Dietary Patterns and Osteoporosis Risk in Postmenopausal Korean Women

Article information

Osong Public Health Res Perspect. 2012;3(4):199-205
Division of Epidemiology and Health Index, Korea National Institute of Health, Osong, Korea.
*Corresponding author. E-mail:
Received 2012 September 25; Revised 2012 October 15; Accepted 2012 October 15.



The prevalence of osteoporosis and related fractures has increased rapidly in Korean women. Proper nutrition intake is associated with the prevention of osteoporosis. We analyzed the association between dietary patterns and the risk of osteoporosis during a 4-year follow-up in postmenopausal Korean women.


Postmenopausal women (n = 1,725) who participated in the Korean Genome and Epidemiology Study were enrolled. Food intake was assessed using a validated semiquantitative food frequency questionnaire, and a quantitative ultrasound device was used to measure the speed of sound at the radius and tibia.


Three major dietary patterns were identified using factor analysis based on baseline intake data: traditional (high intake of rice, kimchi, and vegetables), dairy (high intake of milk, dairy products, and green tea), and western (high intake of sugar, fat, and bread). Multivariate Cox proportional hazards models were used to estimate relative risk for osteoporosis. An inverse association was detected between the dairy dietary pattern and the osteoporosis incidence [relative risk (RR): 0.63, 95% confidence interval (CI): 0.42–0.93, p-trend=0.055 in radius; RR: 0.56, 95% CI: 0.35–0.90, p-trend=0.048 in tibia]. Individuals in the highest quintile for the traditional dietary pattern (p-trend = 0.009 in tibia) and western dietary pattern (p-trend = 0.043 in radius) demonstrated a higher risk of osteoporosis incidence than those in the lowest quintile.


These results suggested that high consumption of milk, dairy products, and green tea may reduce the risk of osteoporosis in postmenopausal Korean women.

1. Introduction

Osteoporosis is a skeletal disorder characterized by compromised bone strength that predisposes the affected individual to an increased risk of fracture [1]. South Korea has already entered the phase of an aging society, and the prevalence of osteoporosis and related fractures is expected to increase rapidly. According to the 2010 Korean National Health and Nutrition Examination Survey (KHANES), the prevalence of osteoporosis was 34.9% in females and 7.8% in males older than 50 years [2]. Nutrients such as calcium, phosphorus, protein, and vitamin D have a relationship with bone mass [3,4]. Proper nutrient intake has a crucial role in both prevention and treatment of osteoporosis [5]. The association of individual nutrient intake with disease outcomes can be difficult to detect, because nutrients are not consumed independently and they also act synergistically with one another in the body [6,7]. To overcome the limitations of studying individual nutrient intake, dietary pattern analysis has emerged as a method of assessing total food consumption [8]. Examining dietary patterns rather than isolated nutrients or foods offers an advantage of studying the effect of the whole diet [9].

Many studies that have used factor or principal component analysis have empirically derived a “healthy” or “prudent” food pattern that is high in fruits, vegetables, fish, low-fat foods, fiber, legumes, etc. These food patterns are inversely associated with the risk for metabolic syndrome [10] and cancer [11-13]. Several studies have investigated the association of dietary patterns with bone mineral density (BMD), fracture, and osteoporosis [14-19]. However, the relation of these patterns and osteoporosis is not well understood in Korean women.

Thus, the aim of the current study was to examine, using prospective cohort study data, whether dietary patterns are related to risk of osteoporosis in postmenopausal Korean women.

2. Materials and Methods

2.1. Participants

The Korean Genome and Epidemiology Study (KoGES) is a longitudinal cohort study that began in 2001 with the aim to study genetic and environmental risk factors for chronic diseases. Korea Centers for Disease Control and Prevention constructed a cohort and performed follow-up studies. As a part of KoGES, the community-based cohort recruited participants from a rural area (Ansung) and an urban area (Ansan). The baseline survey was conducted from May 2001 to February 2003 and comprised 10,038 participants aged 40–69 years. Among those participants, 1725 postmenopausal women completed the food frequency questionnaire at baseline, and the speed of sound (SOS, m/s) at their radius and tibia was measured at baseline, 2-year, and 4-year follow-ups. A total of 261 participants were excluded for the following reasons: reported daily energy intake being lower or higher than 2 standard deviations (SDs) (<347.0 kcal, n = 3 or >3204.6 kcal, n = 63) from the mean value; early menopause prior to 40 years of age (n = 145); on insulin therapy (n = 9); on hormone therapy (n = 19); or on thyroid drug use (n = 22). Finally, we used 1464 participants’ data in the final analysis. Informed consent was obtained from all study participants. The study protocol was approved by the Korea Centers for Disease Control and Prevention Institutional Review Board.

2.2. Measurement of covariates

Data on general characteristics, socioeconomic status, and lifestyle were collected using an intervieweradministered questionnaire. Data on baseline age, age at menarche, age at menopause, marital status (married or other), educational level (graduate elementary school, middle or high school, or college graduate), income status (under $1000, $1000–2000, $2000–3,000, or more than $3000 per month), passive smoking status (yes or no), and exercise (yes or no) were collected.

2.3. Anthropometry and SOS measurement

Height and weight were measured to the nearest 0.1 kg and 0.1 cm, respectively, by trained staff using a scale and a wall-mounted extensometer. Body mass index (BMI) was calculated as the weight in kilograms (kg)/height in square meters (m2). The quantitative ultrasound measurement was used to measure the SOS (Omnisense 7000s, Sunlight Medical Ltd, Petah Tivka, Israel). Measurement sites were at mid-tibia and radius shafts of the nondominant arm and leg. Measurements were performed three times, and the average of the three measurements was used as the final value. Based on the WHO criteria [20], the osteoporosis incident case was that who had normal T-score at baseline and subsequently diagnosed with osteoporosis at primary or secondary follow-up.

2.4. Dietary assessment and food grouping

Usual dietary intake was assessed using a 103-fooditem, semiquantitative food frequency questionnaire (SQFFQ) developed for KoGES. The development procedure of SQFFQ was described in detail elsewhere [21]. Briefly, based on the dietary intake data from the KHANES in 1998, the frequency of servings was classified into nine categories (never or seldom, once a month, 2–3 times a month, 1–2 times a week, 3–4 times a week, 5–6 times a week, once a day, twice a day, and 3 times or more a day). The portion size of each food item was classified into three categories: small, medium, and large. For the food items with different seasonal availability, we requested the participants to mark for how long they ate the food among the following four categories: 3, 6, 9, or 12 months. This SQFFQ has been validated using 12-day diet record data of 124 participants [22]. Nutrient intakes were calculated for each participant using the seventh-edition Food Composition Tables of the Korean Nutrition Society [23].

2.5. Statistical analyses

Descriptive statistics such as mean and SD were used to summarize continuous variables; frequencies were expressed as percentages. Chi-square tests were used for examining the relationship between categorical variables. Dietary patterns were derived using factor analysis (principal component) based on the 18 food groups from SQFFQ using PROC FACTOR. To reduce the complexity of the data, 103 food items were categorized into 18 groups. In general, the food grouping was based on food and nutrient composition similarity (Table 1). The factors were rotated via an orthogonal transformation (the varimax rotation) to obtain a simpler structure with greater interpretability. To determine the number of factors, we considered eigenvalues (>1) and inspection of scree plots. We decided to retain three major factors for further analyses. All subjects had each dietary pattern score by weighing their intake of each food contributing to that pattern by the relative contribution of those foods. Participants were divided into quintiles by score of each factor, and their general characteristics were compared with first quintile and fifth quintiles. Relative risks (RRs) and 95% confidence intervals (CIs) associated with dietary patterns were estimated using the multivariate Cox proportional hazards regression model. To test for trends, we entered the factor scores into the model as continuous terms. A significant difference was defined as p < 0.05. All statistical analyses were performed using SAS 9.1 (SAS institute Inc., Cary, NC, USA).

Food grouping used in dietary pattern analysisa

3. Results

Factor loadings and variances of each dietary pattern are shown in Table 2. Positive loading indicates that the dietary variable is positively associated with the factor and negative loading indicates an inverse association with the factor. Dietary pattern factor 1, labeled as the traditional dietary pattern, included high factor loadings in rice, kimchi, vegetables, and fruits. The dairy dietary pattern (factor 2) had high factor loadings in milk, dairy products, and green tea. The western dietary pattern (factor 3) was characterized by high factor loadings in noodles, breads, sugar, and fat.

Factor loadings matrix for the three dietary patterns identified from the Food Frequency questionnairea

General characteristics of the participants at baseline study are presented in Table 3 according to quintiles of each food pattern. Mean (±SD) values for continuous variables and proportional distributions for categorical variables are presented. In the dairy dietary pattern, the participants in the highest quintile were younger (p < 0.0001), had more urban residents (p < 0.0001), were in lower socioeconomic status (p < 0.0001), exercised more (p < 0.0001), had fewer passive smokers (p = 0.0002), were taller (p < 0.0001), and showed higher SOS values at radius and tibia (p < 0.0001) and lower osteoporosis prevalence at radius (p = 0.031) and tibia (p = 0.0002), compared to those in the lowest quintile. The highest-quintile participants in traditional and western dietary patterns were younger, were more educated, and had higher income than lowest-quintile participants. However, the prevalence of osteoporosis was not different between the highest- and lowestquintile groups in both dietary patterns.

Baseline general characteristics by quintile (Q) of factor scores for 1464 postmenopausal Korean womena

Incident cases of osteoporosis were 324 (73/1000 person-years) at the radius and 390 (110/1000 personyears) at the tibia. There was an inverse association between dairy dietary pattern and osteoporosis risk (Table 4). Compared with the reference group (Q1), the RR for the fifth quintile was 0.63 (95% CI 0.42–0.93, p for trend = 0.055) at radius, after adjusting for age, residual area, exercise, and passive smoking, and 0.56 (95% CI 0.35–0.90, p for trend = 0.048) at tibia. However, traditional and western dietary patterns increased the risk of osteoporosis incidence. In the traditional dietary pattern, RR for the fifth quintile versus the first quintile was 1.82 at tibia (95% CI 1.12–2.96, p for trend = 0.009). In the western dietary pattern, the RR for fifth quintile versus the reference group (Q1) was 1.46 at radius (95% CI 1.02–2.10, p for trend = 0.043).

Relative risk for osteoporosis during the 4-year follow-up by quintile (Q) of the three major dietary patternsa,b

4. Discussion

The KoGES is a series of large-population prospective cohort studies conducted from 2001. Among them we used the data of a 4-year follow-up study of community-based cohorts. We identified three distinct dietary patterns among 1464 postmenopausal Korean women and found associations between dietary patterns and risk of osteoporosis. The risk for osteoporosis at radius and tibia decreased in the highest dairy dietary pattern group, whereas it increased in the highest traditional and western dietary pattern groups.

Okubo et al [15] examined the association between BMD and dietary patterns derived from factor analysis in premenopausal Japanese farmwomen aged 40–55 years. The healthy dietary pattern (high intake of green and dark yellow vegetables, mushrooms, fish and shellfish, fruits, and processed fish) positively correlated with BMD, and the western dietary pattern (high intake of fats and oils, meat, and processed meat) tended to be inversely associated with BMD.

In the 4-year follow-up study of 877 Japanese elderly, it was found that a meat pattern decreased the risk of fallrelated fracture (HR=0.36, 95% CI=0.13–0.94) after adjusting for confounding variables and a vegetable pattern significantly increased the fracture risk (HR=2.67, 95% CI = 1.03–6.90) [17]. In Scottish postmenopausal women, the “healthy” dietary pattern (high intake of fruits, vegetables, rice, pasta, white meat, oily fish, and dairy products) was associated with decreased bone resorption (r=0.081, p < 0.001). However, the processed foods and snack food patterns were associated with a lower BMD (femoral neck: r=–0.056 and r=–0.044, respectively; p < 0.001) [19].

The dietary pattern in middle-aged Greek women, characterized by high consumption of fish and olive oil and low intake of red meat, was positively associated with lumbar spine BMD [16]. In our previous study, the “healthy and light eating” pattern–higher intake of fruits, vegetables, fish, milk, dairy products, and low-calorie foods–showed a higher SOS and a lower risk for osteoporosis/osteopenia than other dietary patterns [24].

Dairy food is rich in calcium, and it is generally accepted that this food group increases peak bone mass. Higher dairy product consumption was associated with greater hip BMD, and calcium supplementation protected bone loss in a longitudinal study of Caucasians [25]. In the Women’s Health Initiative study, femoral neck BMD and high intake of calcium were positively correlated [26], and calcium- and vitamin D-supplemented individuals had less change in total body BMD and lower risk of hip fracture [4,27]. However, 67.9% of elderly Korean women over 65 years of age consume less calcium than the recommended daily allowance (RDA) [28]. The median calcium intake of our study participants was only 443.6 mg, which is 55% of the 800 mg RDA for elderly Korean women. Therefore, postmenopausal Korean women should try to increase their intake of calcium-rich foods to prevent osteoporosis.

In our study, participants with a dairy dietary pattern also tended to drink more green tea. High consumption of green tea had a protective effect on BMD in Japanese elderly women [29]. Although the underlying mechanism of the effect of green tea on BMD is not clear, flavonoids in green tea are thought to increase BMD. Antioxidants may play an important role in the prevention of oxidative stress-related osteoclastogenesis and bone resorption [30].

Even though food compositions in previous studies were different and direct comparison is difficult to determine the effect of dietary pattern on BMD, the healthy dietary pattern showed beneficial effects for BMD and the western (modified) dietary pattern showed negative effects in these studies. These results consistently demonstrate that women with a high western dietary pattern may have lower bone mass density than those without it [15,17,19]. Energy-dense, nutrient-poor dietary patterns were significantly inversely associated with total body bone mineral content [31].

Tucker et al [14] reported that elderly women with a candy dietary pattern were found to have lower BMD at radius, greater bone loss at the femoral neck, and more hip fractures with a high dietary ratio of animal to vegetable protein intake [32]. Groups that are high in meat intake and low in fruit and vegetable intakes have a negative effect on bone health through increased calcium excretion and bone resorption.

We measured bone characteristics using a quantitative ultrasound method. It measures the SOS value rather than the bone mass density, and therefore the result cannot be directly compared with other reports employing dualenergy X-ray absorptiometry (DEXA). However, Taiwanese people showed a significant linear correlation coefficient (r=0.93) between SOS from ultrasound and BMDfromDEXAat a tibia shaft [33]. Therefore, the SOS value can also be used for the bone risk assessment. The follow-up time of 4 years is rather short. Longer follow-up might clear the association between dietary patterns and the incidence of osteoporosis.

In conclusion, the dairy dietary pattern (consisting of high intake of milk, dairy products, and green tea) decreased the risk for osteoporosis and the traditional (high in rice, kimchi, and vegetables) and western (high intake of sugar, fat, and bread) dietary patterns were associated with greater risk for osteoporosis in postmenopausal Korean women. Our results suggest that high intake of milk, dairy products, and green tea is helpful for preventing osteoporosis in postmenopausal Korean women.


This research was supported by the funds 2001-347- 6111-221, 2002-347-6111-221, 2003-347-6111-221, 2004-347-6111-213, 2005-347-2400-2440-215, 2006- 347-2400-2440-215, and 2009-347-4851-302-210-13 of Korea Centers for Disease Control and Prevention.


1. National Institute of Health. Osteoporosis prevention, diagnosis, and therapy. In : National Institute of Health Consensus Development Conference Statement; 2000.
2. Korea Centers for Disease Control and Prevention. National Health Statistics: Korean National Health & Nutrition Examination Survey 2010. p. 68–69.
3. Reid D, New S. Nutritional influences on bone mass. Proc Nutr Soc 1997;56:977–87. 9483664.
4. Feskanich D, Willett W, Colditz G. Calcium, vitamin D. milk consumption, and hip fracture: a prospective study among postmenopausal women. Am J Clin Nutr 2003;77:504–11. 12540414.
5. Kitchin B, Morgan S. Nutritional considerations in osteoporosis. Curr Opin Rheumatol 2003;15:476–80. 12819478.
6. Kant AK. Dietary patterns and health outcomes. J Am Diet Assoc 2004;104:615–35. 15054348.
7. Newby PK, Tucker KL. Empirically derived eating patterns using factor or cluster analysis: a review. Nutr Rev 2004;62:177–203. 15212319.
8. Hu FB. Pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 2002;13:3–9. 11790957.
9. Jacques PF, Tucker KL. Are dietary patterns useful for understanding the role of diet in chronic disease? Am J Clin Nutr 2001;73:1–2. 11124739.
10. Newby PK, Muller D, Tucker KL. Associations of empirically derived eating patterns with plasma lipid biomarkers: a comparison of factor and cluster analysis method. Am J Clin Nutr 2004;80:759–67. 15321819.
11. Terry P, Hu FB, Hansen H, et al. Prospective study of major dietary patterns and colorectal cancer risk in women. Am J Epidemiol 2001;154:1143–9. 11744520.
12. Butler L, Wang R, Koh W, et al. Prospective study of dietary patterns and colorectal cancer among Singapore Chinese. Br J Cancer 2008;99:1511–6. 18813309.
13. Flood A, Rastogi T, Wirfalt E, et al. Dietary patterns as identified by factor analysis and colorectal cancer among middle-aged Americans. Am J Clin Nutr 2008;88:176–84. 18614739.
14. Tucker KL, Chen H, Hannan MT, et al. Bone mineral density and dietary patterns in older adults: the Framingham Osteoporosis study. Am J Clin Nutr 2002;76:245–52. 12081842.
15. Okubo H, Sasaki S, Horiguchi H, et al. Dietary patterns associated with bone mineral density in premenopausal Japanese farmwomen. Am J Clin Nutr 2006;83:1185–92. 16685064.
16. Kontogianni MD, Melistas L, Yannakoulia M, et al. Association between dietary patterns and indices of bone mass in a sample of Mediterranean women. Nutrition 2008;25:165–71. 18849146.
17. Monma Y, Niu K, Iwasaki K, et al. Dietary patterns associated with fall-related fracture in elderly Japanese: a population based prospective study. BMC Geriatr 2010;10:31 10.1186/1471-2318-10-31.
18. et al for the CaMos Research. Langsetmo L, Hanley DA, Prior JC. Dietary patterns and incident low-trauma fractures in postmenopausal women and men aged 50 y: a population-based cohort study. Am J Clin Nutr 2011;93:192–9. 21068350.
19. Hardcastle AC, Aucott L, Fraser WD, et al. Dietary patterns, bone resorption and bone mineral density in early postmenopausal Scottish women. Eur J Clin Nutr 2011;65(3):378–85. 21179049.
20. World Health Organization. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. World Health Organization. Geneva: 1994.
21. Ahn Y, Lee JE, Paik HY, et al. Development of a semiquantitative food frequency questionnaire based on dietary data from the Korea National Health and Nutrition Examination Survey. Nutr Sci 2003;3:173–84.
22. Ahn Y, Kwon E, Shim JE. Validation and reproducibility of food frequency questionnaire for Korean genomic epidemiologic study. Eur J Clin Nutr 2007;61:1435–41. 17299477.
23. The Korean Nutrition Society. Recommended dietary allowances for Koreans 2000. 7th edth ed.
24. Park SJ, Ahn Y, Kim HM, et al. The association of dietary patterns with bone mineral density in middle-aged women: a cohort of Korean Genome Epidemiology Study. Korean J Community Nutr 2007;12:352–60.
25. McCabe LD, Martin BR, McCabe GP, et al. Diary intakes affect bone density in the elderly. Am J Clin Nutr 2004;80:1066–74. 15447921.
26. MacDonald HM, New SA, Golden MH, et al. Nutritional associations with bone loss during the menopausal transition: evidence of a beneficial effect of calcium, alcohol, and fruit and vegetable nutrients and of a detrimental effect of fatty acid. Am J Clin Nutr 2004;79:155–65. 14684412.
27. Dawson-Hughes B, Harris SS. Calcium intake influences the association of protein intake with rates of bone loss in elderly men and women. Am J Clin Nutr 2002;75:773–9. 11916767.
28. Ministry of Health Welfare and family Affairs. The Third Korea National Health & Nutrition Examination Survey (KNHANES III), 2005–Nutrition Survey(I) 2006. 239.
29. Muraki S, Yamamoto S, Ishibashi H, et al. Diet and lifestyle associated with increased bone mineral density: cross-sectional study of Japanese elderly women at an osteoporosis outpatient clinic. J Orthop Sci 2007;12:317–20. 17657549.
30. Sugiura M, Nakamura M, Ogawa K, et al. Dietary patterns of antioxidant vitamin and carotenoid intake associated with bone mineral density: findings from post-menopausal Japanese female subjects. Osteoporos Int 2011;22:143–52. 20480147.
31. McNaughton SA, Wattanapenpaiboon N, Wark JD, et al. An energy-dense, nutrient-poor dietary pattern is inversely associated with bone health in women. J Nutr 2011;141:1516–23. 21653576.
32. Sellmeyer DE, Stone KL, Sebastian A, et al. A high ratio dietary animal to vegetable protein increases the rate of bone loss and the risk of fracture in postmenopausal women. Study of Osteoporotic Fractures Research Group. Am J Clin Nutr 2001;73:118–22. 11124760.
33. Chen T, Chen PJ, Fung CS, et al. Quantitative assessment of osteoporosis from the tibia shaft by ultrasound techniques. Med Eng Phys 2004;26:141–5. 15036181.

Article information Continued

Table 1.

Food grouping used in dietary pattern analysisa

Food or food group Food items included

Rice and rice cake Cooked well-milled rice, cooked rice with barley, cooked rice with other cereals, parched cereal powder, rice cakes (plain rod shape), other rice cakes
Noodles Ramen, noodles with soup, chajangmyon, buckwheat vermicelli/buckwheat noodle, dumpling
Bread Loaf bread, bread with small red beans, other breads, pizza/fast food, cereals
Sugar and fat Cakes, snacks, candy/chocolate, butter/margarine, coffee, sugar, coffee cream
Potatoes Potatoes, sweet potatoes, starch vermicelli, starch jelly
Legumes and nuts Soybean, tofu, stew with soybean pastes, soybean milk, nuts
Kimchi Kimchi, Korean cabbage, kakduki/small radish kimchi, kimchi with liquid, other kimchi (green onion/kodulbbagi/mustard leaves)
Vegetables Green pepper, red pepper leaves, spinach, lettuce, perilla leaf, leek/water dropwort, green-yellow vegetables, radish/salted radish
Doraji/deoduck (kinds of white root), onion, cabbages, cucumber, bean sprouts, carrot, pumpkin gruel/pumpkin juice, young pumpkin, vegetable juice, bracken/sweet potato stalk, pickles
Mushrooms Oyster mushroom, other mushrooms
Fruits Persimmon, dried persimmon, tangerine, muskmelon/melon, banana, pear, apple/apple juice, orange/orange juice, watermelon, peach/plum, strawberry, grape/grape juice, tomato/tomato juice
Meat Dog meat, chicken/chicken leg/chicken wing, pan-roast pork, pork belly, braised pork, ham/sausage, pan-roast beef ribs, thick beef soup/hard-boiled beef ribs, edible viscera
Eggs Eggs
Fish and seafood Sushi, hair tail, eel, yellow croaker, Alaska pollack, mackerel/Pacific saury/Spanish mackerel, dried anchovy, cuttlefish/octopus, tuna, canned, fish paste/crab flavored, crab, clam (small ark shell/little neck clam/clam meat), oyster, shrimp, salt-fermented fish
Seaweeds Laver, kelp/sea mustard
Milk Milk
Dairy products Yoghurt, ice cream, cheese
Coffee and carbonated drink Coffee, carbonated drinks
Green tea and other drinks Green tea, sikhae, citron tea

aFood lists were from the Semiquantitative Food Frequency questionnaire.

Table 2.

Factor loadings matrix for the three dietary patterns identified from the Food Frequency questionnairea

Factor 1 Factor 2 Factor 3

Traditional Dairy Western

Rice and rice cake 0.21 –0.47
Noodles 0.56
Bread 0.26 0.61
Sugar and fat 0.60
Potatoes 0.42 0.29
Legumes and nuts 0.47
Kimchi 0.52 –0.30
Vegetables 0.71
Mushrooms 0.32 0.43
Fruits 0.51
Meat 0.43 0.35
Eggs 0.40 0.29
Fish and seafood 0.56 0.32 0.30
Seaweeds 0.55 0.34
Milk 0.55
Dairy products 0.50
Coffee and carbonated beverages 0.39
Green tea and other drinks 0.22 0.53
Percentage of variance (%) 19.3 8.0 6.6

aFood groups with absolute values <0.20 are not shown for simplicity.

Table 3.

Baseline general characteristics by quintile (Q) of factor scores for 1464 postmenopausal Korean womena

Traditional Dairy Western

Q1 Q5 p Q1 Q5 p Q1 Q5 p

Mean age (SD), y 59.4 (6.8) 58.2 (6.6) 0.0247 60.3 (6.0) 56.7 (6.8) <0.0001 60.0 (6.4) 56.9 (6.7) <0.0001
Mean age of menarche (SD), y 16.4 (1.7) 16.4 (1.9) 0.9540 16.5 (1.7) 16.1 (1.7) 0.0025 16.2 (1.9) 16.3 (2.0) 0.8215
Mean age of menopause (SD), y 48.7 (4.4) 49.0 (3.9) 0.4226 48.8 (4.2) 48.7 (4.1) 0.7269 48.8 (4.2) 48.6 (4.1) 0.7195
Area, n (%) 0.8237 <0.0001 0.0001
   Rural 228 (49.7) 231 (50.3) 277 (66.3) 141 (33.7) 238 (54.7) 197 (45.3)
   Urban 64 (50.8) 62 (49.2) 15 (9.0) 152 (91.0) 55 (36.7) 95 (63.3)
Marital status, n (%) 0.0450 0.0909 0.1184
   Married 226 (48.0) 245 (52.0) 224 (48.1) 242 (51.9) 230 (48.5) 244 (51.5)
   Others (single, widowed) 65 (58.6) 46 (41.4) 67 (56.8) 51 (43.2) 63 (56.8) 48 (43.2)
Education, n (%) 0.0137 <0.0001 <0.0001
   Elementary school degree 205 (53.3) 180 (46.7) 231 (65.4) 122 (34.6) 207 (58.6) 146 (41.4)
   Middle or high school degree 76 (43.4) 99 (56.6) 56 (27.9) 145 (72.1) 77 (37.8) 127 (62.3)
   College degree or above 4 (25.0) 12 (75.0) 2 (7.4) 25 (92.6) 7 (31.8) 15 (68.2)
Income, n (%) 0.0303 <0.0001 <0.0001
   <$1000 188 (54.8) 155 (45.2) 210 (64.6) 115 (35.4) 202 (59.9) 135 (40.1)
   $1000–2000 62 (45.6) 74 (54.4) 52 (37.7) 86 (62.3) 50 (37.9) 82 (62.1)
   $2000–3000 26 (47.3) 29 (52.7) 14 (22.6) 48 (77.4) 14 (28.6) 35 (71.4)
   >$3000 12 (32.4) 25 (67.6) 8 (18.2) 36 (81.8) 17 (33.3) 34 (66.7)
Exercise, n (%) 0.0124 <0.0001 0.2289
   No 242 (52.6) 218 (47.4) 259 (59.3) 178 (40.7) 232 (51.4) 219 (48.6)
   Yes 50 (40.0) 75 (60.0) 33 (22.3) 115 (77.7) 61 (45.5) 73 (54.5)
Passive smoking, n (%) 0.0798 0.0002 0.6932
   No 177 (54.5) 148 (45.5) 143 (42.6) 193 (57.4) 170 (50.3) 168 (49.7)
   Yes 103 (46.8) 117 (53.2) 124 (59.1) 86 (41.0) 101 (48.6) 107 (51.4)
Height, mean (SD), cm 151.5 (5.2) 152.7 (5.4) 0.0047 151.5 (5.2) 154.3 (5.1) <0.0001 152.3 (5.8) 152.7 (5.3) 0.4299
Weight, mean (SD), kg 56.4 (8.3) 58.4 (8.2) 0.0026 57.9 (8.4) 59.3 (8.8) 0.0335 58.1 (8.6) 58.1 (8.1) 0.9468
BMI, mean (SD), m2/kg 24.5 (3.3) 25.0 (3.0) 0.0710 25.2 (3.2) 24.9 (3.1) 0.3456 25.0 (3.2) 24.9 (3.0) 0.5759
Radius SOS, mean (SD), m/s 4,143.5 (180.0) 4,138.8 (183.8) 0.7532 4,092.1 (177.5) 4,169.2 (180.1) <0.0001 4,135.2 (187.1) 4,141.8 (175.4) 0.6609
Tibia SOS, mean (SD), m/s 3,780.8 (168.7) 3,805.2 (158.9) 0.0722 3,770.2 (155.0) 3,848.3 (156.5) <0.0001 3,796.9 (160.0) 3,813.5 (162.2) 0.2134
Prevalence at radius, n (%) 0.5767 0.0305 0.3196
   Normal 267 (50.3) 264 (49.7) 253 (48.4) 270 (51.6) 263 (49.4) 269 (50.6)
   Osteoporosis 25 (46.3) 29 (53.7) 39 (62.9) 23 (37.1) 30 (56.6) 23 (43.4)
Prevalence at tibia, n (%) 0.0958 0.0002 0.0855
   Normal 198 (47.7) 217 (52.3) 193 (45.2) 234 (54.8) 201 (47.9) 219 (52.1)
   Osteoporosis 94 (55.3) 76 (44.7) 99 (62.7) 59 (37.3) 92 (55.8) 73 (44.2)

aThe factors were standardized continuous variables, and each participant received a score for each factor. N = 292 in Q1, and N = 293 in Q5.

BMI = body mass index; SD = standard deviation; SOS = speed of sound.

Table 4.

Relative risk for osteoporosis during the 4-year follow-up by quintile (Q) of the three major dietary patternsa,b

Q1 Q3 Q5

Case n (%) Case n (%) RR (95% CI) Case n (%) RR (95% CI) p for trend

Radius 56 (21.0) Ref 71 (26.8) 1.37 (0.95–1.99) 69 (26.1) 1.46 (1.00–2.13)c 0.0674
Tibia 67 (33.8) Ref 75 (34.9) 1.32 (0.81–2.16) 91 (41.9) 1.82 (1.12–2.96)c 0.0095
Radius 85 (33.6) Ref 56 (21.4) 0.59 (0.41–0.84)d 53 (19.6) 0.63 (0.42–0.93)d 0.0554
Tibia 83 (43.0) Ref 70 (34.3) 0.62 (0.40–0.96)c 70 (29.9) 0.56 (0.35–0.90)d 0.0483
Radius 64 (24.3) Ref 76 (25.2) 1.18 (0.83–1.67) 68 (25.3) 1.46 (1.02–2.10)c 0.0428
Tibia 60 (30.0) Ref 93 (39.4) 1.30 (0.82–2.05) 81 (37.0) 1.46 (0.91–2.33) 0.0628

aParticipants in Q1, Q3, and Q5 were not diagnosed with osteoporosis at the baseline; bModel adjusted for age, residual area, exercise, and passive smoking;cSignificantly different from Q1: p < 0.05; dSignificantly different from Q1: p < 0.01.