There is inconsistent evidence regarding the association between moderate alcohol consumption and bone mineral density (BMD). The aim of this study was to describe the associations between total and beverage-specific alcohol intake and bone loss in older men and women.
A total of 862 randomly selected subjects (mean age 63 years, range 51–81, 51% men) were studied at baseline and 2 years later. BMD was assessed by dual-energy X-ray absorptiometry. Beverage specific and total alcohol intake was assessed by food-frequency questionnaire. Falls risk was determined using the short form Physiological Profile Assessment. Incident fractures were ascertained by questionnaire.
Total alcohol intake in men positively predicted change in BMD at the lumbar spine and hip (β=0.008% and 0.006% per year per gram, P<0.05) after adjustment for confounders, but there was no significant association between alcohol intake and change in BMD in women. Lumbar spine BMD at baseline was negatively associated with frequency of spirits/liquor drinking in men (β=−0.01 g/cm2 per category, P=0.045) and was positively associated with frequency of beer drinking (low alcohol) in women (β=0.034 g/cm2 per category, P=0.002). Change in lumbar spine BMD was positively associated with the frequency of red wine drinking in men (β=0.08% per year per class, P=0.046). Neither beverage-specific nor total alcohol intake was associated with falls risk or fracture.
Alcohol intake especially red wine might prevent bone loss in older men but not women, whereas low-alcohol beer may be protective in women and spirits/liquor may be deleterious in men.
Osteoporotic fractures are an important cause of mortality and morbidity for older people (Piirtola et al., 2008). One established risk factor for osteoporotic fracture is low bone mineral density (BMD), but falls risk is also important as are age, gender, race, diet and other lifestyle factors (Kanis et al., 2005; Dontas and Yiannakopoulos, 2007). Alcohol consumption is a modifiable lifestyle factor with potential to influence both falls risk and BMD, but the relationship between alcohol consumption and fractures, BMD or falls risk in older people is currently unclear. Several studies and reviews showed that moderate alcohol intake was associated with higher bone density in postmenopausal women (Felson et al., 1995; Feskanich et al., 1999; Ganry et al., 2000; Rapuri et al., 2000; Jugdaohsingh et al., 2006; Waugh et al., 2009), but there was insufficient evidence for an association between past or high consumption of alcohol and BMD in women aged 40 to 60 years (Waugh et al., 2009), and the evidence was inconsistent in older men (Papaioannou et al., 2009). Longitudinal studies of the elderly have given inconsistent findings; women with higher alcohol consumption were reported to have greater BMD loss in one study (Hannan et al., 2000) but not in another (Knoke and Barrett-Connor, 2003), whereas alcohol intake in older men was associated with higher BMD in one study (Cawthon et al., 2006) but was reported to have no effect in others (Hannan et al., 2000; Knoke and Barrett-Connor, 2003). Drinking in people aged 25–60 years is strongly associated with unintentional falls at home, resulting in admission to hospital or death (Kool et al., 2008), but the effect of alcohol intake on falls risk for older people is uncertain.
Furthermore, for cardiovascular disease, it has been suggested that the effects of alcohol might vary depending on the type of alcoholic beverage consumed (Grønbæk, 2007). To our knowledge, there is only one study examining the association between intakes of different types of alcoholic beverages and BMD in older people (Tucker et al., 2009). This cross-sectional study suggested that high liquor intake was associated with lower BMD in men but reported limited evidence for other beverage types. It also showed that the alcohol-BMD relationship may be different between men and women. Therefore, the aim of this longitudinal study was to describe the associations between beverage-specific and total alcohol intake, bone loss and falls risk in older adults.
Materials and methods
This study was conducted as part of the Tasmania Older Adult Cohort Study (TASOAC), an ongoing, prospective, population-based study in southern Tasmania, primarily aimed at identifying factors associated with the development and progression of osteoarthritis and osteoporosis in older people. Subjects between the age of 50 and 80 years were selected at baseline from March 2002 until August 2004, randomly using computer generated random numbers from the electoral roll in southern Tasmania (population 229 000), a comprehensive population listing (response 57%), with an equal number of men and women. A total of 1099 randomly selected subjects were studied at baseline. They were followed up 2 year later. The sample was stratified by sex to provide equal numbers of men and women. The study was approved by the Southern Tasmanian Health and Medical Human Research Ethics Committee, and written informed consent was obtained from all participants.
Alcohol intake was assessed by a validated dietary questionnaire (The Cancer Council Victoria, Victoria, Australia) (Hodge et al., 2000; Ambrosini et al., 2003) at baseline and follow-up. Participants were asked to recall how often they drink alcoholic beverages, such as beer (low alcohol or full alcohol), wine (white or red), fortified wines/ port/ sherry and spirits/liqueurs in the last 12 months. The consumption frequency of different alcoholic beverage types were collected as following classes or categories, never, less than once a month, 1–3 days per month, 1 day per week, 2 days per week, 3 days per week, 4 days per week, 5 days per week, 6 days per week and every day. Total alcohol intake was asked in glasses: 30 ml spirits/liqueurs/mixed drinks containing spirits=1 glass, 1 can beer=2 glasses, 1 large bottle of beer (750 ml)=4 glasses, 1 bottle wine (750 ml)=6 glasses, 1 bottle of port or sherry (750 ml)=12 glasses. Alcohol intake was calculated as grams per day. The Pearson correlation coefficient was 0.60 between this food frequency questionnaire and 7-day weighed food records (Hodge et al., 2000).
Bone mass was measured using dual-energy X-ray absorptiometry at the lumbar spine and total hip at baseline and follow-up using the same Hologic Delphi densitometer on array setting (Hologic, Inc., Waltham, MA, USA). The software program was not altered during the study timeframe and a spine phantom was performed to assess any machine drift daily. BMD was examined as areal BMD (g/cm2), which is calculated by dividing the bone mineral content by the area measured. The coefficient of variation is 2% in our hands. Change in BMD was calculated as following: percentage change per annum=(100 × (follow-up BMD−baseline BMD)/ baseline BMD/ time between two scans in years).
Falls risk and fracture
The short form Physiological Profile Assessment (Prince of Wales Medical Research Institute, Sydney, Australia) was used to assess falls risk (Lord et al., 2003). Based on results of five physiological domains (vision, reaction time, proprioception (Lord et al., 2003), strength and balance), the Physiological Profile Assessment uses a discriminant function to compute a falls risk score (standardised score, Z-score) for each individual. Falls risk scores below zero indicate a low risk, scores between 0 and 1 a mild risk, scores between 1 and 2 a moderate risk and scores above 2 indicate a high risk of falling. Information on fracture such as date of occurrence and site was collected by investigators when they interviewed the subjects at baseline and follow-up. Only incident fractures occurring between baseline and follow-up were included in this analysis.
Height was measured to the nearest 0.1 cm (removed headgear and shoes) using a stadiometer. Weight was measured to the nearest 0.1 kg (removed bulky clothing and shoes) using a single pair of electronic scales (Delta Model 707; Seca, Hamburg, Germany) that were calibrated using a known weight at the beginning of each clinic. Body mass index (BMI; weight (kg)/height (m2)) was calculated. Calcium intake was assessed by validated food-frequency questionnaire (as above). Physical activity level was collected with pedometers (HJ-002, Omron Co. Ltd, Tokyo, Japan; Digi-walker SW-200, Yamax, Tokyo, Japan) (Foley et al., 2010). Participants were instructed to wear the pedometer on their dominant side for seven consecutive days except when bathing, sleeping, or performing water-based activities, and follow their normal daily routine. Information on self-reported current smoking status and medication usage was collected by questionnaire.
We initially compared characteristics between subjects who drank more or less than 20 g/day in men and 10 g/day in women based on the median intake and recent Australian guidelines (for healthy men and women, drinking no more than two standard drinks (one Australian standard drink contains 10 g of alcohol) on any day reduces the lifetime risk of harm from alcohol-related disease or injury). T-tests or Mann–Whitney U-tests (where appropriate) were used to compare means. Drinks were classified according to amount of total alcohol intake. Where there were few subjects (<20) in a category of total alcohol intake or frequency of beverage intake, the categories were collapsed. Mean alcohol intake measured at baseline and follow-up was used for analysis on changes of BMD and falls risk.
Univariable and multivariable linear regression with robust standard errors were used to examine the associations between beverage-specific and total alcohol intake and BMD and falls risk at baseline and their percentage changes. These models were adjusted for age, BMI, physical activity, medication, calcium intake and smoking status. Adjusted least-squares means were compared with analysis of covariance between total alcohol drinks or beverage intake categories using post hoc t-tests, with Tukey adjustment for multiple comparisons. Logistic regression was used to assess association between beverage-specific alcohol intakes and fracture. A P-value less than 0.05 (two-tailed) or a 95% confidence interval not including the null point was regarded as statistically significant. All statistical analyses were performed on SAS 9.1 for Windows (SAS Institute, Inc., Cary, NC, USA).
A total of 1099 randomly selected subjects (mean 63 years, range 51–81, 51% men) were studied at baseline, 862 subjects of those were followed up 2.6 years later. Subjects who were lost to follow-up were older (62.6 vs 64.5 years), shorter (167.4 vs 165.3 cm), had higher BMI (27.7 vs 28.5 kg/m2), lower hip BMD (0.97 vs 0.94 g/cm2), were less likely to smoke (8.6 vs 3.4%), and had lower alcohol intake (8.0 vs 4.5 g/day) (all P<0.05).
Average alcohol intake at baseline was 20.9 g/day (median 14.3 g/day, 0–130.5 g/day) for men and 9.9 g/day (median 3.9 g/day, 0–78.3 g/day) for women. Total alcohol intake in eight men was more than 80 g/day but excluding these did not change results so they were retained. Average alcohol intake at follow-up was 19.9 g/day (median 13.2 g/day, 0–110.5 g/day) for men and 9.2 g/day (median 3.6 g/day, 0–78.4 g/day) for women. There were no significant changes in alcohol intake between baseline and follow-up in either sex (P=0.160 in men, P=0.055 in women). Participants’ baseline characteristics are given in Table 1, by sex and by alcohol intake. Men who drank more than 20 g/day of alcohol were heavier and had higher BMI than those who drank less. Women who drank less than 10g/day alcohol were heavier and less physically active than those who drank more, all P<0.05. No other factors differed significantly.
Baseline cross sectional analysis
Total alcohol intake (either grams or drinks per day) was not associated with baseline BMD at the lumbar spine and hip, or baseline falls risk in either men or women (all P>0.05), (Table 2, data not shown for women).
At baseline, the frequency of drinking spirits/liquor was negatively associated with spinal BMD in men (β=−0.01g/cm2 per category, P=0.041) after adjustment for potential factors (Table 3), whereas in women, the frequency of drinking low alcohol beer was positively associated with BMD at the lumbar spine (β=0.034 g/cm2 per category, P=0.002). There were no associations between the baseline frequency of drinking alcoholic beverage types and BMD at hip in either sex (data not shown).
In univariable analysis, alcohol intake (grams per day) was positively associated with percentage change in BMD at the lumbar spine (P=0.036) and hip (P=0.019) in men, and this remained significant after adjustment for potential confounders (age, BMI, medication, physical activity, calcium intake and smoking), BMD at the lumbar spine changed 0.008% per year per gram alcohol intake, P=0.02, whereas BMD at the hip changed 0.006% per year per gram alcohol intake, P=0.04. Similar results were found with drinks of total alcohol (P<0.05 for trend). In women, there were no significant associations between alcohol intake and percentage change of BMD or falls risk (Table 2).
Longitudinally, the frequency of drinking red wine was positively associated with percentage change in BMD at the lumbar spine in men (β=0.08% per year per class, P=0.048) after adjustment for potential factors (Table 3). There were no associations between the frequency of drinking alcoholic beverage types and BMD at hip in either sex (data not shown).
A total of 52 subjects reported fractures during 2.6 years of follow-up, of which five were vertebral fractures. There was no significant association between fracture incidence and alcohol intake at cutoff points of 20 g/day for men and 10 g/day for women (Table 4), nor when alcohol intake was considered as a continuous variable.
This longitudinal study suggests that total alcohol intake and red wine intake may protect against bone loss in older men, but neither alcohol intake nor the type of alcoholic beverage consumed affects bone loss in older women. The cross-sectional analysis supports earlier reports of a negative association between BMD and spirits/liquor intake in men, and suggests a beneficial effect of low alcohol beer intake and BMD in women. Incident fractures and change in falls risk were not associated with either beverage-specific or total alcohol consumed.
Many studies have described a U-shaped or J-shaped curve between alcohol consumption and a number of diseases (de Lorimier, 2000) but this is not clear for bone. Initial reports suggested this association was U-shaped, but the differences were not significant (Burger et al., 1998; Hannan et al., 2000). Recently, an inverted U-shape curve was showed in a cross-sectional study in old men, but not in women (Tucker et al., 2009), whereas a meta analysis has described a linear relationship between femoral neck bone density and alcohol consumption (Berg et al., 2008). In this longitudinal study, we found there was a protective linear relationship between BMD and alcohol consumption (especially red wine) in men. This was consistent with Cardiovascular Health Study, which suggested a positive relation between moderate alcohol intake and BMD at the hip among old adults aged >65 years (Mukamal et al., 2007; Tucker et al., 2009). A cross-sectional study also demonstrated that alcohol intake was linearly related to BMD at the hip and spine in old men aged >65 years (Cawthon et al., 2006) and drinking alcohol was beneficial to the bone health of men and possibly postmenopausal women (Wosje and Kalkwarf, 2007). The mechanisms for the protective effects of alcohol intake on BMD are less clear. Bone-enhancing effects of alcohol might possibly be due to a promotion of endogenous oestrogen synthesis (Mühlbauer et al., 2003; Wosje and Kalkwarf, 2007), especially in those with low circulating oestrogen concentrations, or an ethanol specific effect of moderate alcohol ingestion might directly suppress bone resorption by reducing collagen breakdown, (Sripanyakorn et al., 2009) which might explain why we observed significant associations over time in men but not women.
We report, for the first time from a longitudinal study, a protective association between red wine intake and spine BMD in older men. Wine contains components to enhance health, such as phenols like flavanoids, polyphenols, non-flavanoid phenols, and tannins (de Lorimier, 2000). It's well known that wine, especially red wine, contains high levels of phenolic compounds that favourably influence multiple biochemical systems (de Lorimier, 2000). An experimental study demonstrated that freeze-dried residue from red wine significantly inhibited bone resorption in rats (Mühlbauer et al., 2003) and it has been suggested that phytochemicals (for example, polyphenols) from alcoholic beverages may influence bone health (Jugdaohsingh et al., 2006).
At baseline, total alcohol intake was not associated with bone mass but there were deleterious significant associations between frequency of drinking spirits/liquor and spine BMD in men, and positive associations between low alcohol beer intake and spine BMD in women. The first of these is consistent with an American study reporting that high liquor intakes (>2 drinks/day) in men were associated with significantly lower BMD (Tucker et al., 2009). This highlights the possible importance of non-ethanol components in alcoholic beverages playing a positive role in bone health (Jugdaohsingh et al., 2006). The latter result in women is consistent with a recent report that phalangeal bone density was higher for beer drinkers compared with non-beer drinkers (Pedrera-Zamorano et al., 2009). Tucker found that hip BMD for men was greater when they consumed 1–2 drinks/day of beer compared with non-drinkers (Tucker et al., 2009), but we did not observe this in our study. A protective effect of beer might be due to silicon, as beer is a major dietary source of silicon, which might have a measurable role in promoting bone formation (Jugdaohsingh et al., 2006). However, in Denmark, the risk of hip fracture increased 46% when beer was the preferred drink, which might reflect an increased falls risk (Høidrup et al., 1999).
In this study, cross-sectional and prospective results were inconsistent on effects of different beverages on BMD. Although they were related to each other, there were many comparisons made in this study, thus it is possible that the inconsistent associations may be due to chance. It is also possible that people who drink specific beverage types may differ in other behaviours that have not been picked up by our extensive methods of lifestyle assessment. The variation in cross-sectional and prospective results is also somewhat surprising and difficult to explain unless longer term exposure is more important for some beverages, which would be more likely to give cross-sectional associations (reflecting consistent exposure over time) but not longitudinal associations as we have reported for physical activity and bone mass (Foley et al., 2010). As outlined above, there is some consistency between the results of this study and the literature but it is important to confirm or refute these results in other studies before accepting them as biological.
Falls risk was assessed in the morning when one would expect that alcohol levels would be least. This might, in part, explain the lack of any significant association between alcohol intake and falls risk, unlike the association with significant falls reported in younger adults (Kool et al., 2008). Although it is important to show that there is no effect outside the period of consumption and that assessment of falls risk during consumption would be desirable in future research.
The strengths of our study include its careful assessment of many factors in a population-based setting. Our study also has a number of potential limitations. First, we only have data on frequency of consumption of different alcoholic beverages but not volume (although this is likely to be correlated with volume). Although they were related to each other, there were many comparisons made in this study, thus it is possible that the inconsistent associations may be due to chance. Second, it is most likely the increasing lumbar spine BMD in men is due to spinal osteoarthritis as we have previously reported (Jones et al., 1995), but we did not have assessment of radiographic osteoarthritis to confirm this. Third, subjects who were excluded due to inadequate or missing information appeared also to be older with a lower prevalence of smoking and lower alcohol intake, but the significant association persisted after adjustment for these variables, which suggests that the associations we report in this sample may be generalisable to other older people. Finally, there were only 52 subjects who reported fractures during 2.6 years of follow-up, and this limits our ability to draw conclusions regarding the potential effects of alcohol intake on risk of fracture.
Alcohol intake especially red wine might prevent bone loss in older men but not women, whereas low-alcohol beer in women may be protective and spirits/liquor may be deleterious in men.
Ambrosini GL, van Roosbroeck SA, Mackerras D, Fritschi L, de Klerk NH, Musk AW (2003). The reliability of ten-year dietary recall: implications for cancer research. J Nutr 133, 2663–2668.
Berg KM, Kunins HV, Jackson JL, Nahvi S, Chaudhry A, Harris Jr KA et al. (2008). Association between alcohol consumption and both osteoporotic fracture and bone density. Am J Med 121, 406–418.
Burger H, de Laet CE, van Daele PL, Weel AE, Witteman JC, Hofman A et al. (1998). Risk factors for increased bone loss in an elderly population: the Rotterdam Study. Am J Epidemiol 147, 871–879.
Cawthon PM, Harrison SL, Barrett-Connor E, Fink HA, Cauley JA, Lewis CE et al. (2006). Alcohol intake and its relationship with bone mineral density, falls, and fracture risk in older men. J Am Geriatr Soc 54, 1649–1657.
de Lorimier AA (2000). Alcohol, wine, and health. Am J Surg 180, 357–361.
Dontas IA, Yiannakopoulos CK (2007). Risk factors and prevention of osteoporosis-related fractures. J Musculoskelet Neuronal Interact 7, 268–272.
Felson DT, Zhang Y, Hannan MT, Kannel WB, Kiel DP (1995). Alcohol intake and bone mineral density in elderly men and women. The Framingham Study. Am JEpidemiol 142, 485–492.
Feskanich D, Korrick SA, Greenspan SL, Rosen HN, Colditz GA (1999). Moderate alcohol consumption and bone density among postmenopausal women. J Womens Health 8, 65–73.
Foley S, Quinn S, Jones G (2010). Pedometer determined ambulatory activity and bone mass: a population-based longitudinal study in older adults. Osteoporos Int 21, 1809–1816.
Ganry O, Baudoin C, Fardellone P (2000). Effect of alcohol intake on bone mineral density in elderly women: The EPIDOS Study. Epidemiologie de l’Osteoporose. Am J Epidemiol 151, 773–780.
Grønbæk M (2007). Confounders of the relation between type of alcohol and cardiovascular disease. Ann Epidemiol 17, S13–S15.
Hannan MT, Felson DT, Dawson-Hughes B, Tucker KL, Cupples LA, Wilson PW et al. (2000). Risk factors for longitudinal bone loss in elderly men and women: the Framingham Osteoporosis Study. J Bone Miner Res 15, 710–720.
Hodge A, Giles GG, Patterson A, Brown W, Ireland P (2000). The Anti-Cancer Council of Victoria FFQ. Relative validity of nutrient intakes compared with diet diaries in young of middle-aged women in a study of iron supplementation. Aust N Z J Public Health 24, 576–583.
Høidrup S, Grønbaek M, Gottschau A, Lauritzen JB, Schroll M (1999). Alcohol intake, beverage preference, and risk of hip fracture in men and women. Copenhagen Centre for Prospective Population Studies. Am J Epidemiol 149, 993–1001.
Jones G, Nguyen T, Sambrook PN, Kelly PJ, Eisman JA (1995). A longitudinal study of the effect of spinal degenerative disease on bone density in the elderly. J Rheumatol 22, 921–925.
Jugdaohsingh R, O’Connell MA, Sripanyakorn S, Powell JJ (2006). Moderate alcohol consumption and increased bone mineral density: potential ethanol and non-ethanol mechanisms. Proc Nutr Soc 65, 291–310.
Kanis JA, Borgstrom F, De Laet C, Johansson H, Johnell O, Jonsson B et al. (2005). Assessment of fracture risk. Osteoporos Int 16, 581–589.
Knoke JD, Barrett-Connor E (2003). Weight loss: a determinant of hip bone loss in older men and women. The Rancho Bernardo Study. Am J Epidemiol 158, 1132–1138.
Kool B, Ameratunga S, Robinson E, Crengle S, Jackson R (2008). The contribution of alcohol to falls at home among working-aged adults. Alcohol 42, 383–388.
Lord SR, Menz HB, Tiedemann A (2003). A physiological profile approach to falls risk assessment and prevention. Phys Ther 83, 237–252.
Mühlbauer RC, Lozano A, Reinli A, Wetli H (2003). Various selected vegetables, fruits, mushrooms and red wine residue inhibit bone resorption in rats. J Nutr 133, 3592–3597.
Mukamal KJ, Robbins JA, Cauley JA, Kern LM, Siscovick DS (2007). Alcohol consumption, bone density, and hip fracture among older adults: the cardiovascular health study. Osteoporos Int 18, 593–602.
Papaioannou A, Kennedy CC, Cranney A, Hawker G, Brown JP, Kaiser SM et al. (2009). Risk factors for low BMD in healthy men age 50 years or older: a systematic review. Osteoporos Int 20, 507–518.
Pedrera-Zamorano JD, Lavado-Garcia JM, Roncero-Martin R, Calderon-Garcia JF, Rodriguez-Dominguez T, Canal-Macias ML (2009). Effect of beer drinking on ultrasound bone mass in women. Nutrition 25, 1057–1063.
Piirtola M, Vahlberg T, Löppönen M, Räihä I, Isoaho R, Kivelä SL (2008). Fractures as predictors of excess mortality in the aged-a population-based study with a 12-year follow-up. Eur J Epidemiol 23, 747–755.
Rapuri PB, Gallagher JC, Balhorn KE, Ryschon KL (2000). Alcohol intake and bone metabolism in elderly women. Am J clin nutr 72, 1206–1213.
Sripanyakorn S, Jugdaohsingh R, Mander A, Davidson SL, Thompson RP, Powell JJ (2009). Moderate ingestion of alcohol is associated with acute ethanol-induced suppression of circulating CTX in a PTH-independent fashion. J Bone Miner Res 24, 1380–1388.
Tucker KL, Jugdaohsingh R, Powell JJ, Qiao N, Hannan MT, Sripanyakorn S et al. (2009). Effects of beer, wine, and liquor intakes on bone mineral density in older men and women. Am J Clin Nutr 89, 1188–1196.
Waugh EJ, Lam MA, Hawker GA, McGowan J, Papaioannou A, Cheung AM, Perimenopause BMD Guidelines Subcommittee of Osteoporosis Canada et al. (2009). Risk factors for low bone mass in healthy 40–60 year old women: a systematic review of the literature. Osteoporos Int 20, 1–21.
Wosje KS, Kalkwarf HJ (2007). Bone density in relation to alcohol intake among men and women in the United States. Osteoporos Int 18, 391–400.
This work was supported by the National Health and Medical Research Council of Australia, Tasmanian Government and Royal Hobart Hospital Acute Care Program.
The authors declare no conflict of interest.
About this article
Cite this article
Yin, J., Winzenberg, T., Quinn, S. et al. Beverage-specific alcohol intake and bone loss in older men and women: a longitudinal study. Eur J Clin Nutr 65, 526–532 (2011). https://doi.org/10.1038/ejcn.2011.9
- bone mineral density
- alcoholic beverages
- longitudinal studies