2) versus the lowest (<1.068 g/cm2) tertile when excluding HRT users (OR=0.39; 95% C.I.: 0.14–1.02; p=0.05).
Stratification by gender revealed that these associations were generally more pronounced among females (Table 3) compared to males (Table 4). Specifically, women with a total body BMD in the highest and middle tertiles combined compared to those in the lowest tertile had a lower colorectal adenoma risk (OR=0.29; 95% C.I.: 0.08–0.99; p=0.05) in the minimally adjusted model. Adjustment for additional risk factors (Model 1, Table 3) improved precision slightly when comparing women with total body BMD in the highest tertile (OR=0.08; 95% C.I.: 0.01–0.70; p=0.02) and middle tertile (OR=0.15; 95% C.I.: 0.04–0.94; p=0.04) to the lowest tertile (p-trend=0.02). When excluding HRT users (Model 2, Table 3), we observed similar results for total body BMD when comparing the highest (OR=0.06; 95% C.I.: 0.01–0.75; p=0.03) and middle (OR=0.16; 95% C.I.: 0.02–1.03; p=0.07) to the lowest tertile (p-trend=0.04).
Table 3.
Associations Between Bone Mineral Density (BMD) and Colorectal Adenomas: Females Only
| Skeletal Site BMD Tertile (T) (g/cm2) | Model 1* |
Skeletal Site BMD Tertile (T) (g/cm2) | Model 2† |
||||
|---|---|---|---|---|---|---|---|
| N | OR (95% CI) ** | p | N | OR (95% CI) ** | p | ||
| Total Body | Total Body | ||||||
| T1: <1.130 | 13/20 | 1.00 (Referent) | - | T1: <1.132 | 9/16 | 1.00 (Referent) | - |
| T2:1.130–1.280 | 5/20 | 0.15 (0.04–0.94) | 0.04 | T2: 1.132–1.295 | 4/16 | 0.16 (0.02–1.03) | 0.07 |
| T3: >1.280 | 8/21 | 0.08 (0.01–0.70) | 0.02 0.02‡ |
T3: >1.295 | 5/16 | 0.06 (0.01–0.75) | 0.03 0.04‡ |
| T2 & T3 vs. T1 | 13/61 | 0.13 (0.03–0.70) | 0.02 | T3 & T2 vs. T1 | 9/22 | 0.17 (0.03–0.97) | 0.04 |
| Pelvis | Pelvis | ||||||
| T1: <0.980 | 13/20 | 1.00 (Referent) | - | T1: <0.980 | 9/16 | 1.00 (Referent) | - |
| T2: 0.980–1.123 | 6/21 | 0.15 (0.04–0.95) | 0.04 | T2: 0.980–1.156 | 5/16 | 0.25 (0.04–1.24) | 0.14 |
| T3: >1.123 | 7/20 | 0.12 (0.02–1.00) | 0.05 0.04‡ |
T3: >1.156 | 4/16 | 0.13 (0.01–1.03) | 0.07 0.07‡ |
| T2 & T3 vs. T1 | 13/61 | 0.14 (0.03–0.73) | 0.02 | T3 vs. T2 & T1 | 9/22 | 0.21 (0.03–1.08) | 0.09 |
| Spine | Spine | ||||||
| T1: <1.017 | 13/20 | 1.00 (Referent) | - | T1: <1.018 | 9/16 | 1.00 (Referent) | - |
| T2: 1.017–1.209 | 6/20 | 0.12 (0.02–0.72) | 0.02 | T2: 1.018–1.220 | 4/16 | 0.09 (0.01–0.69) | 0.02 |
| T3: >1.209 | 7/21 | 0.04 (0.01–0.45) | 0.01 0.01‡ |
T3: >1.220 | 5/16 | 0.04 (0.01–0.56) | 0.01 0.01‡ |
| T2 & T3 vs. T1 | 13/61 | 0.09 (0.02–0.51) | 0.01 | T2 & T3 vs. T1 | 9/22 | 0.07 (0.01–0.52) | 0.01 |
Adjusted for age, race, gender, BMI, smoking, alcohol, family history of colon cancer, NSAIDs, physical activity, total calcium, total Vitamin D, total energy intake, menopausal status, age at menopause, years of HRT use
Excludes HRT users. Adjusted for all variables listed above in * except years of HRT use
Odds Ratio (OR) and 95% Confidence Interval (CI) of OR
p-value for trend
Table 4.
Associations Between BMD and Colorectal Adenomas: Males Only
| Skeletal Site BMD Tertile (T) (g/cm2) | N | Model 1* |
|
|---|---|---|---|
| OR (95% CI)† | p | ||
| Total Body | |||
| T1: <1.233 | 19/12 | 1.00 (Referent) | - |
| T2:1.233–1.312 | 9/13 | 0.23 (0.05–0.98) | 0.05 |
| T3: >1.312 | 13/13 | 0.34 (0.08–1.36) | 0.16 0.10** |
| T2 & T3 vs. T1 | 13/38 | 0.34 (0.11–1.07) | 0.07 |
| Pelvis | |||
| T1: <1.026 | 15/12 | 1.00 (Referent) | - |
| T2: 1.026–1.172 | 14/13 | 0.67 (0.19–2.43) | 0.55 |
| T3: >1.172 | 12/13 | 0.64 (0.17–2.41) | 0.51 0.51** |
| T2 & T3 vs. T1 | 41/38 | 0.66 (0.21–2.05) | 0.47 |
| Spine | |||
| T1: <1.105 | 16/12 | 1.00 (Referent) | - |
| T2: 1.105–1.263 | 15/13 | 0.63 (0.18–2.21) | 0.47 |
| T3: >1.263 | 10/13 | 0.25 (0.06–1.03) | 0.06 0.06** |
| T2 vs. T3 & T1 | 45/67 | 0.32 (0.10–1.09) | 0.07 |
Adjusted for age, race, gender, BMI, smoking, alcohol, family history of colon cancer, NSAIDs, physical activity, total calcium, total Vitamin D, total energy intake
Odds Ratio (OR) and 95% Confidence Interval (CI) of OR
p-value for trend
In the spinal region, women in the highest (OR=0.24; 95% C.I.: 0.06–0.94; p=0.04) and middle (OR=0.16; 95% C.I.: 0.03–0.84; p=0.03) BMD tertiles compared to those in the lowest tertile had a lower risk (p-trend=0.02) in the minimally adjusted model (not shown). Results for the more fully adjusted model (Model 1, Table 3) and the model excluding HRT users (Model 2, Table 3) were similar. In the pelvic region, women in the highest and second-highest BMD tertiles combined compared to those in the lowest tertile had a decreased risk of adenomas in minimally adjusted (OR=0.27; 95% C.I.: 0.08–0.94; p=0.04) and more fully adjusted (OR=0.14; 95% C.I.: 0.03–0.73; p=0.02) models; however, the association was only marginally significant when excluding HRT users (OR=0.21; 95% C.I.: 0.03–1.08; p=0.09).
In males, a significant association was observed in the fully adjusted model (Model 1, Table 4) when comparing men in the middle tertile to men in the lowest tertile of total body BMD (OR=0.23; 95% C.I.: 0.05–0.98; p=0.05); however, the p-value for trend was only 0.10. Although the risk of colorectal adenomas tended to decrease with increasing BMD in pelvic and spinal regions in males, we did not observe any statistically significant associations in minimally or fully adjusted models (Model 1, Table 4).
In addition, we found that postmenopausal women in the highest and middle tertiles combined compared to those in the lowest tertile of total body BMD had a lower risk of colorectal adenomas when including (OR=0.08; 95% C.I.: 0.01–0.83; p=0.02) and excluding (OR=0.04; 95% C.I.: 0.01–0.52; p=0.02) HRT users (not shown). An inverse association was also observed for spinal BMD when comparing the highest and middle tertiles combined compared to the lowest tertile when including (OR=0.08; 95% C.I.: 0.01–0.75; p=0.03) and excluding (OR=0.05; 95% C.I.: 0.01–0.75; p=0.04) HRT users. In the pelvic region, a significant inverse association was observed when including (OR=0.06; 95% C.I.: 0.01–0.62; p=0.01) but not when excluding (OR=0.09; 95% C.I.: 0.01–0.62; p=1.11) HRT users. Due to the small number of premenopausal women in our study, we could not examine potential associations between BMD and adenomas in premenopausal women.
Discussion
Our results show, for the first time, that total body BMD is inversely associated with colorectal adenomas. Stratification by gender revealed that the associations were generally more pronounced among females in our study population even when HRT users were excluded. Postmenopausal women in the highest and middle tertiles of total body BMD combined compared to those in the lowest tertile were also observed to have a decreased risk of adenomas. Associations between pelvic and spinal region BMD were similar to those observed for total body BMD.
Although our results are not directly comparable to previous findings in colorectal cancer, under the adenoma-to-carcinoma paradigm (2), our findings are indeed consistent with and complement prior studies reporting that higher bone mass levels decrease the risk of colorectal cancer (38) and colon cancer in postmenopausal women (39, 40). Differences in the specific bone mass regions examined and certain population attributes may also help explain variation in effect sizes across studies. For example, we found that the highest and middle tertiles of spinal (>1.209 g/cm2; 1.017–1.209 g/cm2) and total body (>1.280 g/cm2; 1.130–1.280 g/cm2) BMD in females, who were ~57 years old on average, reduced risk of colon adenomas by approximately 80% to 85%, respectively; and, this risk was further reduced (~90%) in postmenopausal women. Ganry et al. (39) observed that the highest compared to the lowest BMD tertile in the femoral neck (>0.819 vs. <0.708 g/cm2), trochanteric (>0.734 vs. <0.622 g/cm2) and Ward’s triangle (>0.655 vs. <0.546 g/cm2) regions, which are sites previously used to diagnose osteoporosis (43), decreased the risk of colon cancer by about 20% in older (mean age of ~71 years), postmenopausal women. When comparing the highest to the lowest BMD tertiles in the metacarpal cortical area (tertile values not reported), Zhang et al. (40) found a 60% decreased risk of colon cancer among older (mean age of ~62 years), postmenopausal women. Using data from the first National Health and Nutrition Examination Survey (NHANES I), which consisted of 2,818 men and 3,228 women who were, on average, ~50 years old, Nelson et al. (38) observed a significant trend for a decreasing risk of colorectal cancer with increasingly higher BMD quartiles in the left hand using Radiography Absorptiometry (RA). Although measures of bone density using RA have been shown to correlate with those from DXA scans, RA has several limitations including its calibration in arbitrary units (44), which inhibits the ability to direct compare bone density levels. Furthermore, the composition of bone varies by anatomical region, with the spine and pelvis having a higher concentration of metabolically active trabecular bone and the femur and legs having more cortical bone (45). Menopause also results in significant bone loss, which is believed to occur primarily through changes in trabecular and intracortical bone remodeling; and, postmenopausal women receiving estrogen therapy have higher BMD than those not receiving this treatment (45).
Moreover, prior studies reporting an association between BMD and colorectal cancer varied in terms of the other covariates included in their models – some authors used models adjusted only for age, gender, race and BMI while others adjusted for additional risk factors such as family history of colorectal cancer, smoking, alcohol, physical activity, NSAID use, etc. In our study, models including age, gender, race and BMI were similar to more fully adjusted models. Incorporating the additional risk factors improved the precision of effect estimates in some cases (e.g., in female total body BMD analyses) but the additional adjustment did not materially change the overall interpretation. This suggests that when taking into account only a few easily and reliably measurable attributes (age, gender, race, BMI), BMD may serve as an objective biomarker of colorectal adenomas.
The use of an objective, composite marker, such as BMD, as an additional mechanism to help identify patients who may benefit from earlier screening for colorectal adenomas (and cancer) may be particularly appealing for several reasons. First, calcium, vitamin D, estrogen and physical activity, which are known to affect BMD through complex biological and physiological interactions (24–26, 28, 46), have all been independently associated with colon adenomas and colon cancer, however, results have not been consistent across studies (9–12, 14, 15, 19–22). These inconsistencies may be due, in part, to problems with accurately quantifying lifetime or long-term exposure to dietary and lifestyle factors using self-report questionnaires and ‘snap-shots’ of serum biomarkers (35–37). Furthermore, although the exact underlying mechanisms are not currently known, calcium, vitamin D and estrogen have all been shown to modify levels of cell differentiation and apoptosis in the colon epithelium (16–18, 47). Moreover, methods used to screen for BMD are non-invasive and quite efficient, with a whole body DXA scan taking only approximately 10 to 15 minutes.
We note that two of the three previous studies only reported significant associations between BMD and colon cancer in women (39, 40); and, the one study that included men did not report results separately for males and females (38). In our study, we observed a dose-response between higher total body BMD tertiles and colorectal adenomas in women. The lack of a dose-response finding among men in our study may be due, in part, to the smaller amount of variation in BMD among men compared to women (total body BMD standard deviation (s.d.): 0.124 vs. 0.146 g/cm2; pelvis BMD s.d.: 0.145 vs. 0.206 g/cm2; spine BMD s.d.: 0.146 vs. 0.242 g/cm2). Nevertheless, given our current and the previous findings, one could speculate that BMD may have greater substantive value in females, particularly of middle-to-older age, which, arguably, could be undergoing BMD screening for osteoporosis and fracture risk anyway. Interestingly, it has been reported that women have a substantially lower adenoma detection rate compared to men, which, as Roy and Bianchi (48) suggest, may indicate a decreased efficacy for colonoscopy screening to prevent colorectal cancer in women compared to men. Although Roy and Bianchi (48) acknowledge that their appraisal of the literature is quite controversial, they contend that the current evidence is strong enough to provide “a plausible reason to investigate other biomarkers that may help protect against colon cancer in women”. Perhaps, BMD, although non-specific, could serve as such a marker.
In our study population, female cases, on average, reported an earlier age at menopause compared to controls (Table 1). Although we confirmed menopause using serum FSH and E2, the age at menopause was obtained by self-report and, therefore, may be biased. However, the bias is most likely non-differential between cases and controls since the information was obtained prospectively (i.e., prior to the colonoscopy outcome). It is also possible that the lower mean age at menopause we observed in cases compared to controls may reflect some other underlying unmeasured systemic condition.
There are several limitations of our study. The most important limitation relates to our small sample size, which led to small cell counts and prohibited a separate evaluation of premenopausal women. In addition, a larger sample would help to more definitively determine if there is (or is not) a dose-response association between higher BMD and colorectal adenomas in men. Although we controlled for factors that may affect estrogen levels in our analysis (e.g., BMI, post-menopausal status, years of HRT use), the possibility of residual confounding by an unmeasured estrogen measure cannot be ruled out. In particular, we did not have information to evaluate how the specific type or classification of HRT drug (e.g. estrogen alone vs. estrogen and progesterone) might have affected results. However, excluding HRT users from the analyses did not materially change the results. Furthermore, we were not able to quantify BMD at the specific sites previously reported in colon cancer (femoral neck, Ward’s triangle, trochanter, metacarpal cortical area) using our total body DXA scan; however, we did use surrogate areas (pelvis, spine) to attempt to evaluate regions of similar bone composition.
In conclusion, we observed, for the first time, an inverse association between higher levels of BMD and colorectal adenomas, which stratification by gender revealed was more pronounced in females even when HRT users were excluded. Although additional larger, prospective studies are needed, our results suggest that BMD may be a biomarker for colorectal cancer precursor lesions.
Acknowledgments
This work was supported, in part, by National Institutes of Health (NIH) National Cancer Institute (NCI) [K07 CA129162 to N.L.N.; U54 CA116867, R01 CA136726 to L.L.]
Reference List
- 1.Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, 2009. CA Cancer J Clin. 2009;59:225–49. doi: 10.3322/caac.20006. [DOI] [PubMed] [Google Scholar]
- 2.Hill MJ, Morson BC, Bussey HJ. Aetiology of adenoma--carcinoma sequence in large bowel. Lancet. 1978;1:245–7. doi: 10.1016/s0140-6736(78)90487-7. [DOI] [PubMed] [Google Scholar]
- 3.Winawer SJ, Zauber AG, O’Brien MJ, Ho MN, Gottlieb L, Sternberg SS, Waye JD, Bond J, Shapiro M, Stewart ET, Panish J, Ackroyd F, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med. 1993;329:1977–81. doi: 10.1056/NEJM199312303292701. [DOI] [PubMed] [Google Scholar]
- 4.Cram P, Fendrick AM, Inadomi J, Cowen ME, Carpenter D, Vijan S. The impact of a celebrity promotional campaign on the use of colon cancer screening: the Katie Couric effect. Arch Intern Med. 2003;163:1601–5. doi: 10.1001/archinte.163.13.1601. [DOI] [PubMed] [Google Scholar]
- 5.Gross CP, Andersen MS, Krumholz HZ, McAvay GJ, Proctor D, Tinetti ME. Relation between Medicare screening reimbursement and stage at diagnosis for older patients with colon cancer. JAMA. 2006;296:2815–22. doi: 10.1001/jama.296.23.2815. [DOI] [PubMed] [Google Scholar]
- 6.Giacosa A, Frascio F, Munizzi F. Epidemiology of colorectal polyps. Tech Coloproctol. 2004;8:s243–s247. doi: 10.1007/s10151-004-0169-y. [DOI] [PubMed] [Google Scholar]
- 7.Gupta S. Colorectal polyps: the scope and management of the problem. Am J Med Sci. 2008;336:407–17. doi: 10.1097/MAJ.0b013e31817d2402. [DOI] [PubMed] [Google Scholar]
- 8.Young J, Jass JR. The case for a genetic predisposition to serrated neoplasia in the colorectum: hypothesis and review of the literature. Cancer Epidemiol Biomarkers Prev. 2006;15:1778–84. doi: 10.1158/1055-9965.EPI-06-0164. [DOI] [PubMed] [Google Scholar]
- 9.Cho E, Smith-Warner SA, Spiegelman D, Beeson W, van den Brandt PA, Colditz GA. Dairy foods, calcium, and colorectal cancer: a pooled analysis of 10 cohort studies. J Natl Cancer Inst. 2004;96:1015–22. doi: 10.1093/jnci/djh185. [DOI] [PubMed] [Google Scholar]
- 10.Huncharek M, Muscat J, Kupelnick B. Colorectal cancer risk and dietary intake of calcium, vitamin D, and dairy products: a meta-analysis of 26,335 cases from 60 observational studies. Nutr Cancer. 2009;61:47–69. doi: 10.1080/01635580802395733. [DOI] [PubMed] [Google Scholar]
- 11.Yin L, Grandi N, Raum E, Haug U, Arndt V, Brenner H. Meta-analysis: longitudinal studies of serum vitamin D and colorectal cancer risk. Ailment Pharmacol Ther. 2009;30:113–25. doi: 10.1111/j.1365-2036.2009.04022.x. [DOI] [PubMed] [Google Scholar]
- 12.Baron JA, Beach MA, Mandel JS, van Stolk RU, Haile RW, Sandler RS, Rothstein R, Summers RW, Snover DC, Beck GJ, Fankyl H, Pearson L, et al. Calcium supplements and colorectal adenomas. Polyp Prevention Study Grou p. Ann N Y Acad Sci. 1999;889:138–45. doi: 10.1111/j.1749-6632.1999.tb08731.x. [DOI] [PubMed] [Google Scholar]
- 13.Wallace K, Baron JA, Cole BF, Sandler RS, Karagas MR, Beach MA, Haile RW, Burke CA, Pearson LH, Mandel JS, Rothstein R, Snover DC. Effect of calcium supplementation on the risk of large bowel polyps. J Natl Cancer Inst. 2004;96:921–5. doi: 10.1093/jnci/djh165. [DOI] [PubMed] [Google Scholar]
- 14.Wei MY, Garland CF, Gorham ED, Mohr SB, Giavonucci E. Vitamin D and prevention of colorectal adenoma: a meta-analysis. Cancer Epidemiol Biomarkers Prev. 2008;17:2958–69. doi: 10.1158/1055-9965.EPI-08-0402. [DOI] [PubMed] [Google Scholar]
- 15.Grau MV, Baron JA, Sandler RS, Haile RW, Beach ML, Church TR, Heber D. Vitamin D, calcium supplementation, and colorectal adenomas: results of a randomized trial. J Natl Cancer Inst. 2003;95:1765–71. doi: 10.1093/jnci/djg110. [DOI] [PubMed] [Google Scholar]
- 16.Penman ID, Liang QL, Bode J, Eastwood MA, Arends MJ. Dietary calcium supplementation increases apoptosis in the distal murine colonic epithelium. J Clin Pathol. 2000;53:302–7. doi: 10.1136/jcp.53.4.302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fedirko V, Bostick RM, Flanders WD, Long Q, Shaukat A, Rutherford RE. Effects of Vitamin D and Calcium Supplementation on Markers of Apoptosis in Normal Colon Mucosa: A Randomized, Double-Blind, Placebo-Controlled Clinical Trial. Cancer Prev Res (Phila Pa) 2009;2:213–23. doi: 10.1158/1940-6207.CAPR-08-0157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Harris DM, Go VL. Vitamin D and colon carcinogenesis. J Nutr. 2004;134:3463S–71S. doi: 10.1093/jn/134.12.3463S. [DOI] [PubMed] [Google Scholar]
- 19.Grodstein F, Newcomb PA, Stampfer MJ. Postmenopausal hormone therapy and the risk of colorectal cancer: a review and meta-analysis. Am J Med. 1999;106:574–82. doi: 10.1016/s0002-9343(99)00063-7. [DOI] [PubMed] [Google Scholar]
- 20.Hébert-Croteau N. A meta-analysis of hormone replacement therapy and colon cancer in women. Cancer Epidemiol Biomarkers Prev. 1998;7:653–9. [PubMed] [Google Scholar]
- 21.Woodson K, Lanza E, Tangrea JA, Alberts PS, Slattery M, Pinsky J, Caan B, Paskett E, Iber F, Kikendall JW, Lance P, Shike M, et al. Hormone replacement therapy and colorectal adenoma recurrence among women in the Polyp Prevention Trial. J Natl Cancer Inst. 2001;93:1799–805. doi: 10.1093/jnci/93.23.1799. [DOI] [PubMed] [Google Scholar]
- 22.Grodstein F, Martinex ME, Platz E, Giovannucci E, Colditz GA, Kautzky M, Fuchs C, Sampfer MJ. Postmenopausal hormone use and risk for colorectal cancer and adenoma. Ann Intern Med. 1998;128:705–12. doi: 10.7326/0003-4819-128-9-199805010-00001. [DOI] [PubMed] [Google Scholar]
- 23.Mirza FS, Prestwood KM. Bone health and aging: implications for menopause. Endocrin Metab Clin North Am. 2004;33:741–59. doi: 10.1016/j.ecl.2004.07.001. [DOI] [PubMed] [Google Scholar]
- 24.Bauer DC, Browner WS, Cauley JA, Orwoll ES, Scott JC, Black DM, Tao JL, Cummings SR. Factors associated with appendicular bone mass in older women. The Study of Osteoporotic Fractures Research Group. Ann Intern Med. 1993;118:657–65. doi: 10.7326/0003-4819-118-9-199305010-00001. [DOI] [PubMed] [Google Scholar]
- 25.Cummings SR, Browner WS, Grady D, Ettinger B. Should prescription of postmenopausal hormone therapy be based on the results of bone densitometry? Ann Intern Med. 1990;113:565–7. doi: 10.7326/0003-4819-113-8-565. [DOI] [PubMed] [Google Scholar]
- 26.Zhang Y, Kiel DP, Kreger BE, Cupples LA, Ellison RC, Dorgan JF, Schatzkin A, Levy D, Felson JT. Bone mass and the risk of breast cancer among postmenopausal women. N Engl J Med. 1997;336:611–7. doi: 10.1056/NEJM199702273360903. [DOI] [PubMed] [Google Scholar]
- 27.Branca F. Physical activity, diet and skeletal health. Public Health Nutr. 1999;2(3A):391–6. doi: 10.1017/s1368980099000531. [DOI] [PubMed] [Google Scholar]
- 28.Brot C, Jorgensen N, Madsen OR, Jensen LB, Sorensen OH. Relationships between bone mineral density, serum vitamin D metabolites and calcium:phosphorus intake in healthy perimenopausal women. J Intern Med. 1999;245:509–16. doi: 10.1046/j.1365-2796.1999.00474.x. [DOI] [PubMed] [Google Scholar]
- 29.Papaioannou A, Kennedy CC, Cranney A, Hawker G, Brown JP, Kaiser SM, Leslie WD, O’Brien CJ, Sawka AM, Khan A, Siminoski K, Tarulli G, et al. Risk factors for low BMD in healthy men age 50 years or older: a systematic review. Osteoporos Int. 2009;20:507–18. doi: 10.1007/s00198-008-0720-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Waugh EJ, Lam MA, Hawker GA, McGowan J, Papaioannou A, Cheung AM, Hodsman AB, Leslie WD, Siminoski K, Jamal SA. Perimenopause BMD Guidelines Subcommittee of Osteoporosis Canada. Risk factors for low bone mass in healthy 40–60 year old women: A systematic review of the literature. Osteoporos Int. 2008;20:1–21. doi: 10.1007/s00198-008-0643-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Harriss DJ, Atkinson G, Batterham A, George K, Cable NT, Reilly T, Haboubi N, Renehan AG, Colorectal Cancer, Lifestyle, Exercise And Research Group Lifestyle factors and colorectal cancer risk (2): a systematic review and meta-analysis of associations with leisure-time physical activity. Colorectal Dis. 2009;11:689–701. doi: 10.1111/j.1463-1318.2009.01767.x. [DOI] [PubMed] [Google Scholar]
- 32.Rosenberg L, Boggs D, Wise LA, Palmer JR, Roltsch MH, Makambi KH, Adams-Campbell LL. A follow-up study of physical activity and incidence of colorectal polyps in African-American women. Cancer Epidemiol Biomarkers Prev. 2006;15:1438–42. doi: 10.1158/1055-9965.EPI-06-0079. [DOI] [PubMed] [Google Scholar]
- 33.Wallace K, Baron JA, Karagas MR, Cole BF, Byers T, Beach MA, Pearson LH, Burke CA, Silverman WB, Sandker RS. The association of physical activity and body mass index with the risk of large bowel polyps. Cancer Epidemiol Biomarkers Prev. 2005;14:2082–6. doi: 10.1158/1055-9965.EPI-04-0757. [DOI] [PubMed] [Google Scholar]
- 34.Wolin KY, Yan Y, Colditz GA, Lee IM. Physical activity and colon cancer prevention: a meta-analysis. Br J Cancer. 2009;100:611–6. doi: 10.1038/sj.bjc.6604917. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Millen AE, Bodnar LM. Vitamin D assessment in population-based studies: a review of the issues. Am J Clin Nutr. 2008;87:1102S–5S. doi: 10.1093/ajcn/87.4.1102S. [DOI] [PubMed] [Google Scholar]
- 36.Miller AB, Linsiesen J. Achievements and future of nutritional cancer epidemiology. Int J Cancer. 2009 Oct 28; doi: 10.1002/ijc.25006. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 37.Schatzkin A, Subar AF, Moore S, Park Y, Potischman N, Thompson FE, Leitzmann M, Hollenbeck A, Morrissey JG, Kipnis V. Observational epidemiologic studies of nutrition and cancer: the next generation (with better observation) Cancer Epidemiol Biomarkers Prev. 2009;18:1026–32. doi: 10.1158/1055-9965.EPI-08-1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nelson RL, Turyk M, Kim J, Persky V. Bone mineral density and the subsequent risk of cancer in the NHANES I follow-up cohort. BMC Cancer. 2002;2:22–31. doi: 10.1186/1471-2407-2-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ganry O, Lapotre-Ledoux B, Fardellone P, Dubreuil A. Bone mass density, subsequent risk of colon cancer and survival in postmenopausal women. Eur J Epidemiol. 2008;23:467–73. doi: 10.1007/s10654-008-9256-0. [DOI] [PubMed] [Google Scholar]
- 40.Zhang Y, Felson DT, Ellison RC, Kreger BE, Schatzkin A, Dorgan JF, Cupples LA, Levy D, Kiel DP. Bone mass and the risk of colon cancer among postmenopausal women: the Framingham study. Am J Epidemiol. 2001;153:31–7. doi: 10.1093/aje/153.1.31. [DOI] [PubMed] [Google Scholar]
- 41.Martinez MEMJR, Graver E, Whitacre RC, Woolf K, Ritenbaugh C, Alberts DC. Reliability and validity of a self-administered food frequency questionnaire in a chemoprevention trial of adenoma recurrence. Cancer Epidemiol Biomarkers Prev. 1999;8:941–6. [PubMed] [Google Scholar]
- 42.Staten LK, Taren DL, Howell WH, Tobar M, Poehlman ET, Hill A, Reid PM, Ritenbaugh C. Validation of the Arizona Activity Frequency Questionnaire using doubly labeled water. Med Sci Sports Exerc. 2001;33:1959–67. doi: 10.1097/00005768-200111000-00024. [DOI] [PubMed] [Google Scholar]
- 43.Lewiecki EM, Gordon CM, Baim S, Leonard MB, Bishop NJ, Bianchi ML, Kalkwarf HJ, Langman L, Plotkin H, Rauch F, Zemel BS, Binkley N, et al. International Society for Clinical Densitometry 2007 Adult and Pediatric Official Positions. Bone. 2008;43:1115–21. doi: 10.1016/j.bone.2008.08.106. [DOI] [PubMed] [Google Scholar]
- 44.Gulam M, Thorton MM, Hodsman AB, Holdsworth DW. Bone mineral measurement of phalanges: comparison of radiographic absorptiometry and area dual X-ray absorptiometry. Radiology. 2000;216:586–91. doi: 10.1148/radiology.216.2.r00au10586. [DOI] [PubMed] [Google Scholar]
- 45.Allen MR. Surface-specific bone formation effects of osteoporosis: pharmacological treatments. Clinic Rev Bone Miner Metab. 2008;6:69. [Google Scholar]
- 46.Cummings SR, Ettinger B, Delmas PD, Kenemans P, Stathopoulos V, Verweij P, Mol-Arts M, Kloosterboer L, Mosca M, Christiansen C, Bilezikian J, Kerzberg EM, et al. The effects of tibolone in older postmenopausal women. N Engl J Med. 2008;359:697–708. doi: 10.1056/NEJMoa0800743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Wada-Hiraike O, Warner M, Gustafsson JA. New developments in oestrogen signalling in colonic epithelium. Biochem Soc Trans. 2006;34:1114–6. doi: 10.1042/BST0341114. [DOI] [PubMed] [Google Scholar]
- 48.Roy HK, Bianchi LK. Differences in colon adenomas and carcinomas among women and men: potential clinical implications. JAMA. 2009;302:1696–7. doi: 10.1001/jama.2009.1499. [DOI] [PMC free article] [PubMed] [Google Scholar]