Professional Documents
Culture Documents
Diane Feskanich, ScD, Walter C. Willett, MD, DrPH, Meir J. Stampfer MD,
DrPH, and Graham A. Colditz, MD, DrPH
June 1997, Vol. 87, No. 6 American Journal of Public Health 993
Feskanich et al.
TABLE 2-Relative Risks (RRs) with 95% Confidence Intervals (Cls) for Hip and Forearm Fractures, by Frequency of
Milk Consumption, in a Cohort of 77 761 Womena Aged 34 through 59 Years at Baseline in 1980 and Followed
for 12 Years
s1/wk 213 264 41 1.00 1.00 ... 342 1.00 1.00 ...
2-6/wk 140 518 33 1.22 1.36 0.86,2.16 235 1.06 1.04 0.88, 1.23
1/d 152 795 31 1.05 1.23 0.76, 1.98 231 0.95 0.92 0.78,1.09
22/d 138 644 28 1.08 1.45 0.87, 2.43 238 1.12 1.05 0.88, 1.25
Pfor trendd .92 .22 .26 .71
TABLE 3-Relative Risks (RRs) with 95% Confidence Intervals (Cis) for Hip and Forearm Fractures, by Calcium Intake,
in a Cohort of 77 761 Womena Aged 34 through 59 Years at Baseline in 1980 and Followed for 12 Years
through 69 years of age. For forearm were more likely to drink milk as adults.
Results
fractures, incidence rates in these two age The use of calcium supplements increased
In 645 221 person-years of fol- groups were 104 and 381, respectively. sharply from 9% of the population in 1982
low-up (median 8.3 years per participant), At baseline, the frequency of milk to 50% in 1986 and declined somewhat
133 hip fractures and 1046 forearm frac- consumption was positively associated during the remaining follow-up period.
tures due to low or moderate trauma were with dietary intakes of calcium, vitamin D,
documented. The incidence (per 100 000 protein, phosphorus, and potassium (Table Adult Milk Consumption
person-years) of hip fractures increased 1). Women drinking more alcohol were
from 15 among women 40 through 44 less likely to drink milk, and women who In the 12 years of follow-up, we
years of age to 140 among women 65 drank milk more frequently as teen- agers found no evidence to suggest a decreased
994 American Journal of Public Health June 1997. Vol. 87, No. 6
Milk and Bone Fractures
June 1997, Vol. 87, No. 6 American Joumnal of Public Health 995
Feskanich et al.
Discussion dius,'-4 spine,46 and proximal femur47 association was observed with hip frac-
with calcium supplementation, particu- ture.41
These data suggest that more fre- larly in older women and when diets are With a fourfold difference in calcium
quent milk consumption and higher di- low in calcium.37'38 However, demon- intake between the 10th and 90th popula-
etary calcium intakes in middle-aged and strated reduction in bone loss over the 2 or tion percentiles, it seems unlikely that the
older women do not provide any substan- 3 years of most clinical trials may not lack of association between dietary cal-
tial protection against hip or forearm substantially alter the risk of fracture if cium and fracture risk in our cohort can be
fractures. Increased consumption of milk bone remodeling reaches a new steady due to insufficient variation in diet.
and dairy foods, which are sources of both state27 or if trabeculae remain discon- Misclassification of calcium intake could
calcium and vitamin D, have been found nected.'7 A 45% reduction in vertebral attenuate associations, but it would not
to improve calcium balance24 and to lower fractures was reported among elderly explain the positive association observed
rates of bone loss in the spine25 and women after 4 years of calcium supple- between dietary calcium and hip fractures.
femoral neck26 in clinical trials with adult mentation,13 but only among those who Also, the same food-frequency question-
women. However, it is possible that the entered the study with a previous vertebral naire was used to detect a clear inverse
observed improvements were transient fracture, and 43% fewer hip fractures association between calcium intake and
effects of bone remodeling that may not were observed after 18 months in calcium- kidney stones in a population of men.42
be sustained in the long term.27 treated women,9 although the results We cannot exclude the possibility that
Calcium is essential for building cannot necessarily be attributed to cal- fracture risk might be decreased at dietary
peak bone mass,28'29 but the extent to cium since the women also received a levels above those obtained in our cohort.
which adult milk consumption helps to vitamin D supplement. This study was limited by the fact
maintain bone structure is unknown. Of the previously published prospec- that the women in our population were
Drinking more than one glass of milk per tive population studies of calcium and hip relatively young for hip fractures. How-
day as an adult eliminated a positive fractures, only one reported a protective ever, we continued to see no beneficial
association between lifetime caffeine in- effect. Holbrook et al.8 found a 60% effect of higher milk consumption on hip
take and reduced bone densities at the hip reduction in the risk of hip fracture among fracture risk even when analyses were
and lumbar spine,30 and drinking milk older men and women consuming more limited to women over 60 years of age.
with every meal in adulthood was signifi- than 765 mg/day compared with those Our cohort also exhibited relatively low
cantly associated with higher bone density with lower intakes. However, the study age-adjusted incidence rates for hip frac-
at the midradius, spine, and hip.3 ' was limited by the small number of ture compared with those reported from
However, milk consumption in child- fractures (n = 33) and the use of a 1-day national hospital discharge data,43 which
hood, adolescence, and adulthood are dietary recall to assess calcium intake may limit the generalizability of our
often highly correlated, making it difficult
to determine which periods are truly during the 14 years of follow-up. Calcium results.
responsible for the higher adult bone did not demonstrate a protective effect in Many of the known and suspected
densities. Higher milk consumption nationally representative populations in factors that affect bone density were
during childhood and adolescence has Britain'2 and the United States," although measured and included in multivariate
been associated with higher radial, femo- the US data suggested that benefits may analyses. However, unmeasured factors
ral, and vertebral bone densities in be limited to older postmenopausal women might still confound the observed relation-
studies with middle-aged and older not on estrogen replacement therapy. Two ships between diet and fracture incidence.
women,32-34 while the relationships with other large prospective studies also failed For example, supplemental vitamin D has
adult milk consumption were either null to find a protective effect of calcium on been shown to reduce fracture incidence
or positive but not significant. Although hip fractures.'0"'4 among elderly men and women,44 and
not statistically significant, our finding In our cohort, women consuming therefore measurement of sunlight expo-
of a possible beneficial effect of teen- greater amounts of calcium from dairy sure as a proxy for endogenous produc-
aged milk consumption on risk of adult foods had modest but significantly in- tion of cholecalciferol may be an impor-
hip fracture is compatible with these creased risks of hip fracture while no tant factor in assessing fracture risk.
studies. increase in fracture risk was observed for Recall bias is not a concem in this
In dairy-consuming regions of China the same levels of calcium from nondairy study since the dietary data were collected
and Yugoslavia, women had higher radial sources. We have no reason to believe that prior to fracture incidence. The recall of
bone densities35 and a lower rate of hip dairy calcium itself was responsible for teenaged diet was collected after follow-
fracture,36 respectively, than women liv- the observed increase in risk of hip up was begun, although we have no
ing in regions with low dairy intake. In fracture; rather, some other characteristic reason to assume that women would bias
both reports, the authors believed that the of women who consume dairy foods or their recall based on their suspicion of
results were due to the formation of higher some other nutrient or nonnutrient compo- fracture risk. To minimize this possibility,
peak bone densities among those in the nent of dairy foods might have contrib- we controlled for current milk intake in
dairy regions. In fact, although the meta- uted to the elevated risk. Since protein can the analyses. Selection bias was mini-
carpal cortical area was greater for women increase urinary calcium39 and perhaps mized by eliminating women who re-
in the dairy region of Yugoslavia, the fracture incidence,' we considered the ported taking calcium supplements, since
difference decreased with age, suggesting possibility that dairy protein was respon- the growing public awareness of osteopo-
that higher calcium intakes did not retard sible for the increase in risk of hip rosis may have prompted calcium use
age-related bone loss. fractures. However, in our cohort, higher disproportionately among women who
Numerous clinical studies have dem- protein intake was associated with an believed themselves to be at the highest
onstrated reduced bone loss in the ra- increased risk of forearm fracture, but no risk.
996 American Journal of Public Health June 1997, Vol. 87. No. 6
Milk and Bone Fractures
The results of this study are most 11. Looker AC, Harris TB, Madans JH, Sem- 28. Fehily AM, Coles RJ, Evans WD, Elwood
directly applicable to middle-aged White pos CT. Dietary calcium and hip fracture PC. Factors affecting bone density in
women in the United States. Although it is risk: the NHANES I Epidemiologic Fol- young adults. Am J Cliii Nuttr 1992;56:579-
low-up Study. Osteoporos Iit. 1993;3:177- 586.
possible that therapeutic levels of calcium 184. 29. Johnston CC, Miller JZ, Slemenda CW, et
supplementation may protect against os- 12. Wickham CAC, Walsh K, Cooper C. et al. al. Calcium supplementation and increases
teoporotic fractures at particular bone Dietary calcium, physical activity, and risk in bone mineral density in children. N Enigl
sites and in defined groups of women, it is of hip fracture: a prospective study. BMJ. J Med. 1992;327:82-87.
1989;299:889-892. 30. Barrett-Connor E, Chang JC, Edelstein SL.
unlikely that high consumption of milk or Coffee-associated osteoporosis offset by
other food sources of calcium during 13. Recker RR, Kimmel DB, Hinders S,
Davies KM. Anti-fracture efficacy of cal- daily milk consumption. JAMA. 1994;271:
midlife will confer substantial protective cium in elderly women. J Bone Miner Res. 280-283.
effects against hip or forearm frac- 1994;9(suppl):S 154. 31. Soroko S, Holbrook TL, Edelstein S,
Barrett-Connor E. Lifetime milk consump-
tures. Cl 14. Cummings SR, Nevitt MC, Browner WS, tion and bone mineral density in older
et al. Risk factors for hip fracture in white women. Amn J Public Health. 1994;84:
women. N Engl J Med. 1995;332:767-773. 1319-1322.
15. Hei SL, Slemenda CE, Johnston CC. 32. Sandler RB, Slemenda CW, LaPorte RE, et
Acknowledgments Baseline measurement of bone mass pre- al. Postmenopausal bone density and milk
This study was supported by research grants dicts fracture in white women. Ann1 Intern consumption in childhood and adoles-
AR 41383 and CA 40356 from the National Med. 1989; 111:355-361. cence. Am J Clint Nutr 1985;42:270-274.
Institutes of Health and by the US Department 16. Cummings SR, Black DM, Nevitt MC, et 33. Stracke H. Renner G, Knie G, Leidig G,
of Agriculture, Agricultural Research Service, al. Appendicular bone density and age Minne H, Federlin K. Osteoporosis and
under contract 53 3K06-5- 10. predict hip fractures in women. JAMA. bone metabolic parameters in dependence
1990;263:665-668. upon calcium intake through milk and milk
17. Parfitt AM. Trabecular bone architecture in products. Eur J Cliii Nutr 1993;47:617-
References the pathogenesis and prevention of frac- 622.
1. Riis B, Thomsen K, Christiansen C. Does ture. Am J Med. 1987;82:68-72. 34. Murphy S, Khaw KT, May H, Compston
calcium supplementation prevent post- 18. Salvini S, Hunter DJ, Sampson L, et al. JE. Milk consumption and bone mineral
menopausal bone loss? a double-blind, Food-based validation of a dietary question- density in middle aged and elderly women.
controlled clinical study. N Enigl J Med. naire: the effects of week-to-week variation BMJ. 1994;308:939-941.
1987;316:173-177. in food consumption. Int J Epidemiol. 35. Hu JF, Zhao XH, Jia JB, Parpia B,
2. Ettinger B, Genant HK, Cann CE. Post- 1989; 18:858-867. Campbell TC. Dietary calcium and bone
menopausal bone loss is prevented by 19. Willett WC, Reynolds RD, Cottrell- density among middle-aged and elderly
treatment with low-dosage estrogen with Hoehner S, Sampson L, Browne ML. women in China. Amn J Clin Nutr 1993;58:
calcium. Ann,l Initern Med. 1987; 106:40-45. Validation of a semiquantitative food fre- 219-227.
3. Smith EL, Gilligan C, Smith PE, Sempos 36. Matkovic V, Kostial K, Simonovic I,
CT. Calcium supplementation and bone quency questionnaire: comparison with a Buzina R, Brodarec A, Nordin BE. Bone
loss in middle-aged women. Am J Clini 1-year diet record. J Anm Diet Assoc. status and fracture rates in two regions of
Nutr 1989;50:833-842. 1987;87. Yugoslavia. Amii J Cliti Nutr 1979;32:540-
4. Dawson-Hughes B, Dallal GE, Krall EA, 20. Willett WC. Nutritionial Epidenziology. 549.
Sadowski L, Sahyoun N, Tannenbaum S. A New York, NY: Oxford University Press; 37. Cumming RG. Calcium intake and bone
controlled trial of the effect of calcium 1989. mass: a quantitative review of the evi-
supplementation on bone density in post- 21. Colditz GA, Martin P, Stampfer MJ, et al. dence. Ccalcif Tissue IJut. 1990;47: 194-20 1.
menopausal women. N Enigl J Med. Validation of questionnaire information on 38. Dawson-Hughes B. Calcium supplementa-
1990;323:878-883. risk factors and disease outcomes in a tion and bone loss: a review of controlled
5. Elders PJM, Netelenbos JC, Lips P, et al. prospective cohort study of women. Am J clinical trials. Am J Cliiz Nutr 1991 (suppl);
Calcium supplementation reduces verte- Epidemziol. 1986; 123:894-900. 54:274S-280S.
bral bone loss in perimenopausal women: a 22. Kleinbaum DG, Kupper LL, Morganstein 39. Hu JF, Zhao XH, Parpia B, Campbell TC.
controlled trial in 248 women between 46 H. Epidemiologic research: principles and Dietary intakes and urinary excretion of
and 55 years of age. J Cliti Endocrinol quantitative methods. New York, NY: Van calcium and acids: a cross-sectional study
Metab. 1991 ;73:533-540. Nostrand Reinhold Co; 1982. of women in China. Am J Cliti Nutr
6. Dawson-Hughes B, Jacques P, Shipp C. 23. Cox DR. Regression models and life- 1993;58:398-406.
Dietary calcium and bone loss from the tables. J Royal Stat Soc. 1972;34: 187-220. 40. Abelow BJ, Holford TR, Insogna KL.
spine in healthy postmenopausal women. 24. Recker RR, Heaney RP. The effect of milk Cross-cultural association between dietary
Am J Clinz Nutr 1987;46:685-687. supplements on calcium metabolism, bone animal protein and hip fracture: a hypoth-
7. Reid IR, Ames RW, Evans MC, Gamble metabolism and calcium balance. Am J esis. Calcif Tissuie IJut. 1992;50: 14-18.
GD, Sharpe SJ. The effect of calcium Clini Nitr. 1985;41 :254-263. 41. Feskanich D, Willett WC, Stampfer MJ,
supplementation on bone loss in postmeno- 25. Baran D, Sorensen A, Grimes J, et al. Colditz GA. Protein consumption and bone
pausal women. N Enzgl J Med. 1993;328: fractures in women. Am J Epidemiol.
460-464. Dietary modification with dairy products 1996; 143:472-479.
8. Holbrook TL, Barrett-Connor E, Wingard for preventing vertebral bone loss in 42. Curhan GC, Willett WC, Rimm EB,
DL. Dietary calcium and risk of hip premenopausal women: a three-year pro- Stampfer MJ. A prospective study of
fracture: 14 year prospective population spective study. J Clini Endocrinol Metab. dietary calcium and other nutrients and the
study. Laincet. 1988;2:1046-1049. 1990;70:264-270. risk of symptomatic kidney stones. N Etigl
9. Chapuy MC, Arlot ME, Duboeuf F. et al. 26. Nelson ME, Fisher EC, Dilmanian FA, J Med. 1993;328:833-838.
Vitamin D3 and calcium to prevent hip Dallal GE, Evans WJ. A l-y walking 43. Farmer ME, White LR, Brody JA, Bailey
fractures in elderly women. N Enzgl J Med. program and increased dietary calcium in KR. Race and sex differences in hip
1992;327: 1637-1642. postmenopausal women: effects on bone. fracture incidence. Am J Public Health.
10. Paganini-Hill A, Chao A, Ross RK, Hender- Am J Clin Nutr 1991;53:1304-131 1. 1984;74: 1374-1380.
son BE. Exercise and other factors in the 27. Heaney RP. Nutritional factors in bone 44. Heikinheimo RJ, Inkovaara JA. Harju EJ,
prevention of hip fracture: the Leisure health in elderly subjects: methodological et al. Annual injection of vitamin D and
World Study. Epidemiology. 1991 ;2: and contextual problems. Am J Clini Nutr fractures in aged bones. Calcif Tissuxe Jutt.
16-25. 1989;50: 1182-1189. 1992;51: 105-1 10.
June 1997, Vol. 87. No. 6 American Journal of Public Health 997