Variables for the prediction of femoral bone mineral status in American women.

Introduction: A reliable procedure for identifying persons at risk for osteoporosis and subsequent fracture is needed so that preventive measures may be initiated.

Material and Methods: Participants included 7,532 women, ages 20 and older, surveyed in the National Health and Nutrition Examination Survey III (NHANES, 1988-1994). Influences of race, body composition, exercise, alcohol intake, smoking status, as well as the effect of nutritional intake of calcium, phosphorus, magnesium, iron, zinc, sodium, and potassium on bone mineral density (BMD) were assessed.

Results: Advancing age, low body weight, low exercise expenditure, and smoking were significant predictors for low BMD. Nutritional variables examined were not significant in the predictive models.

Conclusions: The absence of calcium from the predictive models indicates the need for re-evaluation of the current recommended intake levels of this nutrient. A greater emphasis on factors such as exercise and achieving adequate weight is recommended.

Discussion: Providing women with the knowledge of their risk for low BMD may influence lifestyle behaviors, which may ultimately result in the prevention of bone injury.

Key Words: bone mineral density (BMD), osteoporosis, dual-energy x-ray absorptiometry (DXA)

**********

Osteoporosis is a major health threat for 44 million people in the United States, 80% of whom are women. (1) Often called the "silent thief," osteoporosis occurs when bone mineral density (BMD) is diminished to the extent that bones become brittle and weak. Bone mineral is silently removed until bone is so weak that even low-impact movement can induce bone fracture. There is no cure for osteoporosis; however, bisphosphonates have been found effective in preserving bone mass (2,3) as well as increasing bone mineral density. (4-6) Modifiable lifestyle factors have been suggested as determinants of osteoporosis, indicating that this is a preventable condition. (7)

Bone fractures are among the more detrimental consequences of osteoporosis, with hip fracture being the most serious for both the individual and for health care agencies. Hip fractures are painful and almost always require hospitalization. Complications may arise from either the fracture itself or from necessary surgery. Postoperative complications that may occur include urinary tract infections, pneumonia, and decubitus ulcers. (8) Depression is common in hip fractures and may be exacerbated by the loss of independence and need for in-home or nursing home care. (9)

BMD is thought to account for 75 to 85% of bone strength (7) and has been identified as a major determinant of risk for osteoporosis. (10,11) Maximum adult bone density, called peak bone mass, is achieved roughly within the age range of 30 to 35 years. (12-14) Thereafter, a reduction in BMD, which is thought to be a natural consequence of aging, is observed. (15,13) The timing of peak BMD varies depending on the lifestyle of the individual (eg, exercise, smoking, and nutrition). (16) Failure to achieve sufficient peak BMD at maturity may result in increased fracture risk as a result of reduced bone mass.

Genetic influences account for up to 80% of the maximum BMD achieved. (16) However, data indicate a smaller genetic contribution to BMD at the proximal femur and forearm than observed in the spine, suggesting that lifestyle factors such as diet and exercise may have an important role in determining the risk for low BMD and osteoporotic fracture at these sites. (15) More recent investigations have shown that inadequate intakes of calcium, inactivity, and adverse habits such as smoking and alcohol consumption contribute to a low peak BMD development and increased risk of fractures. (12,16)

The strong focus on the role of calcium has resulted in less emphasis on the role of other nutrients in bone health and osteoporosis. With the large variety of dietary intake practices in the US, the range of intake levels of many nutrients is extreme. The influence of several nutrients, either in excessive or depleted amounts, on bone health remains undetermined. Development of a predictive model regarding the influence of nutrient intake on bone status would greatly benefit both individuals at risk and health care professionals in the determination of low BMD and osteoporosis as well as dietary interventions for decreasing the risk for this debilitating condition.

Materials and Methods

Participants

The study consisted of 7,532 females, ages 20 and older, who participated in the Third National Health and Nutrition Examination Survey (NHANES III). The NHANES III sample is representative of the total civilian, noninstitutionalized population two months of age or older in the United States from 1988 to 1994. (18) For the purposes of this study, only female participants were assessed. Women are considered at greater risk than men for osteoporosis and subsequent fracture. The rate of bone loss observed throughout the adult lifecycle is higher in women than in men and most of the US adults with low femur BMD are women. (15,10)

Data Collection

NHANES III data were collected through household interviews and physical examinations. Interviewers and other staff received intensive training with retraining programs conducted throughout the survey to ensure that high skill levels were maintained.

Measurement of the Variables

The nutritional predictor variables examined included protein as a percent of caloric intake, calcium, phosphorus, magnesium, zinc, sodium, potassium, iron, and caffeine. Vitamin D intake was not included in this study as NHANES III did not gather information regarding intake of this nutrient. The 24-hour recall method was applied for estimating quantitative nutrient intakes. Trained dietary interviewers administered the recall using an automated interactive dietary interview and coding system (NHANES III Dietary Data Collection [DDC] system) developed for NHANES III. The DDC system was designed to obtain detailed information about foods, such as brand names, ingredients, and preparation methods. (20) Food composition data were obtained using the US Department of Agriculture Survey Nutrient Database.

Caffeine consumption data were obtained from food frequency information collected during the examination. Participants provided answers to the questions: "How often did you have regular colas and sodas, not diet; how often did you have regular coffee with caffeine; how often did you have tea with caffeine?" Intake was recorded as the number of times per month cola, coffee, or tea was consumed.

Alcohol intake was also obtained from food frequency information. Participants reported the number of times in the past month they had consumed beer, wine, or hard liquor. The United States Department of Health and Human Services Dietary Guidelines for Americans recommends that women limit alcoholic intake to one drink per day. (21) Based on this recommendation, participants were grouped as alcohol consumers that drink one or more alcoholic beverages per day or those who consume less than one alcoholic drink per day.

Smoking behavior was also obtained from self-reported information. Participants provided answers to the questions "have you smoked at least 100 cigarettes during your entire life" and "do you smoke cigarettes now?" Participants were grouped as smokers (those who have smoked more than 100 cigarettes and are currently smoking), former smokers (those who have smoked more than 100 cigarettes but are currently not smoking), and nonsmokers (those who had never smoked more than 100 cigarettes). (22)

To examine the influence of body composition on BMD, recorded weight and body mass index (BMI) was obtained from the examination files. Triceps, thigh, and suprailium skinfold measures were obtained by trained technicians during the examination component. The Siri equation was utilized to calculate percent body fat using these skinfold measures. (23)

Dual energy x-ray absorptiometry (DXA) was used to measure bone density of the proximal femur. Measurements were obtained using three densitometers (Hologic ADR 1,000; Waltham, MA) located in mobile examination centers (MECs). Each DXA instrument was monitored by anthropomorphic phantom scans which were performed weekly. For comparability of the three DXA instruments, an additional phantom was circulated between MECs and scanned for each of the two measurement phases. (24,25)

Statistical Analysis

The data for this study were analyzed to develop predictive models for the diagnosis of low BMD and osteoporosis for women. Four sites of the proximal femur measured for BMD were used: femoral neck, trochanter, intertrochanter, and total proximal femur.

The independent variables included protein as a percent of total caloric intake, calcium, phosphorus, magnesium, zinc, sodium, potassium, iron, and caffeine intakes as well as height, BMI, and percent body fat as obtained by skinfold measures. Physical activity levels, alcohol and tobacco use, and race were controlled variables. Logistic regression analysis was used to determine the relevant variance of each variable, in combination with the other variables, in predicting BMD at these four sites.

Several of the independent variables were assessed as categories for the purpose of logistic regression analysis. The participants were grouped into four BMI categories. A BMI of less than 20 defined the underweight group. Participants with a BMI between 20 and 25 were classified as acceptable weight. Those with a BMI greater than 25 but no more than 30 comprised the overweight group, while participants with a BMI greater than 30 were classified as obese. (22)

The dependent variables were divided into categories of bone status according to Looker et al. (10) For the total proximal femur, normal BMD was defined as > 0.82 g/[cm.sup.2] and low BMD as < 0.82. Femur neck categories were defined as normal BMD being > 0.74 g/[cm.sup.2] and low BMD < 0.74. Trochanter measures were differentiated at 0.61 g/[cm.sup.2] and intertrochanter as 0.95 g/[cm.sup.2]. These values correspond with the World Health Organization's diagnostic criteria for low bone mass. (10)

Data were analyzed using Standard Analysis System (SAS) and JMP statistical program (2001). An alpha level of 0.05 was used to test for significance.

Results

Participant demographic information is displayed in Table 1. The average age of the study sample was 48.4 years (SD = 18.9). The average weight of the participants was 70.88 kg (SD = 17.23), with a mean percent body fat of 28.3. Results of BMD categorization are reported in Table 2.

Table 3 displays the comparison of the mean values for percent body fat, BMI, BMD, protein, calcium, phosphorus, magnesium, zinc, potassium, iron, and sodium intakes with values indicated as recommended for healthy women. Participants were, on the average, above the levels recommended for a healthy percent body fat and BMI. However, these women had mean BMD values above the lower limits recommended. Dietary intake means for protein and phosphorus were reported above the recommended daily amounts. Intakes of calcium, magnesium, and zinc were below the recommended level. Mean iron intake level was below the 15 mg recommended for women aged 20 to 50 years, but above the recommendation of 10 mg per day for women over the age of 50.

Results of the logistic regression analysis are displayed in Tables 4 through 7. Table 4 presents the model generated for predicting total proximal femur BMD. Increasing age, lower weight, low phosphorus intake, increased percent body fat, underweight as determined by BMI, current smoker, and low exercise intensity were significant risk factors for predicting low bone density at this site. Being a non-Hispanic black was found to be protective of BMD and exerted the strongest influence on the outcome variable. Being under-weight, as determined by BMI, was the next strongest predictor, followed by current smoker status.

The predictive model for femur neck BMD (Table 5) revealed increasing age, lower weight, increased percent body fat, and underweight as determined by BMI as significant risk factors for low BMD. Being Mexican-American, non-Hispanic black, or classified as "other" for race-ethnicity was protective of BMD at this site. As with the model generated for total proximal femur BMD, non-Hispanic black race and underweight as determined by BMI were the strongest predictors for BMD.

Logistic regression analysis of BMD at the trochanter site (Table 6) revealed advancing age, low weight, increased percent body fat, low calcium intake, obese status as determined by BMI, smoker status, and low exercise intensity as significant predictors. Again, being non-Hispanic black decreased the risk of low BMD and exerted the strongest influence on BMD at this site. Being obese as defined by BMI, low exercise intensity, and being a smoker were also strong predictors.

The predictive model for intertrochanter BMD (Table 7) revealed increasing age, lower weight, low phosphorus intake, high protein intake, underweight as determined by BMI, smoker status, and low exercise intensity as significant factors. Non-Hispanic black race-ethnicity was again a protective variable and the strongest predictor of BMD. A BMI classification as underweight, smoker status, and low exercise intensity were also strong predictors.

Discussion

Identification of risk factors for low BMD and osteoporosis is useful in developing preventive strategies that result in reduced exposure to these factors. Assessing an individual for risk factors can be used as a screening strategy for indicating susceptibility for bone mineral depletion. Findings of this study were used to develop an assessment tool for risk level, and prevention of low BMD among women.

To aid clinicians in identifying women at risk for osteoporosis, the independent variables in this study are classified according to their strength as risk factors for diminished BMD (Table 8). Variables that remained in the logistic regression models were classified as significant risk factors, while variables significant in univariate analyses, but not in multivariate tests, were classified as low risk factors.

Age is a strong risk factor for diminished BMD. (19,26,27) Although this is not a modifiable risk factor, screening programs should be designed that target older women, informing them of their increased risk and need for modifying eating and activity behaviors that may reduce the incidence of bone mineral loss.

Previous studies have shown that overweight women have a protective effect against depletion of bone mineral. (22,28-30) Results of this study classify indices of body composition: weight, BMI, and percent body fat, as significant risk factors for low BMD of the proximal hip. These risk factors indicate that the underweight woman is at greater risk for osteoporosis when compared with the woman with an acceptable weight. However, the increased risk of cardiometabolic disease among obese individuals legislates against counseling women to gain excessive weight merely to reduce bone mineral loss. A BMI measure of overweight produced no significant differences in BMD as compared with the normal weight group. Being overweight (not obese) is preferable to being underweight for maintenance of proximal femur BMD; however, being overweight has no protective advantage over being normal weight.

Identifying women who are underweight, have a low BMI, or low percent body fat is a beneficial screening tool for the identification for those at greatest risk for developing osteoporosis. Educational programs should include strategies to assist underweight women in achieving and maintaining a healthy body weight. This is of particular importance with young women who have yet to reach peak bone mass.

Weight-bearing physical activity has been shown to be an important stimulus for remodeling and strengthening of bone. (31) Vincent and Braith (32) reported increases in biochemical indices of bone turnover in elderly women who performed high intensity resistance exercise, as compared with low intensity resistance. Exercise conveying high impact forces may be more protective than activities producing minimal or low levels of muscle and bone stress. The predictive models generated in this study support this finding in that low exercise intensity was found to be a significant risk factor for low BMD at the trochanter, intertrochanter, and total femur sites. These results reflect only high intensity exercise, with all else classified as low intensity exercise due to the lack of a suitable equation for categorizing this data set to include moderate intensity exercise. Therefore, the effect of moderate levels of physical activity is not examined, separate from low intensity exercise and no exercise. Furthermore, the questionnaire did not allow a precise quantification of exercise volume, now considered to be an important factor in protection against morbidity and mortality. (19,33)

Educating women on the importance of exercise for the maintenance of bone health is essential. Helping women identify an exercise regimen that they enjoy should ensure greater compliance. This is important with the elderly, osteoporotic woman who may be fearful of any physical activity. Research (34) suggests that jumping activities may be the most effective exercise modality for prevention of bone mineral loss.

Studies investigating the effect of smoking on BMD have varied results. Mazess and Barden (35) found smoking to be a significant risk factor for decreased spinal BMD in young females. Turner et al (22) found no significant relationship between smoking and bone fracture risk in women aged 50 years and older, while Law and Hackshaw (36) found that post-menopausal women who smoked decreased their BMD by an additional 2% each decade of life. The findings of the present study suggest that smoking is a significant risk factor for low BMD or osteoporosis at the proximal femur. These results may be explained by the lower body weight often observed in women who smoke. (36) These women may also practice other adverse health behaviors that can negatively impact bone health, such as poor intakes of calcium, phosphorus, and protein, and sedentary lifestyles.

Excessive protein intake, as well as inadequate protein intake, has been found to be detrimental to BMD. (14,37-39) Results of this study indicate protein intake to be significant at the intertrochanter site only. Percentage of caloric intake as protein was not found to be significant for any predictive models. The mean protein intake in this study was 0.9 g/kg of body weight. This is only slightly above the recommended daily intake of 0.8 g/kg of appropriate body weight. (40) This intake level may not be sufficient enough to induce calciuria and lead to loss of bone mineral.

The significance of dietary calcium in the attainment of maximal BMD is well documented. (40-43) However, the evidence of a direct relationship between dietary calcium and bone mineral status is inconsistent. In this study, calcium intake was retained as a significant factor only at the trochanter site. These findings support those of Kanders et al (44) in that the effect of physical activity was more significant than calcium intake as a risk factor for low BMD in adult females. The mean calcium intake in this study was only 659 mg per day, two-thirds of the dietary reference intake (DRI) for women 19 to 50 years old and 55% of the DRI for women over the age of 50. Even at deficient intake levels, calcium was not found to be a significant predictor of BMD. Perhaps the threshold level of calcium needed to prevent BMD loss is lower than commonly stated.

Evidence that high calcium intakes are not significant for achievement or maintenance of healthy BMD is becoming more apparent in the scientific literature. (17,35,39,45) Angus et al (17) suggested that bone mass is influenced by dietary factors other than calcium. The role of dietary calcium in achieving and maintaining healthy bone needs to be re-evaluated. Health care providers may unwittingly be recommending unnecessary high doses of calcium for women, while artificially creating hopes of preventing bone loss and osteoporosis.

Past studies have shown that deficient intakes of phosphorus lead to decreased mineralization of bone matrix (46) while excessive intakes result in an increase in bone resorption leading to bone loss. (47) Results of this study found phosphorus to be significant in predicting BMD of the intertrochanter and total proximal femur regions, supporting findings of past studies. The influence of phosphorous intake on BMD suggests that assessment of phosphorus intake should be a consideration in the determination of risk for low BMD or osteoporosis.

The nutritional variables of magnesium, zinc, sodium, potassium, and iron were not found to be significant predictor variables for BMD of the proximal femur. Risk factors for diminished BMD may be specific for certain types of bone. Vertebral bone is 66 to 90% trabecular bone while the bone tissue of the hip is only 25% trabecular bone. (48) These nutrients may exert a greater influence on trabecular bone development and maintenance than on cortical bone.

Underreporting of sodium intake could be an aspect in the lack of significance found with sodium as a risk factor for low BMD. NHANES III obtained sodium intake values from data on the sodium content of foods eaten over a 24-hour period. Sodium added to foods during or after preparation was not assessed. A more precise report of sodium intake needs to be obtained to accurately assess the effect of sodium intake.

Caloric, fat, carbohydrate, zinc, and sodium intakes, and the percent of caloric intake as protein were found to be low risk factors for low BMD for each of the dependent variables. Protein intake was identified as a low risk factor for the femoral neck, trochanter and total femur. Phosphorus was significant at the femur neck and trochanter, calcium at the total femur, and potassium at the intertrochanter site. Nutritional evaluation by a registered dietitian is important for identifying women with inadequate dietary intakes of these nutrients. Altering nutritional regimens to prevent bone loss and increase bone density is a behavior modification that can be achieved with proper education on food choices.

This study did not find caffeine intake to be a significant risk factor for low BMD. Subjects in NHANES III reported the number of coffee, tea, or regular cola and soda beverages consumed per month. Intake of diet colas and sodas was not assessed, possibly resulting in a significant underreporting of daily caffeine intake which may have influenced results.

Several physiologic and psychological effects occur in the body as a result of caffeine ingestion. The adverse effects of caffeine use (stomach upset, nervousness, irritability, headaches, and diarrhea) (40) must be weighed against any possible benefits. Caffeine-containing beverages generally contain no nutritional value and often replace beverages such as milk, that would supply nutrients that are beneficial to bone status. Therefore, limiting consumption of these beverages is advisable in the prevention and treatment of low BMD.

The present study did not find significance with alcohol intake and BMD. However, alcohol is often consumed at the expense of healthy foods that may provide proper nutrition for bone. Excessive alcohol intake interferes with the body's use of nutrients, making them ineffective in maintaining healthy bone as well as other important metabolic functions. (27)

Berarducci et al (49) report that primary care providers typically assess osteoporosis-associated risk factors more often in women 50 years or older. Information and education related to osteoporosis prevention were more frequently provided to this age group. The importance of educating young women on the impact their current lifestyle and nutritional habits may have on the future of their bone status is crucial and often overlooked. Decreases in bone mass may occur as early as the third decade of life. (16) Therefore timely interventions are important in preventing additional unnecessary bone loss. The number of young women at risk of developing osteoporosis is unknown. Identifying and educating susceptible young women may make a difference that will impact the rest of their lives.

Conclusion

Health care providers and health educators are in a position to facilitate the decision to adopt new behaviors in the prevention or treatment of low BMD or osteoporosis. Achieving and maintaining an appropriate body weight, starting and maintaining an exercise program, smoking cessation, and appropriate dietary practices are issues that health professionals need to address with female clients of all ages. Knowledge of increased risk for osteoporosis due to advancing age and ethnicity may affect an individual's lifestyle behavior, which may result in the prevention of future bone injury. As with chronic diseases, early detection of low BMD is crucial. Early initiation of preventive practices is of critical importance but is often overlooked. Health educators need to implement interventions that address the specific needs of all women, young and postmenopausal, to decrease the chances of being robbed by this "silent thief."

References

1. National Institutes of Health. Osteoporosis Overview. 2005. Available at: http://www.osteo.org/osteo.html. Accessed July 15, 2005.

2. Meunier PJ, Confavreux E, Tupinon I, et al. Prevention of early postmenopausal bone loss with cyclical etidronate therapy. J Clin Endocrinol Metab 1997;82:3740.

3. Russell RG, Rogers MJ. Bisphosphonates: from the laboratory to the clinic and back again. Bone 1999;25:97-106.

4. Chan AS. Ten years of alendronate treatment for osteoporosis in postmenopausal women. N Engl J Med 2004;351:190-192.

5. Cohen SB. An update on bisphosphonates. Curr Rheumatol Rep 2004;6:59-65.

6. Fleisch HA. Bisphosphonates: preclinical aspects and use in osteoporosis. Ann Med 1997;29:55-62.

7. Notelovitz M. Osteoporosis: Prevention, diagnosis and management. Professional Communications, Inc; 1997.

8. Kanis JA. Textbook of Osteoporosis. Cambridge, England, University Press, 1996.

9. Melton LJ. Hip fractures: a worldwide problem today and tomorrow. Bone 1993;14:S1-S8.

10. Looker AC, Orwoll ES, Johnston CC, et al. Prevalence of low femoral bone density in older US adults from NHANES III. J Bone Miner Res 1997;12:1761-1768.

11. Teegarden D, Lyle RM, McCabe GP, et al. Dietary calcium, protein, and phosphorus are related to bone mineral density and content in young women. Am J Clin Nutr 1998;68:749-754.

12. Carmichael KA, Carmichael DH. Bone metabolism and osteopenia in eating disorders. Medicine 1995;75:254-267.

13. Recker RR, Davies KM, Hinders SM, et al. Bone gain in young adult women. JAMA 1992;268:2403-2408.

14. Swaminathan R. Nutritional factors in osteoporosis. Int J Clin Pract 1999;53:540-548.

15. Burger H, de Laet CE, van Daele PL, et al. Risk factors for increased bone loss in an elderly population. Am J Edpdemiol 1998;147:871-879.

16. Johnston CC, Slemenda CW. Determinants of peak bone mass. Osteoporos Int 1993;1:S54-S55.

17. Angus RM, Sambrook PN, Pocock NA, et al. Dietary intake and bone mineral density. Bone Min 1988;4:265-277.

18. Vital and health statistics: Sample design: Third national health and nutrition examination survey. Washington, DC: US Dept of Health and Human Services: 1992. Publication No. (PHS) 92-1387.

19. Paffenbarger RS, Kampert JB, Lee I, et al. Changes in physical activity and other lifeway patterns influencing longevity. Med Sci Sports Exerc 1993;26:857-865.

20. McDowell MA, Briefel RR, Alaimo K, et al. Energy and macronutrient intakes of persons ages 2 months and over in the United States: Third national health and nutrition examination survey. Phase I 1988-91 (Advance Data from Vital and Health statistics No. 255). Hyattsville, MD, National Center for Health Statistics, 1994.

21. Dietary guidelines for Americans: If you drink alcoholic beverages, do so in moderation. Washington, DC: US Department of Agriculture; 1993. Home and Garden Bulletin No. 253-8.

22. Turner LW, Wang MW, Fu Q. Risk factors for hip fracture among southern older women. South Med J 1998;91:533-540.

23. Adams GM. Exercise Physiology Laboratory Manual, 3rd ed. Boston, McGraw-Hill, 1998.

24. Looker AC, Wahner HW, Dunn WL, et al. Updated data on proximal femur bone mineral levels of US adults. Osteoporos Int 1998;8:468-489.

25. National Health and Nutrition Examination Survey III, 1988-94 (CD-ROM Series 11, No. 2A) [Data file]. Hyattsville, MD, National Center for Health Statistics

26. McArdle WD, Katch FI, Katch VL. Exercise Physiology: Energy, Nutrition, and Human Performance, 4th ed. Baltimore, Williams & Wilkins, 1996.

27. Whitney EN, Rolfes SR. Understanding Nutrition, 8th ed. Belmont, CA, West/Wadsworth, 1999.

28. Lewis RD, Modlesky CM. Nutrition, physical activity, and bone health in women. Int J Sport Nutr Exerc Metab 1998;8:250-284.

29. Bohannon AD. Osteoporosis and African American women. J Womens Health Gender-Based Med 1999;8:609-615.

30. Lindsay R, Cosman F, Herrington BS, et al. Bone mass and body composition in normal women. J Bone Miner Res 1992;7:55-63.

31. Madsen KL, Adams WC, Van Loan MD. Effects of physical activity, body weight and composition, and muscular strength on bone density in young women. Med Sci Sports Exerc 1998;30:14-120.

32. Weiler HA, Janzen L, Green K, et al. Percent body fat and bone mass in healthy Canadian females 10 to 19 years of age. Bone 2000;27:203-207.

33. Renfro J, Brown JB. Understanding and preventing osteoporosis. A AOHN J 1998;46:181-191.

34. Vincent KR, Braith RW. Resistance exercise and bone turnover in elderly men and women. Med Sci Sports Exerc 2000;34:17-23.

35. Blair SN, Cheng Y, Holder JS. Is physical activity or physical fitness more important in defining health benefits? Med Sci Sports Exerc 2001;33:S379-S399.

36. Bassey EJ, Rothwell MC, Littlewood JJ, et al. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone Miner Res 1998;13:1805-1813.

37. Mazess RB, Barden HS. Bone density in premenopausal women: Effects of age, dietary intake, physical activity, smoking, and birth-control pills. Am J Clin Nutr 1991;53:132-142.

38. Law MR, Hackshaw AK. A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: recognition of a major effect. BMJ 1997;315:841-846.

39. Lutz J. Calcium balance and acid-base status of women as affected by increased protein intake and by sodium bicarbonate ingestion. Am J Clin Nutr 1984;39:281-288.

40. Metz JA, Anderson JJ, Gallagher PN. Intakes of calcium, phosphorus, and protein, and physical activity level are related to radial bone mass in young adult women. Am J Clin Nutr 1993;58:537-542.

41. Hannan MT, Felson DT, Dawson-Hughes B. Risk factors for longitudinal bone loss in elderly men and women: The Framingham osteoporosis study. J Bone Miner Res 2000;15:710-720.

42. Bonjour JP, Carrie AL, Ferrari S, et al. Calcium-enriched foods and bone mass growth in prepubertal girls: A randomized, double-blind, placebo-controlled trial. J Clin Invest 1997;99:1287-1294.

43. Branca F, Vatuena S, et al. Calcium, physical activity and bone health--building bones for a stronger future. Public Health Nutr 2001;4:17-123.

44. Ilich JZ, Kerstetter JE. Nutrition in bone health revisited: A story beyond calcium. J Am Coll Nutr 2000;19:715-737.

45. Kanders B, Dempster DW, Lindsay R. Interaction of calcium nutrition and physical activity on bone mass in young women. J Bone Miner Res 1998;3:1445-1149.

46. Maggiolini M, Bonofiglio D, Giorno A, et al. The effect of dietary calcium intake on bone mineral density in healthy adolescent girls and young women in Southern Italy. Int J Epidemiol 1999;28:479-484.

47. Cashman KD, Flynn A. Optimal nutrition: Calcium, magnesium and phosphorus. Proc Nutr Soc 1999;58:477-487.

48. Schaafsma G, van Beresteyn ECH, Raymakers JA, et al. Nutritional aspects of osteoporosis. World Rev Nutr Diet 1987;49:121-159.

49. Bilezikian JP, Raisz LG, Rodan GA. Principles of Bone Biology. San Diego, Academic Press, 1996.</p> <pre> Democracy arises out of the notion that those who are equal in any respect are equal in all respects; because men are equally free, they claim to be absolutely equal. --Aristotle </pre> <p>Martha Bass, PHD, RD, M. Allison Ford, PHD, Barry Brown, PHD, Andy Mauromoustakos, PHD, and Roseanne S. Keathley, PHD

From Sam Houston State University-Health Education, Huntsville, TX and the Dept. of Health, Kinesiology, Recreation and Dance, University of Arkansas, Fayetteville, AR.

Reprint requests to Martha Bass, PhD RD, Sam Houston State University-Health Education, PO Box 2176, Huntsville, TX. Email: [e-mail]

Accepted August 2, 2005.

RELATED ARTICLE: Key Points

* Advancing age, low body weight, low exercise expenditure, and smoking are significant predictors for low bone mineral density.

* Calcium was not a predictor of low BMD, suggesting the recommended intake level may need to be reevaluated.

* Health providers and health educators need to place a greater emphasis on exercise and achieving adequate weight in order to reduce the risk of osteoporosis and subsequent bone fractures.

Table 1. Study population characteristics (n = 7,532)

Variables                                     n      %

Age (yrs)
  20-54                                       4,700  62
  [greater than or equal to] 55               2,832  38
Race
  White                                       3,251  43
  Black                                       2,129  28
  Hispanic                                    1,827  24
  Other                                         325   5
Body mass index
  Underweight                                   597   8
  Normal weight                               2,360  31
  Overweight                                  2,348  31
  Obese                                       2,211  30
Physical activity
  Low                                         6,670  89
  High                                          862  11
Smoking status
  Smoke                                       1,649  22
  Former smoker                               1,300  17
  Nonsmoker                                   4,583  61
Alcohol use
  < 1 drink per day                           7,185  96
  [greater than or equal to] 1 drink per day    337   4

Table 2. Bone mineral density categorizations (a)

Category         n      %

Total femur
  Low            2,471  32.8
  Healthy        5,061  67.2
Femur neck
  Low            2,870  38.1
  Healthy        4,662  61.9
Trochanter
  Low            2,620  34.8
  Healthy        4,912  65.2
Intertrochanter
  Low            2,241  29.7
  Healthy        5,291  70.3

(a) Looker et al, 1997.

Table 3. Mean values with values indicated as recommended for healthy
women

Variable                                 Mean      Recommendation

% Body fat                                  28.30  25 (a)
BMI                                         27.50  21.3-22.1 (a)
Total proximal femur BMD (g/[cm.sup.2])      0.89  > 0.82 (b)
Femur neck BMD (g/[cm.sup.2])                0.79  > 0.74 (b)
Trochanter BMD (g/[cm.sup.2])                0.66  > 0.61 (b)
Intertrochanter BMD (g/[cm.sup.2])           1.05  > 0.95 (b)
Protein (g)                                 65.00  46-50 (c)
Calcium (mg)                               659.00  age 19-50: 1,000 (c)
                                                   age > 50: 1,200 (c)
Phosphorus (mg)                          1,033.60  700 (c)
Magnesium (mg)                             241.20  age 19-30: 310 (c)
                                                   age > 30: 320 (c)
Zinc (mg)                                    9.30  12 (c)
Potassium (mg)                           2,323.30  3,500 (c)
Iron (mg)                                   12.40  age 20-50: 15 (c)
                                                   age > 50: 10 (c)
Sodium (mg)                              2,786.9   2,400 (c)

(a) McArdle et al, 1996.
(b) Looker et al, 1997.
(c) Whitney & Rolfes, 1999.

Table 4. Predictive model for low total proximal femur BMD: Logistic
regression

Model: Y (a) = 0.3839 + 0.0757 (age) - 0.0868 (weight) - 0.00015
               (phosphorus) + 0.0230 (% body fat) - 0.8735 (race:
               non-Hispanic black) + 0.6135 (BMI: underweight) + 0.4768
               (smoker) + 0.4213 (exercise intensity: low)

(a) The log of the probability of low BMD vs healthy BMD at the total
femur site

Table 5. Predictive model for low femur neck BMD: Logistic regression

Model: Y (a) = 0.1530 + 0.0826 (age) - 0.0761 (weight) + 0.0259 (% body
               fat) - 0.3839 (race: Mexican-American) - 1.1400 (race:
               non-Hispanic black) - 0.4184 (race: other) + 0.4776 (BMI:
               underweight)

(a) The log of the probability of low BMD vs healthy BMD at the femur
neck site

Table 6. Predictive model for low trochanter BMD: Logistic regression

Model: Y (a) = 0.3039 + 0.0637 (age) - 0.0768 (weight) + 0.0235 (% body
               fat) - 0.00021 (calcium) - 0.5931 (race: non-Hispanic
               black) + 0.4692 (BMI: obese) + 0.3890 (smoker) + 0.4453
               (exercise intensity: low)

(a) The log of the probability of low BMD vs healthy BMD at the
trochanter site

Table 7. Predictive model for low intertrochanter BMD: Logistic
regression

Model: Y (a) = 0.4532 + 0.0724 (age) - 0.0772 (weight) - 0.0004
               (phosphorus) + 0.0047 (protein) - 0.9019 (race:
               non-Hispanic black) + 0.6012 (BMI: underweight) + 0.4360
               (smoker) + 0.3322 (exercise intensity: low)

(a) The log of the probability of low BMD vs healthy BMD at the
intertrochanter site

Table 8. At-risk indicators for low femoral bone mineral density

                      Total proximal femur     Femur neck

Significant           [up arrow] Age           [up arrow] Age
  risk factors        [down arrow] Weight      [down arrow] Weight
                      [down arrow] Intensity   [up arrow] % Body fat
                        exercise
                      [up arrow] % Body Fat    BMI: underweight
                      [down arrow] Phosphorus  Race: Mexican-American
                        intake                   and black are
                                                 protective
                      Smoker
                      BMI: underweight
                      Race: black is
                        protective
Low risk factors (a)  Calories                 Calories
                      Protein                  Protein
                      Fat                      Fat
                      Carbohydrates            Carbohydrates
                      Calcium                  Phosphorus
                      Zinc                     Zinc
                      Sodium                   Sodium
                      % Kcal as protein        Potassium
                                               % Kcal as protein

                      Trochanter              Intertrochanter

Significant           [up arrow] Age          [up arrow] Age
  risk factors        [down arrow] Weight     [down arrow] Weight
                      [down arrow] Intensity  [down arrow] Intensity
                        exercise                exercise
                      [up arrow] % Body fat   [down arrow] Phosphorus
                      [down arrow] Calcium      intake
                        Intake                [up arrow] Protein intake
                      Smoker                  Smoker
                      BMI: obese              BMI: underweight
                      Race: black is          Race: black is protective
                        protective
Low risk factors (a)  Calories                Calories
                      Protein                 Fat
                      Fat                     Carbohydrates
                      Carbohydrates           Zinc
                      Phosphorus              Sodium
                      Zinc                    Potassium
                      Sodium                  % Kcal as protein
                      % Kcal as protein

(a) Significant (P < 0.05) in univariate analyses