Cardiovascular disease (CVD) and osteoporosis share some common risk factors such as old age, smoking, alcoholic drinking, hypertension, diabetes mellitus, and hyperlipidemia. Although previous studies have investigated the association of bone mineral density (BMD) with CVD, the results were conflicting. There are limited studies on the association of BMD loss rate with CVD. We therefore conducted a 5-year prospective study to examine the relation among BMD, bone loss, and risk of CVD in a Chinese cohort. Of 9,657 community residents 30 to 75 years old, 6,092 were enrolled in the study and followed annually for 5 years. At baseline demographic data, BMD, smoking and drinking statuses, medical history, and blood samples were collected. Cox proportional hazards analysis was used to evaluate the association of BMD and incidence of CVD. Over the 5-year follow-up period, CVD developed in 118 subjects. Baseline BMD, bone loss rate, current smoking, daily alcoholic ingestion, and higher osteoprotegerin and leptin levels were independently associated with increased risk of CVD, whereas higher baseline adiponectin level was associated with decreased risk of CVD in women and men. In conclusion, uncovering the relation linking osteoporosis and CVD is important for understanding the pathogenesis of these 2 common disorders.
Some but not all previous studies have suggested that low bone mineral density (BMD) is related to an increased risk of cardiovascular disease (CVD). There is no study on the association between bone loss rate and risk of CVD. Furthermore, the relation between these 2 common disorders is not fully elucidated. The purpose of this study was to examine the association of BMD and bone loss rate with risk of CVD in a Chinese population.
Methods
Subjects were sampled from 5 randomly selected communities in the city of Chongqing. This study was approved by the institutional review board of Daping Hospital and all subjects provided informed consent.
Baseline screening of the present study was performed from January 2003 through August 2003. In the selected communities, there were 9,657 subjects 30 to 75 years old. Subjects were interviewed by holding meetings in community centers. For subjects who were absent at the meeting time or were not able to attend because of physical disabilities, our staff went to their homes to perform the interview.
Of 7,221 subjects who were screened at baseline, 2,436 subjects were not available at the time of screening (n = 1,339) or declined to participate (n = 1,097) and thus excluded from the present study. Subjects who complained of signs or symptoms of CVD (angina or chest pain) were excluded from the study. Our study was comprised of 7,115 subjects who were free of CVD at baseline. CVD included angina, coronary heart disease, myocardial infarction, angioplasty, coronary artery bypass surgery, congestive heart failure, or hemorrhagic stroke.
The following data were collected by formal questionnaire and cardiologists. Procedures were administered by trained interviewers consisting of experienced cardiologists and senior nurses. The Cohen kappa statistic was used to measure reliability across interviewers. Agreement on data collection was found to be excellent based on the same sample of subjects (kappa 0.93).
Demographic data included age, gender, and educational level (lower educational level, i.e., illiterate or primary school, and higher educational level, i.e., higher than secondary school).
Dual energy x-ray absorptiometry (Prodigy Fan Beam Densitometer, Lunar Corp., GE Medical Systems, Madison, Wisconsin) was used to determine BMD. BMD of the femoral neck bone was used as a measurement of global bone health. BMD was measured at baseline and 3 years after enrollment. Annual percent change (100 [BMD 1 − BMD 2]/BMD 1 × length of follow-up in years) and absolute annual change in BMD during the 3-year follow-up period were calculated. BMD and percent change of BMD were categorized into quartiles.
Smoking status was classified as (1) former smokers who had quit smoking for ≥6 months, (2) current smokers, or (3) those who never smoked. Drinking status was classified as daily drinking, weekly drinking, monthly drinking, or occasional drinking as defined previously.
Subjects’ medical history was collected from medical records. Data included CVD, stroke, peripheral artery disease, cardiomyopathy, surgery, or trauma within the previous month, malignant disease, febrile conditions, chronic obstructive pulmonary disease, chronic hepatitis, chronic renal insufficiency, hypertension, diabetes mellitus, and hypercholesterolemia.
Blood pressure measurements and electrocardiography were performed on site, and fasting blood samples were collected to measure glucose, total cholesterol, osteoprotegerin (OPG), receptor activator of nuclear factor-κB ligand (RANKL), adiponectin, and leptin. Body mass index was measured. Subjects with abnormalities implying potential diseases that were not previously diagnosed were introduced to Daping Hospital in Chongqing for further investigation. Diagnosis of diseases including coronary heart diseases, hypertension, diabetes mellitus, hypercholesterolemia, obesity, atrial fibrillation, chronic renal insufficiency, and chronic hepatitis were based on codes from the International Classification of Diseases, Ninth Revision .
CVD events were defined as angina, myocardial infarction, angioplasty, coronary artery bypass surgery, congestive heart failure, and ischemic or hemorrhagic stroke during the follow-up period. Subjects were followed annually for 5 years from September 2003 through August 2008.
Baseline variables were analyzed using t test for independent normally distributed continuous data, chi-square test for categorical data, and Mann–Whitney U test for ordinal and categorical variables that were not normally distributed. Cox proportional hazards models were used to assess associations between BMD or bone loss rate and risk of CVD development with adjustment for potential confounders. Variables taken into account as potential confounders were continuous variables such age, gender, and education. Statistical analyses were performed using SPSS 15.0 for Windows (SPSS, Inc., Chicago, Illinois).
Results
At baseline 7,115 subjects without CVD were enrolled in the present study. Five years later, 118 developed CVD, 68 (0.96%) died, 802 (11.3%) declined, and 153 (2.2%) moved away during follow-up. Thus 6,092 subjects (85.6%) completed the follow-up.
Compared to subjects who did not develop CVD, those who did were older, more frequently men, current smokers and consumed alcoholic drinks daily, had lower BMD and plasma adiponectin levels, and higher plasma leptin and OPG levels ( Table 1 ).
| Variable | Did Not Develop CVD (n = 5,974) | Developed CVD (n = 118) | p Value |
|---|---|---|---|
| Age (years), mean ± SD | 55.8 ± 10.4 | 62.6 ± 8.8 | <0.001 |
| Men | 2,419 (41%) | 73 (62%) | <0.001 |
| Lower education level (≤6 years) | 1,229 (21%) | 31 (26%) | 0.13 |
| Bone mineral density (g/cm 2 ), mean ± SD | 0.94 ± 0.12 | 0.81 ± 0.10 | <0.001 |
| Body mass index (kg/m 2 ), mean ± SD | 24.4 ± 3.6 | 24.7 ± 2.4 | 0.343 |
| Osteoprotegerin (ng/ml), mean ± SD | 1.16 ± 0.33 | 1.29 ± 0.51 | <0.001 |
| Receptor activator of nuclear factor-κB ligand (ng/ml), mean ± SD | 1.54 ± 0.44 | 1.59 ± 0.71 | 0.20 |
| Leptin (ng/ml), mean ± SD | 16.8 ± 6.3 | 22.1 ± 11.1 | <0.001 |
| Adiponectin (μg/ml), mean ± SD | 16.8 ± 6.5 | 12.8 ± 6.1 | <0.001 |
| Never smoker | 2,313 (39%) | 29 (25%) | |
| Former smoker | 2,371 (40%) | 44 (37%) | |
| Current smoker | 1,290 (22%) | 45 (38%) | <0.001 |
| Occasional alcoholic drink | 1,754 (29%) | 28 (24%) | |
| Daily alcoholic drink | 1,198 (20%) | 35 (30%) | 0.037 |
| Weekly alcoholic drink | 1,216 (20%) | 27 (23%) | |
| Monthly alcoholic drink | 1,806 (30%) | 28 (24%) | |
| Hypertension ⁎ | 2,538 (43%) | 73 (62%) | <0.001 |
| Diabetes mellitus | 626 (11%) | 27 (23%) | <0.001 |
| Hyperlipidemia † | 1,859 (31%) | 44 (41%) | 0.027 |
⁎ Defined as systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or use of antihypertensive mediation.
† Defined as total cholesterol ≥200 mg/dl, low-density lipoprotein cholesterol ≥130 mg/dl, or use of lipid-lowering medication.
Unadjusted hazard risks (HRs) of CVD for each component were 3.70 for BMD (95% confidence interval [CI] 2.04 to 6.70, quartiles 1 and 4), 3.14 for OPG (95% CI 1.80 to 5.47), 2.44 for current smoker (95% CI 1.52 to 3.91), 2.40 for diabetes (95% CI 1.56 to 3.68), 2.05 for hypertension (95% CI 1.41 to 2.97), 1.82 for daily drinking (95% CI 1.11 to 2.99), 1.71 for hyperlipidemia (95% CI 1.19 to 2.45), 1,09 for leptin (95% CI 1.07 to 1.11), and 0.87 for adiponectin (95% CI 0.84 to 0.91). These associations remained unchanged after adjustment for age and gender ( Table 2 ).
| Variable | Unadjusted HR for CVD (95% CI) | Adjusted HR for CVD (95% CI) ⁎ |
|---|---|---|
| Bone mineral density † | ||
| Quartile 1 | 3.70 (2.04–6.70) | 3.51 (1.94–6.35) |
| Quartile 2 | 2.36 (1.26–4.40) | 2.32 (1.24–4.33) |
| Quartile 3 | 1.56 (0.80–3.04) | 1.51 (0.77–2.96) |
| Quartile 4 | 1 | 1 |
| Osteoprotegerin | 3.14 (1.80–5.47) | 3.06 (1.76–5.34) |
| Receptor activator of nuclear factor-κB ligand | 1.28 (0.88–1.87) | 1.40 (0.91–2.17) |
| Leptin | 1.09 (1.07–1.11) | 1.08 (1.06–1.10) |
| Adiponectin | 0.87 (0.84–0.91) | 0.88 (0.84–0.91) |
| Cigarette smoking | ||
| Never | 1 | 1 |
| Former | 1.46 (0.92–2.31) | 1.43 (0.90–2.26) |
| Current | 2.44 (1.52–3.91) | 2.47 (1.54–3.95) |
| Alcohol drinking | ||
| Occasional | 1 | 1 |
| Daily | 1.82 (1.11–2.99) | 1.78 (1.08–2.93) |
| Weekly | 1.38 (0.81–2.33) | 1.44 (0.85–2.44) |
| Monthly | 0.97 (0.57–1.63) | 0.92 (0.54–1.55) |
| Hypertension | 2.05 (1.41–2.97) | 2.02 (1.39–2.93) |
| Diabetes mellitus | 2.40 (1.56–3.68) | 2.31 (1.50–3.56) |
| Hyperlipidemia | 1.71 (1.19–2.45) | 1.67 (1.16–2.41) |
⁎ Adjusted for age and gender.
† Quartile 1, <0.762; quartile 2, 0.763 to 0.886; quartile 3, 0.887 to 1.013; quartile 4, >1.014.
We further examined the association between BMD and CVD in women or men. Compared to BMD in the highest quartile (quartile 4), women with BMD values in the lowest quartile (quartile 1) had an HR of 3.24 (95% CI 1.29 to 8.15). The association remained statistically significant after adjustment (HR 3.46, 95% CI 1.36 to 8.77).
Similarly, men with BMD values in the lowest quartile (quartile 1) had an unadjusted HR of 3.98 (95% CI 1.83 to 8.66) and adjusted HR of 3.65 (95% CI 1.65 to 8.07) compared to men with BMD values in the highest quartile (quartile 4; Table 3 ).
| BMD Quartile ⁎ | Women | Men | ||
|---|---|---|---|---|
| Unadjusted HR (95% CI) | Adjusted HR (95% CI) † | Unadjusted HR (95% CI) | Adjusted HR (95% CI) † | |
| 1 | 3.24 (1.29–8.15) | 3.46 (1.36–8.77) | 3.98 (1.83–8.66) | 3.65 (1.65–8.07) |
| 2 | 2.05 (0.78–5.39) | 1.69 (0.63–4.53) | 2.71 (1.19–6.15) | 2.93 (1.27–6.75) |
| 3 | 1.25 (0.44–3.62) | 1.24 (0.43–3.59) | 1.87 (0.79–4.47) | 1.55 (0.64–3.75) |
| 4 | 1 | 1 | 1 | 1 |
⁎ Quartile 1, <0.762; quartile 2, 0.763 to 0.886; quartile 3, 0.887 to 1.013; quartile 4, >1.014.
† Adjusted for age, smoking, drinking, hypertension, diabetes, hyperlipidemia, osteoprotegerin, receptor activator of nuclear factor-κB ligand, leptin, and adiponectin.
Compared to subjects whose bone loss rates were in the lowest quartile (quartile 1), subjects whose bone loss rates were in the highest quartile (quartile 4) had an HR of 3.14 (95% CI 1.81 to 5.45). The association remained statistically significant after adjustment (HR 2.90, 95% CI 1.65 to 5.08). In further analyses for women and men, these associations remained the same, with higher HRs in men than in women. These results suggest that bone loss rate is also associated with increased risk of CVD ( Table 4 ).
