By Ghazaleh Barghgir, MD, and Nancy Selfridge, MD

Dr. Barghgir is Clinical Skills Facilitator, Department of Clinical Foundations at Ross University School of Medicine, Barbados, West Indies.

Dr. Selfridge is Professor, Department of Clinical Foundations at Ross University School of Medicine, Barbados,
West Indies.

SYNOPSIS: An analysis of data from the Study of Women’s Health Across the Nation cohort did not reveal a significant association between daily dairy intake frequency, femoral and spine bone mineral density loss, and non-traumatic fracture risk among women transitioning to menopause.

SOURCE: Wallace TC, Jun S, Zou P, et al. Dairy intake is not associated with improvements in bone mineral density or risk of fractures across the menopause transition: Data from the Study of Women’s Health Across the Nation. Menopause 2020;27:879-886.

Osteoporosis is a disease characterized by an increase in osteoclastic bone resorption not balanced by osteoblastic bone formation, resulting in low bone mass and an increased risk of fractures. Osteoporosis often is diagnosed via dual-energy X-ray absorptiometry (DEXA) scans and defined as a bone mineral density (BMD) 2.5 standard deviations or more below the norm for healthy young women (T score of less than -2.5 standard deviations). Osteoporosis has been estimated to affect 53.6 million people, representing approximately 54% of U.S. adults age 50 years and older.1

Osteoporosis is more prevalent in postmenopausal women and in the United States (25% to 30% of global prevalence).2 In 2010-2011, there were 131,443 osteoporosis-attributable fractures, resulting in 64,884 acute care admissions and 983,074 acute hospital days. Acute care costs were $1.5 billion, an 18% increase since 2008.3 Risk factors associated with osteoporosis include age, female sex, ethnicity, family history of osteoporosis, smoking, vitamin D or calcium deficiency, high caffeine intake, immobilization, late menarche, early menopause, and being underweight.4 Bone loss accelerates during menopause; thus, prevention tactics aimed at age-related bone loss are important potential interventions.

One potential modifiable risk factor is dietary dairy product intake, since these foods are rich in nutrients (calcium, magnesium, phosphorus, vitamin D, zinc, and protein) beneficial for bone health maintenance.5,6 Currently, the 2015-2020 Dietary Guidelines for Americans (DGA) recommend adults consume three servings/day of fat-free or low-fat dairy, with one serving equivalent to one cup of milk (e.g., one cup of low-fat yogurt or half an ounce of hard cheese).7 No long-term clinical data exist regarding the relationship between bone health and dairy intake for women transitioning into menopause. Wallace et al aimed to help fill this gap by examining dairy intake and bone health outcomes using publicly available data from the Study of Women’s Health Across the Nation (SWAN).

SWAN data collection began in 1996 for 3,302 pre- and perimenopausal women age 42 to 53 years. Criteria for inclusion were having an intact uterus and at least one ovary and no hormone usage within the three months before screening. Five clinical sites in the United States included Oakland, CA; Los Angeles, Boston, Detroit, and Pittsburgh. Enrolled participants were followed annually to gather information on their demographics, clinical data, and anthropomorphic data. There were 2,335 women with complete baseline femoral neck and/or lumbar spine BMD data. Of these, women with osteoporosis; diabetes; unknown menopausal status; and missing information on dairy intake, physical activity, and smoking status were eliminated. Ultimately, 1,955 participants were included in this BMD and fracture analysis. Loss of femoral neck bone density mass over 10 years was calculated using the following formula: ([BMD at visit 10 - BMD at baseline] / BMD at baseline) × 100.

Menopausal status was determined on annual questionnaires and characterized as premenopause (menstrual bleeding within the past three months and no change in bleeding pattern over the last year), early perimenopause (bleeding in the past three months with decreased menstrual regularity over the past year), late perimenopause (no bleeding for the past three to 11 months), and postmenopause (no bleeding in the past 12 months).

Participants missing data on their femoral neck bone density measurement at baseline (n = 7) and for visit 10 (n = 587) and those missing a final menstrual period date (n = 252) were excluded from the analysis, leaving 1,109 women with adequate data for the study cohort analysis.

BMD of the femoral neck and spine was assessed at each annual follow-up visit using DEXA scanning. The occurrence of fractures was entirely self-reported for visits 1-7 but confirmed at visits 7-10 by a review of medical records and radiology reports. The investigators excluded face, toe, and digit fractures not typically seen in osteoporosis and fractures caused by significant trauma (fall from a height > 6 inches, motor vehicle accident, sports activity-related falls, and fractures when struck by a heavy object).

A modified block food frequency questionnaire assessing the average intake of 137 food items was administered at baseline, visit 5, and visit 9 to collect data on eating habits and dairy product consumption. In the analysis, the average number of dairy servings were used, and any missing dietary data in the questionnaires were imputed using the last observation carried forward method. Study subjects then were classified into four groups based on their cumulative average daily dairy intake: < 0.5 servings, 0.5-1.5 servings, 1.5-2.5 servings, and 2.5 servings. Linear trend estimations were used for participants in each dairy-intake group to assess differences in continuous variables, such as weight, body mass index (BMI), age, and BMD. Chi-square tests were similarly applied to the groups to analyze difference in categorical variables. The statistical significance was set at P < 0.01.

The association of dairy intake with 10-year BMD loss rate was analyzed using a general linear model. A Cox proportional hazard model was applied to calculate hazard ratios (HR) for non-traumatic fractures with 95% confidence intervals. Because of the small number of fractures reported in the cohort, HRs between only the < 1.5 dairy serving per day and 1.5 dairy serving per day groups were used. The authors used several different models in the analysis, adjusting for a variety of variables. The fully adjusted models controlled for race, baseline height, age, activity level, smoking status, time-varying weight, menopausal status, alcohol use, calcium supplementation, and caloric intake.

The final analysis yielded some noteworthy associations. Women consuming more dairy at baseline also were more likely to be premenopausal, heavier, taller, nonsmokers, alcohol consumers, and somewhat more physically active. Compared to Chinese, African-American, and Japanese participants, non-Hispanic white individuals were more likely to consume a higher daily intake of dairy. There were no significant differences between the dairy intake groups at baseline for age, BMI, bone density measurements of femoral neck and lumbar spine, use of calcium supplements, or history of fractures.

Despite adjustment for potential confounding variables and sensitivity analysis, the authors did not find significant differences in changes in BMD among the four dairy intake groups. The authors did not observe differences in HRs and relative risks for non-traumatic fractures based on daily dairy intake group in fully adjusted models.


This analysis contributes to the current body of knowledge concerning dairy food intake associations with bone density and fracture risk or prevention. A significant strength of this study was the SWAN cohort data collection was designed to assess changes in BMD and occurrence of fractures and also included data for many potential confounding variables that strengthened the ultimate data analyses. A major weakness of the study was the fact the cohort did not include any Hispanic women.

A significant factor to consider while interpreting results of this study is that dairy intake was low overall among SWAN participants, with 65% of participants consuming < 1.5 servings per day (well below the three servings/day recommended in the DGA) and was notably lower for ethnic groups other than non-Hispanic whites. In fact, only 7% of the SWAN cohort met the DGA dairy intake recommendations. This low overall dairy consumption in the cohort may have been insufficient to affect BMD loss and fracture outcomes. Further, the study did not explore associations according to different types of dairy food consumption (e.g., yogurt vs. cheese vs. milk), only focusing on a derived composite of total dairy intake. This is noteworthy given the varying nutritional content and inflammatory potential in different dairy products.8 Potential bias in data collection would be present in all self-reported data, including food and supplement intake and fracture occurrence before visit 5. Although the investigators adjusted all analyses for confounding variables, residual confounding remains a source of potential bias.

Although the authors found no association between higher dairy intake and reduced risk of bone mineral loss and fracture occurrence in perimenopausal women, physicians can and should continue to recommend adequate dietary calcium intake, which can be accomplished through consumption of green leafy vegetables (spinach, collard greens, and kale), canned fish with bones (salmon, sardines), tofu, edamame, beans, lentils, and, only if their patients can tolerate them, dairy products (milk, fortified soy milk, etc.). Additional high-quality research is needed, assessing larger cohorts with higher daily dairy intake, accounting for early life nutritional and dairy intake adequacy, differentiating types of dairy products consumed (fermented vs. non-fermented), including data on exercise volume/type, and extending investigation into the post-menopausal period to determine the best evidence-based recommendations for optimizing bone density before menopause, preserving bone density, and reducing fracture risk thereafter. 


  1. Wright NC, Looker AC, Saag KG, et al. The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine. J Bone Miner Res 2014;29:2520-2526.
  2. Hiligsmann M, Cornelissen D, Vrijens B, et al. Determinants, consequences and potential solutions to poor adherence to anti-osteoporosis treatment: Results of an expert group meeting organized by the European Society for Clinical and Economic Aspects of Osteoporosis, Osteoarthritis and Musculoskeletal Diseases (ESCEO) and the International Osteoporosis Foundation (IOF). Osteoporos Int 2019;30:2155-2165.
  3. Hopkins RB, Burke N, Von Keyserlingk C, et al. The current economic burden of illness of osteoporosis in Canada. Osteoporos Int 2016;27:3023-3032.
  4. Hyassat D, Alyan T, Jaddou H, Ajlouni KM. Prevalence and risk factors of osteoporosis among Jordanian postmenopausal women attending the National Center for Diabetes, Endocrinology and Genetics in Jordan. Biores Open Access 2017;6:85-93.
  5. Heaney RP. Calcium, dairy products and osteoporosis. J Am Coll Nutr 2000;19(2 Suppl):83S-99S.
  6. Heaney RP. Dairy and bone health. J Am Coll Nutr 2009:28(Suppl 1):82S-90S.
  7. U.S. Department of Health and Human Services, U.S. Department of Agriculture. 2015-2020 Dietary Guidelines for Americans. 8th ed. Washington;2015.
  8. Bordoni A, Danesi F, Dadevet D, et al. Dairy products and inflammation: A review of the clinical evidence. Crit Rev Food Sci Nutr 2017;57:2497-2525.