Factors affecting bone mineral density in men
- First Online:
- Cite this article as:
- Atalar, E., Aydin, G., Keles, I. et al. Rheumatol Int (2009) 29: 1025. doi:10.1007/s00296-008-0768-4
- 102 Views
In this study, in 131 men aged 20–75 years, we investigated correlations between bone mineral density (BMD) in the lumbar spine and femoral neck and endogenous factors (age, body mass index) as well as exogenous factors (calcium intake, physical activity, smoking, caffeine, socioeconomic and educational levels). The age had a negative effect on femoral neck BMD in patients overall, and on both lumbar spine and femoral neck BMD in patients under 50. Physical activity has effects on femoral neck BMD in men above 50. Lumbar vertebral BMD negatively correlated with smoking in patients overall, and this correlation persisted when patients aged 50 and older were analyzed separately. Femoral neck BMD was positively correlated with body mass index in men aged 50 and older. Given the variety of findings in the research literature regarding risk factors for low BMD, we suggest that genetic and geographic factors should be considered.
KeywordsBone mineral densityMaleBone lossAge-related
In societies living in different geographical regions, bone mineral density (BMD) can show differences due to ethnic and genetic factors, as well as traditional dietary habits and physical activity . Calcium-rich diets, physical activity, high socioeconomic levels as well as early menarche and short times since menopause are thought to have positive effects on BMD [2–5]. Smoking and alcohol, excessive consumption of caffeine-rich foods, wearing clothes that prevent adequate sun exposure, and low body mass index (BMI) are thought to have negative effects on BMD [6–10].
Many studies have investigated the question of when peak bone mass is reached, but results have varied. Up to the age of 18, at least 90% of the peak bone mass is reached . However, in some people, the bone mass continues to increase by 5–12% into the third decade . Depending on gender and the part of the skeleton investigated, the time of reaching the peak bone mass may vary . Once it is reached, the peak bone mass generally remains constant up to the fifth decade, when progressive bone loss begins .
The purpose of this study was to examine, in males aged 20–75 years, factors which might affect BMD of the lumbar spine and femoral neck, including endogenous (age, BMI) and exogenous factors (calcium intake, physical activity, smoking, caffeine intake, socioeconomic and educational levels).
This study included 131 male patients referred to our center for any musculoskeletal complaint. Mean age of the patients was 47.43 years (range 20–75). The study was approved by our institution’s local ethics committee, and written informed consent was obtained from the patients.
Patients with current or past systemic diseases which could affect bone metabolism (diseases of the liver, kidney, thyroid, parathyroid, adrenal glands or reproductive organs; diabetes, neoplasia, gastrectomy) were excluded from the study. Patients were also excluded if they were currently taking, or had taken in the past, medications known to affect bone metabolism (steroids, thiazide diuretics, hormones, antiepileptic drugs, heparin, lithium, bisphosphonates, Vitamin D, etc.). Thus the patients included in the study were those with spine and joint pains due to mechanical and/or degenerative causes, e.g. rotator cuff injury, osteoarthritis, or mechanical low back pain.
BMD of lumbar vertebrae L2–L4 and of the left femoral neck were measured anteroposteriorly via dual-energy X-ray absorptiometry (DEXA, Norland XR-36 Corporation, Wisconsin, USA). BMI was calculated in kg/m2.
Through the use of a standardized questionnaire, the following information was obtained for each patient: calcium intake (mg/week), physical activities (metabolic equivalent tasks, METs/week ), caffeine intake (mg/week), alcohol intake (g/week), smoking status, educational level (illiterate, primary school, middle school, high school, college or university) and socioeconomic level (low, middle, high).
Calculation of calcium intake was based on calcium-rich foodstuffs consumed most in the region. Weekly consumption of calcium-rich foodstuffs in the last year was asked about and calcium intake was calculated in mg/week.
Physical activities and equivalents
MET values for activities
Sitting at home or at work
Mixed standing/walking at home or at work
Walking (very brisk)
Table tennis (ping pong)
Lawn and garden
8 MET × 0.002 × number of flights of stairs × 7
Patients were divided into four groups with respect to smoking status: non-smokers, former smokers, those who currently smoked less than one pack per week, and those who currently smoked more than this.
In measuring socioeconomic levels, three groups were defined, according to the per capita monthly income of the members of the patients’ families: low (US $0–199), middle (US $200–799) and high (above US $800).
In terms of educational level attained, the patients were divided into the following groups: illiterate, primary school, middle school, high school, college or university.
For correlation analyses, age categories were defined in 10-year intervals: 20–29, 30–39, 40–49, 50–59, 60–69, 70–79. Two larger age groups were also defined, as under 50 years, and 50 and above. Lumbar spine and femoral neck BMD were included in the study as dependent variables. Relations of these two variables with the others (age, BMI, calcium intake, physical activity, smoking, caffeine, alcohol, socioeconomic level, educational level) were tested using Spearman’s ρ correlation in terms of variables which were not normally distributed or which were ordinal variables, while the Pearson’s r statistic was used for normally distributed variables.
Among the participants in the study, a history of alcohol use was very rare, and for this reason this variable was not analyzed. Multivariate linear regression analysis was made to determine the extent to which the variations in bone density were explained by these variables. The most suitable regression model was reached by means of a stepwise regression method.
SPSS 11 software was used for the statistical analysis. For all the test statistics used in the study, significance was defined as P < 0.05.
Distribution of findings across the under-50 and 50-and-over age groups
All patients in the study (n = 131)
Men under 50 (n = 71)
Men aged 50 or older (n = 60)
Physical activity (METs/week)
Vertebral (L2–L4) BMD (g/cm2)
Femoral neck BMD (g/cm2)
Mean bone mineral densities in the 10-year interval age groups
Age groups (years)
No. of patients
Femoral neck BMD
Vertebral L2–L4 BMD
Factors related to lumbar spine and femoral neck BMD
Among the patients overall, lumbar BMD was negatively correlated with smoking (r = −0.262, P < 0.05). Femoral neck BMD was positively correlated with income and education levels (r = 0.244, P < 0.05; r = 0.265, P < 0.05) and was negatively correlated with age and smoking (r = −0.495, P < 0.05; r = −0.191, P < 0.05). Linear regression analysis showed that lumbar spine BMD was affected by smoking (R2 = 0.075), while femoral neck BMD was affected by age and BMI (R2 = 0.288).
Femoral neck BMD in patients aged 50 or above was significantly lower than that in patients under age 50 (P < 0.05). However, no significant difference was found between these two age groups with regard to lumbar spine BMD (P > 0.05).
In the patients under age 50, lumbar spine BMD was negatively correlated with age and caffeine intake (r = −0.268, P < 0.05; r = −0.244, P < 0.05). In this same age group, femoral neck BMD was likewise negatively correlated with age and caffeine intake (r = −0.349, P < 0.05; r = −0.244, P < 0.05). Linear regression analysis showed that lumbar spine and femoral neck BMD were both affected by age (R2 = 0.060, R2 = 0.120).
In men 50 or above, lumbar spine BMD was negatively correlated with smoking (r = −0.463, P < 0.05). In this same age group, femoral neck BMD was positively correlated with BMI (r = 0.449, P < 0.05) and negatively with age (P < 0.05, r = −0.283). Linear regression analysis showed that lumbar spine BMD was affected by smoking (R2 = 0.189) and femoral neck BMD was affected by BMI and physical activity (R2 = 0.299).
It is known that during growth many factors affect the increase of bone mass. These factors are primarily genetic factors, gender, dietary intake of calcium and protein, endocrine factors (sex steroids, calcitriol, IGF-1, thyroid hormone, growth hormone), mechanical forces (physical activity, weight), height, pubertal phase, prepubertal and postpubertal hormonal situation [16–20]. Genetic factors play an important role in the attainment of peak adult bone mass [4, 11, 21]. Due to ethnic and genetic factors, as well as traditional diets and physical activity habits, differences in BMD are also found across communities living in different geographical regions .
In our study, we observed that age had a negative effect on femoral neck BMD in patients overall, and on both lumbar spine and femoral neck BMD in patients under 50. We think that an absence of an effect of age on lumbar vertebrae in the whole group and in men above 50 might be related to an increase in BMD due to bone degeneration .
Minerals, particularly calcium, phosphate and magnesium are very important in regulating skeletal development . Nguyen et al. , in a study of 690 elderly men, found a significant positive correlation between dietary intake of calcium and lumbar spine and femoral neck BMD. Similarly, Kelly et al. , in a study of men aged 21–79 years, reported a significant positive correlation between lumbar spine and femoral neck BMD and calcium intake. Tanaka et al. , in a study of 325 men aged above 50 years, did not find a relation between calcium intake and osteoporosis. Welten et al. , in a 15-year longitudinal study of 84 healthy males and 98 females, did not find a statistically significant relation between calcium intake and lumbar spine BMD.
A generally accepted view is that an increase in calcium intake in childhood and adolescence will cause a greater increase in bone mass, and that a greater peak bone mass will be gained . It has been reported that positive effects of sufficient calcium intake on BMD are observed most evidently in the prepubertal and early pubertal stages, and that the effect of calcium intake and exercise on formation of optimal bone mass rapidly decrease after adolescence [16, 17]. Morgan et al.  report that calcium intake is effective in decreasing bone loss in postmenopausal women, but that calcium and vitamin D intake are not sufficient for decreasing bone loss in men. Jaime et al. , in a study of 277 men above age 50, reported that calcium intake is effective for increasing femoral neck BMD in black men but not in white men.
In our study, we did not find a statistically significant relation between BMD and calcium intake. However, our study was cross-sectional and evaluated the calcium intake of patients only within the past year. It did not cover the prepubertal and pubertal stages in which the effect of calcium intake on bone mineral density is well established. Longitudinal studies would give more definitive results on this matter.
Bone mass can be increased by means of daily weight-bearing physical activities and it is widely believed that exercises producing significant impact loading on the skeleton cause increases in bone mass. Grimston et al. , in a study of boys and girls 10–16 years of age, showed that exercises producing impact loading were associated with increases in BMD in the femoral neck and lumbar spine. Walking is also known to have positive effects on bones. Walking increases the existing bone mass and decreases the pace of bone loss. In a study made in 1,075 women and 690 men, higher physical activity levels and higher quadriceps strength were associated with high BMD at the weight-bearing site of the femoral neck, but not at the lumbar spine, in both men and women . Welten et al. , in a 15-year prospective study in 84 males and 98 females aged 13–28, showed that the factor with the highest effect on vertebral BMD was physical activity in men, and body weight in women. In a 4-year study of 5,049 male patients with a mean age of 73, it was found that hip fracture incidence in a group making an average of an hour of exercise a day was half that in the other group making an average of half an hour of daily exercise . It has been established that daily physical activity is protective against hip fractures both in men and in women .
In our study, we saw that physical activity has effects on femoral neck BMD in men above 50. We could not find a similar positive effect in men under 50. There are studies showing that daily physical activities have effects on BMD as mentioned above [2, 30, 31]; there are also studies which suggest that positive effects of additional exercise programs are more prominent [2, 29, 32]. Among the individuals in our study, an important part of physical activities was daily life activities. Most of our patients were not in the habit of playing sports for the purpose of exercising. This may help to explain why we did not find the expected positive relation in men under 50, and between lumbar BMD and physical activity in men age 50 or above.
Caffeine consumption has been reported to have negative effects on BMD [7, 8]. However, Glynn et al. , in a study of 523 men over age 50, could not find a relation between caffeine consumption and proximal femur BMD. Hannan et al. , in a study of men and women aged 67–90, showed that caffeine consumption did not affect lumbar spine and femoral neck BMD. Hegarty et al.  showed a positive effect of tea consumption on lumbar spine BMD in elderly women. In our study also, tea drinking had an important place as a form of caffeine consumption; however, we did not find any positive or negative effect of increased caffeine consumption on BMD.
Smoking is a risk factor for loss of bone mass. It affects the bone mass and the risk of fracture in various ways. It decreases calcium absorption and suppresses sex steroids in both men and women . Smoking increases the lifetime risk of developing vertebral and hip fracture . Tanaka et al. , in a study of 325 men above age 50, showed that those with a history of smoking in any period of their life had less bone mass.
In our study, lumbar spine BMD was negatively correlated with smoking in patients overall, and this correlation persisted when patients aged 50 and older were analyzed separately. It is thought that smoking suppresses sex steroids and thereby has a negative effect on BMD [26, 37].
Body mass index
In our study, we saw a positive significant relation between femoral neck BMD and BMI in men aged 50 and older. We could not find a similar relation in men under 50.
Low body weight is among the parameters associated with low BMD [28, 38]. The most important determinants of BMD during childhood and adolescence are body weight and sexual maturity . Nguyen et al. , in their study of men and women over 60, showed a significant positive correlation between BMI and lumbar spine BMD. Cetin et al. , in a study of men aged 43–73, found a significant positive correlation between BMI and lumbar spine BMD; however, they did not find a significant relation between femoral neck BMD and BMI. Smerdely et al.  have shown that low BMI in men aged 70 and above has negative effects on femoral neck BMD. Tanaka et al.  found a positive relation between low BMI and risk of development of osteoporosis in men aged 50 and above.
Increased BMI puts a mechanical load on the skeleton and creates positive effects on bone density. In overweight persons calcium absorption is higher and sensitivity to parathyroid hormone has dropped. Presence of excessive fat tissue increases transformation of androstenedione to estrone. These effects cause an increase in and protection of bone mass in both sexes [3, 4, 39].
Level of income and education levels
In our study, we did not find a significant relation between these and femoral neck or lumbar spine BMD.
In conclusion, the findings of our study are consistent with only some of the other studies in the literature. One possible explanation for the conflicting results of different studies may be that BMD is affected by genetic and geographic factors as well as the factors which are being studied. Differences in the methods of assessing BMD may also have an effect. Comprehensive longitudinal studies analyzing the effects of endogenous and exogenous factors, particularly genetic factors, are needed.