Critique 161: The effects of alcohol consumption on the risk of hip fracture — 31 March 2015

Zhang X, Yu Z, Yu M, Qu X.  Alcohol consumption and hip fracture risk.  Osteoporos Int 2015;26:531–542.  DOI 10.1007/s00198-014-2879-y

Authors’ Abstract

Summary:  The present meta-analysis shows that a nonlinear association between alcohol consumption and the risk of hip fracture was observed.  Light alcohol consumption was inversely significantly associated with hip fracture risk, whereas heavy alcohol consumption was associated with an elevated hip fracture risk.

Introduction:  Previous studies examining the association between alcohol consumption and the risk of hip fracture have reported conflicting findings.  Therefore, we conducted a meta-analysis of prospective cohort studies to assess the association between alcohol consumption and the risk of hip fracture.

Methods:  PubMed and EMBASE were searched for prospective cohort studies on the relationship between alcohol consumption and the risk of hip fractures.  Relative risks (RR) with 95 % confidence intervals (CI) were derived using random-effects models throughout the whole analysis.

Results:  Eighteen prospective cohort studies were included with 3,730,424 participants and 26,168 hip fracture cases.  Compared with non-drinkers, the pooled RR of hip fractures for alcohol consumption was 1.03 (95 % CI, 0.91–1.15), with high heterogeneity between studies (P<0.001, I2=72.6%).  A nonlinear relationship between alcohol consumption and the risk of hip fracture was identified (P nonlinearity = 0.003).  Compared with non-drinkers, the pooled RRs of hip fractures were 0.88 (95 % CI, 0.83–0.89) for light alcohol consumption (0.01–12.5 g/day), 1.00 (95 % CI, 0.85–1.14) for moderate alcohol consumption (12.6–49.9 g/day), and 1.71 (95 % CI, 1.41–2.01) for heavy alcohol consumption (≥50 g/day).

Conclusions:  There was no evidence of publication bias.  In conclusion, a nonlinear association between alcohol consumption and the risk of hip fracture was observed in this meta-analysis.  Further, light alcohol consumption was inversely significantly associated with hip fracture risk, whereas heavy alcohol consumption was associated with an elevated hip fracture risk.

Forum Comments

Especially among the elderly, hip fractures and other fractures are common causes of morbidity and especially of disability, and increase the risk of subsequent mortality.  Osteoporosis of bones and instability of gait are generally considered to be key underlying factors.  Because heavy alcohol intake can lead to a decrease in bone mineral density and to problems with balance and stability, the role of alcohol consumption as a contributing factor to the risk of hip fracture has been questioned for many years.

For several decades, many epidemiologic studies have reported that moderate drinking is associated with an increase in bone mineral density (Holbrook & Barrett-Connor, Felson et al, Ganry et al, McLernon et al), although the specific mechanisms are not clear (Jugdaohsingh et al).  Further, many studies have also reported lower risk of hip fracture among light-to-moderate drinkers (Mukamal et al, Berg et al), although the risk is increased among heavy drinkers, especially alcoholics (Yuan et al).

In a previous meta-analysis on alcohol and hip fracture by Berg et al in 2007, the authors stated: “We pooled effect sizes for 2 specific outcomes (hip fracture and bone density) and synthesized data qualitatively for 4 outcomes (non-hip fracture, bone density loss over time, bone response to estrogen replacement, and bone remodeling).  Compared with abstainers, persons consuming from more than 0.5 to 1.0 drinks per day had lower hip fracture risk (relative risk=0.80 [95% confidence interval, 0.71-0.91]), and persons consuming more than 2 drinks per day had higher risk (relative risk=1.39 [95% confidence interval, 1.08-1.79]). A linear relationship existed between femoral neck bone density and alcohol consumption.”

The present meta-analysis was based on data from 18 prospective studies that included 3,730,424 participants, among whom 26,168 hip fractures were recorded.  Studies included subjects from the USA, France, Europe, Sweden, the Netherlands, Australia, Canada, Denmark, and Japan.  The analysis concluded that there was a nonlinear association between alcohol consumption and the risk of hip fracture, with a decrease in risk for subjects reporting light alcohol consumption but an elevated hip fracture risk for heavier drinkers.

Specific comments on the paper:  This was a meta-analysis of a very large number of subjects from prospective studies, with more than 26,000 cases of hip fracture.  The follow-up period of studies ranged from 3 to 30 years.  All of the studies included were adjusted for age, and almost all for most other known risk factors, including smoking, physical activity, and BMI.  Alcohol intake was categorized as none, light (up to 12.5 g/day, or one typical drink by US standards), moderate (from 12.6 to 49 g/day, or up to about 4 to 5 typical drinks), and heavy for those with greater consumption.  Data are not presented to estimate effects if “moderate” drinking was defined as only up to 24 or 36 g/day, levels that would relate more to current guidelines.

Overall, Forum members considered this to be a very good summary on an important topic.  The main results show a decrease in risk of hip fracture for light drinkers, no effect for moderate drinkers, and an increased risk for heavy drinkers.  In sub-analyses by gender, males showed a significantly lower risk of hip fracture for both light and moderate drinking, but a reduced risk among women was seen only for light drinkers.  For both genders, heavy drinkers showed an increase in risk.

There was a reasonable discussion by the authors of associations that have been shown in other studies of the effects of alcohol intake on bone mineral density and various aspects of bone metabolism.  As stated by member Van Velden: “This article gives a balanced view on the influence of alcohol on bone density, and we support the findings.”

Forum member Zhang also considered this to be a nice review paper.  He stated: “The conclusion that there is a J-shaped association between alcohol consumption and risk of fracture appeared to be fair.  In contrast to a J-shaped association between alcohol consumption and risk of fracture, it appears that there was a linear relationship between alcohol consumption and bone density at the femoral neck, up to about 2 drinks/day.  Although biological mechanisms linking alcohol consumption to bone density are not well understood, several studies have found that alcohol consumption may increase the concentration of serum estradiol and liver estrogen receptors (Gavaler)(Chung).”

Reviewer Thelle noted: “My only small concern is that the frequency of osteoporosis differs considerably on a global level, and the studies included in this analysis are mainly from North America and Europe.  I am not sure whether this actually matters with regard to their conclusion.  The paper is otherwise balanced and only confirms that being in the middle is a safe area.”

Reviewer Finkel pointed out that there is always a problem comparing people who drink very moderately with those who consume alcohol heavily.  There are many lifestyle and other differences in these groups, which raises the possibility of confounding, regardless of how many factors are ‘controlled for’ in the analysis.  Bone health is surely not the only difference.  Forum member Svilaas agreed, and added that this study provides further evidence of the adverse effects of heavy drinking.

Differential effects according to type of alcoholic beverage:  Forum member Estruch commented: “After reviewing the paper and others, I also support the findings. However, I think that the manuscript did not remark on the effects of non-alcoholic components of some beverages (mainly wine and some beer) on bone. In fact, postmenopausal wine drinkers have shown lower hip fracture incidence. In a study of 31,785 men and women in Denmark, moderate drinkers of all beverages showed no increase in fracture risk over that of abstainers, but increased risk was seen for heavier drinkers. Overall, subjects who preferred wine or spirits had a reduced risk of hip fracture compared to those who preferred beer, among whom there was an increase in risk (Høldrup et al).

“Similarly, in the Women’s Health Initiative Clinical Trials and Observational Study (involving 115,655 women), those who preferred wine were at lower risk of hip fracture than non-drinkers, past drinkers and those with other alcohol preferences (Kubo et al). Interestingly, this protective effect has been related to the polyphenol content of wine since studies in animals have shown that resveratrol protected against bone loss during immobilization (Habold et al) and may potentially promote osteoblast formation (Rayalam et al). In addition to flavonoids, beer is also rich in phytoestrogens, especially in lignans, B-vitamins and other minor components. Thus, wine and beer may exert a higher protective effect on bone than other alcoholic beverages, suggesting that constituents other than alcohol may contribute to bone health.” Reviewer Lanzmann-Petithory noted that, unfortunately, other studies have not supported protection from beer; in fact, as stated, the large study in Copenhagen by Høldrup et al showed a significantly higher risk of hip fracture among subjects who preferred beer.

Estruch added: “Finally, it should be noted that, up to now, there has been no long-term randomized intervention study that has analyzed the long-term effects of moderate wine or beer intake (with and without alcohol) on bone mineral density and the possible beneficial mechanisms involved. These studies are needed.”

References from Forum review

Berg KM, Kunins HV, Jackson JL, Nahvi S, Chaudhry A, Harris KA Jr, Malik R, Arnsten JH.  Association between alcohol consumption and both osteoporotic fracture and bone density.  Am J Med 2008;121:406-418.  doi: 10.1016/j.amjmed.2007.12.012.

Chung KW.  Effects of chronic ethanol intake on aromatization of androgens and concentration of estrogen and androgen receptors in rat liver.  Toxicology 1990;62:285–295.

Felson DT, Zhang Y, Hannan MT, Kannel WB, Kiel DP.  Alcohol intake and bone mineral density in elderly men and women. The Framingham Study.  Am J Epidemiol 1995;142:485–492.

Ganry O, Baudoin C, Fardellone P.  Effect of alcohol intake on bone mineral density in elderly women: The EPIDOS Study.  Epidémiologie de l’Ostéoporose.  Am J Epidemiol 2000;151:773-780.

Gavaler JS. Oral hormone replacement therapy: factors that influence the estradiol concentrations achieved in a multiracial study population. J Clin Pharmacol 2002;42:137–144.

Habold C, Momken I, Ouadi A, Bekaert V, Brasse D. Effect of prior treatment with resveratrol on density and structure of rat long bones under tail-suspension. J Bone Miner Metab 2011;29:15-22. doi: 10.1007/s00774-010-0187-y.

Holbrook TL, Barrett-Connor E.  A prospective study of alcohol and bone mineral density.  BMJ 1993;306:1506-1509.

Høldrup S, Grønbaek M, Gottschau A, Lauritzen JB, Schroll M.  Alcohol intake, beverage preference, and risk of hip fracture in men and women. Copenhagen Centre for Prospective Population Studies.  Am J Epidemiol 1999;149:993-1001.

Kubo JT, Stefanick ML, Robbins J, et al.  Preference for wine is associated with lower hip fracture incidence in post-menopausal women.  BMC Women’s Health 2013;13:36. doi:10.1186/1472-6874-13-36

Jugdaohsingh R, O’Connell MA, Sripanyakorn S, Powell JJ.  Moderate alcohol consumption and increased bone mineral density: potential ethanol and non-ethanol mechanisms.  Proceedings of the Nutrition Society 2006;65:291–310.

McLernon DJ, Powell JJ, Jugdaohsingh R, Macdonald HM.  Do lifestyle choices explain the effect of alcohol on bone mineral density in women around menopause?  Am J Clin Nutr 2012;95:1261-1269.

Mukamal KJ, Robbins JA, Cauley JA, Kern LM, Siscovick DS.  Alcohol consumption, bone density, and hip fracture among older adults: the cardiovascular health study.  Osteoporos Int 2007;18:593-602.

Rayalam S, Della-Fera MA, Baile CA.  Synergism between resveratrol and other phytochemicals: implications for obesity and osteoporosis.  Mol Nutr Food Res 2011;55:1177-1185. doi: 10.1002/mnfr.201000616.

Yuan Y, Dawson N, Cooper GS, et al.  Effects of Alcohol-Related Disease on Hip Fracture and Mortality: A Retrospective Cohort Study of Hospitalized Medicare Beneficiaries.  Am J Public Health 2001;91:1089–1093.

Forum Summary

Among elderly people, falls leading to hip fracture are a major health problem, leading to severe morbidity and mortality.  Underlying factors that increase the risk of hip and other fractures include osteoporosis and low bone mineral density, as well as an unsteady gait making falls more common.  The role that alcohol consumption among the elderly may relate to hip fracture has been a topic of concern for many decades, but data from a number of prospective studies now suggest that light-to-moderate drinking may actually decrease the risk of such fractures, while heavy drinking may increase the risk.

The present meta-analysis is based on prospective studies that have yielded more than 26,000 incidences of hip fracture.  It concludes that there is a “J-shaped” association between alcohol consumption, especially of wine, and the risk of hip fractures, with lower risk for a reported consumption of about 1 drink/day, no effect from drinking between 1 and 4 to 5 drinks, and an increased risk from heavier drinking.  Part of the “protection” from light drinking is apparently from the effects of alcohol on increasing bone mineral density.

Forum reviewers considered this to be a very well-done meta-analysis that provides balanced information on the risk of this common public health problem.  Noting that the major differences seen in this study were between light drinkers and heavy drinkers, it was commented that there are many differences between people who are moderate in their habits, including moderate drinking, and people who usually consume more than 4 or 5 drinks/day; thus, there is still the possibility of other lifestyle factors confounding this association.  Also, it was noted that most of the subjects in this analysis were North American or European, so applicability to other racial/ethnic groups may be limited.

In summary, this large and very well-done meta-analysis supports a protective effect of light alcohol consumption on the risk of hip fracture, with an increase in bone density from alcohol being a probable important factor.  Data suggest that wine consumption may have the most favorable effect, perhaps indicating that polyphenols and other compounds may also play a role.  The study shows that heavy drinking (an average of 4 to 5 or more drinks/day) is associated with an increase in the risk of hip fracture.

*                     *                      *                      *

Comments on this critique by the International Scientific Forum on Alcohol Research have been provided by the following members:

Giovanni de Gaetano, MD, PhD, Department of Epidemiology and Prevention, IRCCS Istituto Neurologico Mediterraneo NEUROMED, Pozzilli, Italy

R. Curtis Ellison, MD. Section of Preventive Medicine & Epidemiology, Department of Medicine, Boston University School of Medicine, Boston, MA, USA

Ramon Estruch, MD, PhD.  Associate Professor of Medicine, University of Barcelona, Spain

Harvey Finkel, MD, Hematology/Oncology, Boston University Medical Center, Boston, MA, USA

Dominique Lanzmann-Petithory,MD, PhD, Nutrition/Cardiology, Praticien Hospitalier Hôpital Emile Roux, Paris, France

Erik Skovenborg, MD, specialized in family medicine, member of the Scandinavian Medical Alcohol Board, Aarhus, Denmark

Arne Svilaas, MD, PhD, general practice and lipidology, Oslo University Hospital, Oslo, Norway

Dag S. Thelle, MD, PhD, Senior Professor of Cardiovascular Epidemiology and Prevention, University of Gothenburg, Sweden; Senior Professor of Quantitative Medicine at the University of Oslo, Norway

Gordon Troup, MSc, DSc, School of Physics, Monash University, Victoria, Australia

David Van Velden, MD, Dept. of Pathology, Stellenbosch University, Stellenbosch, South Africa

Yuqing Zhang, MD, DSc, Clinical Epidemiology, Boston University School of Medicine, Boston, MA, USA