Moderate Alcohol Consumption Suppresses Bone Turnover in Adult Female Rats



Chronic alcohol abuse is a major risk factor for osteoporosis but the effects of moderate drinking on bone metabolism are largely uninvestigated. Here, we studied the long-term dose-response (0, 3, 6, 13, and 35% caloric intake) effects of alcohol on cancellous bone in the proximal tibia of 8-month-old female rats. After 4 months of treatment, all alcohol-consuming groups of rats had decreased bone turnover. The inhibitory effects of alcohol on bone formation were dose dependent. A reduction in osteoclast number occurred at the lowest level of consumption but there were no further reductions with higher levels of consumption. An imbalance between bone formation and bone resorption at higher levels of consumption of alcohol resulted in trabecular thinning. Our observations in rats raise the concern that moderate consumption of alcoholic beverages in humans may reduce bone turnover and potentially have detrimental effects on the skeleton.


ALCOHOLICS OFTEN have radiographic and histomorphometric evidence of osteopenia and a greatly reduced bone mineral density (BMD).(1–3) Histomorphometric analysis of bone biopsy specimens and measurement of biochemical markers of bone metabolism have revealed consistent evidence that alcohol excess inhibits bone formation.(4–13) The effects of ethanol on bone resorption are less certain; increases, decreases, and no change have been reported.(2,4–7,11,13)

The inability to assign a role for bone resorption in mediating alcohol-induced bone loss highlights the difficulties associated with performing studies in alcoholics. Human studies often are difficult to interpret because of the small number of patients who can be studied and wide variations of the patient population in age, duration, and pattern of alcohol abuse and accompanying risk factors. It is difficult to distinguish the direct effects of ethanol from secondary factors such as magnesium and zinc deficiency, reduced mechanical loading of the skeleton caused by decreased physical activity and weight loss, malabsorption caused by chronic pancreatitis, and skeletal abnormalities associated with increased cigarette smoking and increased use of aluminum-containing antacids. The role of abnormal liver function is especially controversial, with some investigators reporting bone loss in alcoholics free of liver disease and others reporting no bone loss.(4,8,12,14–17) It is especially difficult to control for nutrition, in part because alcoholics have a larger caloric intake than their peers but frequently are underweight.(18,19)

A rat model for alcohol abuse has been developed to circumvent the limitations of human studies. Weight and nutrition can be controlled carefully in this animal model. Growing rats who consume ethanol at a rate (adjusted for the difference in body mass) comparable with alcoholics develop osteopenia and other abnormalities in bone and mineral metabolism.(20,21) All of the changes described in the rat model have been reported in alcoholic patients.(2–4,6,7,11–13,15,17)

The skeletal response to low and moderate alcohol consumption is relevant to a much larger segment of the adult population than is alcohol abuse but has not been studied extensively in either humans or laboratory animals. The present investigation was designed to investigate the long-term dose-response effects of ethanol on the skeleton of adult female rats. Specifically, the study was designed to determine the minimum consumption of ethanol required to induce bone loss.


Animal experiment

Female Sprague-Dawley rats (Harlan, Indianapolis, IN, USA) were received at 8 months of age (body weight [BW], 279 ± 3 g; mean ± SE). Subsequent procedures performed on animals were approved by the Mayo Institutional Animal Care and Use Committee in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Rats were weight-matched into seven study groups (n = 9–12) comprised of four ethanol-treated groups (3, 6, 13, and 35% caloric intake), a control group fed ad libitum, a control group pair fed to the 6% ethanol-treated group, and a control group killed at the start of the experiment (baseline). As described for a previous study,(22) rats were housed individually in a temperature- and humidity-controlled animal facility on a 12-h light/dark cycle. During the first week of the study, all animals were acclimated to a modified Lieber-DeCarli liquid diet (BioServe, Frenchtown, NJ, USA). This diet contains 1.3 g/liter calcium and 1.7 g/liter phosphorous. Protein, fat, and carbohydrates contribute 18, 35, and 47% of caloric intake, respectively. The alcohol-supplemented groups were subsequently fed, ad libitum, a liquid diet containing increasing concentrations of ethanol (95% vol/vol) until receiving the appropriate percent of total caloric intake at the end of the first treatment week. Treatment continued for 4 months. Rats were not given water. Control rats did not receive ethanol and were fed a liquid diet isocalorically supplemented with maltose/dextran, as per the manufacturer's instructions. All animal weights were recorded weekly for the last 12 weeks of the study. On the first day of ethanol treatment, all rats were injected with a 20 mg/kg BW dose of a bone fluorochrome (tetracycline-hydrochloride; Sigma, St. Louis, MO, USA) perivascularly at the base of the tail. The baseline control rats were killed 24 h later. The remaining rats were injected (20 mg/kg BW) with the fluorochromes calcein (Sigma) and demeclocycline (Sigma) 9 days and 2 days before death, respectively. Rats were anesthetized using ketamine and xylazine, weighed, and killed by decapitation. Tibiae were harvested, defleshed, and fixed by immersion in 70% ethanol for bone histomorphometry. Wet weights of uteri were recorded.

Cancellous bone histomorphometry

Tibiae were processed for plastic embedding without demineralization and sectioned as described.(22) Histomorphometry of cancellous bone was performed on unstained longitudinal sections (5 μm thick) in a standard sampling site in the proximal tibial metaphysis.(22) Terminology and abbreviations are consistent with the standardized histomorphometric nomenclature.(23)

All measurements were conducted on an image-analysis system employing OsteoMeasure software (OsteoMetrics, Atlanta, GA, USA) as described.(22)

The following measured and derived histomorphometric values were obtained: total cancellous bone area (BA) measured at a magnification of ×10 in a metaphyseal sampling site divided by the total area of the tissue sampling site (2.88 mm2) and expressed as a percentage; cancellous BA was compared between the ad libitum and the pair-fed control groups to determine if there was a significant effect of diet on BA; and trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular separation (Tb.S) were estimated using the method of Parfitt et al.(24)

Fluorochrome-based measurements and derived values consisting of double-labeled perimeter, mineral apposition rate (MAR), and bone formation rate (BFR) were determined as described.(22) The measurements were performed at a magnification of ×10.

The length of cancellous bone perimeter covered by osteoclasts was measured in toluidine blue-stained sections at ×20, divided by total bone perimeter, and expressed as a percentage. Osteoclasts were morphologically distinguished as large, multinucleated cells with a foamy cytoplasm juxtaposed to bone surface.


All values are expressed as means ± SE. Significant differences between alcohol-treated groups and controls were determined by Fisher's protected least significant difference post hoc test for multiple group comparisons after detection of significance by one-way analysis of variance (ANOVA). Significance was considered at values of p < 0.05. Dose-response effects were evaluated by linear regression analysis.


Pair feeding had no effect on any measured value. For this reason, the pair-fed and ad libitum control groups were combined.

The effects of ethanol on BW; change in BW, food, and ethanol consumption; and uterine weight are shown in Table 1. The initial BWs did not differ between the treatment groups. Ethanol had minor effects on final BW and food consumption. The lowest concentration of ethanol (3% caloric intake) increased final BW and rate of change in BW as well as consumption of the diet, whereas the highest concentration (35% of caloric intake) significantly decreased food consumption and rate of change in BW. There was a nearly linear increase in total ethanol consumed per day as the concentration of ethanol was increased in the diet. Ethanol treatment tended to decrease uterine weight; the reductions were significant for the groups with 6% and 35% caloric intakes.

Table Table 1.. Body and Uterine Weights
original image

The effects of ethanol on static bone histomorphometry are summarized in Table 2. BA/tissue area (TA) was significantly decreased in the 13% and 35% caloric intake groups and Tb.Th was decreased for intake levels at and above 6%. Tb.N and Tb.S were not significantly influenced by ethanol.

Table Table 2.. Cancellous Bone Architecture
original image

The effects of ethanol on cancellous bone dynamic and cellular histomorphometry are summarized in Table 3. Ethanol treatment decreased mineralizing perimeter (M.Pm) and BFR (perimeter referent). Ethanol had no effect on MAR. All concentrations of ethanol treatment reduced osteoclast perimeter (Oc.Pm) to a similar magnitude.

Table Table 3.. Fluorochrome and Osteoclast Measurements
original image

We were unable to measure longitudinal bone growth because of inadequate separation between the tetracycline label given at the start of the experiment and the mineralized hypertrophic cartilage.

The dose-response effects of alcohol are summarized in Table 4. Linear regression revealed significant dose-dependent decreases in BA/TA, Tb.Th, M.Pm, and BFR. There was no dose-dependent effect of alcohol on BW, uterine wet weight, Tb.N, Tb.S, MAR, or Oc.Pm.

Table Table 4.. Dose-Response Effects of Ethanol Analyzed by Linear Regression
original image


The observed results are disturbing because we did not observe a no-effect dose for alcohol consumption. Pronounced changes in bone metabolism were observed at the lowest consumption level of alcohol. Dietary intake of alcohol comprising as little as 3% of total calories dramatically reduced histological indices of bone turnover. Higher consumption levels of alcohol resulted in alterations in trabecular architecture and even net cancellous bone loss.

Analysis of the cancellous bone architecture revealed that alcohol-induced bone loss was caused by a reduction in Tb.Th; Tb.N was not altered. Oc.Pm was not increased, indicating that cancellous bone loss was caused by a disturbance in the bone remodeling balance rather than increased bone remodeling. Indeed, the present results indicate that alcohol consumption results in reduced bone remodeling as evidenced by the decreases in histomorphometric indices of bone resorption (Oc.N) and bone formation (M.Pm).

Cancellous bone remodeling occurs when focal resorption (remodeling unit) is initiated on a previously quiescent trabecular surface.(25) A small amount of bone is resorbed and the resulting resorption lacuna is filled shortly thereafter as a result of new bone formation. There are two cellular mechanisms that could lead to trabecular thinning: (1) excessive erosion during the resorption phase and (2) incomplete refilling of the erosion cavity during the formation phase. The maximum inhibition of osteoclast number was achieved at the lowest level of alcohol consumption, whereas M.Pm showed a striking dose-dependent decrease. These findings suggest that incomplete filling of the erosion cavity during the formation phase of the remodeling cycle is the more likely cellular mechanism for trabecular thinning. The MAR was not altered, suggesting that alcohol inhibits onset of the bone formation phase of the remodeling cycle but not its continuation once initiated. These results are consistent with Dyer et al.(26) who concluded that alcohol inhibits osteoblast proliferation and activity in the rat. This observation also is consistent with in vitro studies showing that ethanol delays recruitment of osteoblasts but has little effect on bone matrix protein gene expression and peptide secretion by mature osteoblasts.(27,28) Similarly, ethanol did not increase apoptosis in vitro,(25) suggesting that osteoblast lifespan is unaffected.

At first glance, the observed nondose-dependent reduction in uterine weight in rats fed some doses of ethanol suggests that gonadal insufficiency contributes to the bone changes. However, the histological changes are not consistent with this possibility. Ovariectomy results in greatly elevated bone turnover(29,30) whereas alcohol reduced bone turnover. In addition, the pattern of bone loss differs from ovariectomy. Gonadal insufficiency decreases Tb.N, whereas ethanol resulted in a decrease in Tb.Th.(22) The mechanism for the uterine atrophy is not clear but alcohol has been reported to induce a variety of pathophysiological changes in the uterus, including uterine atrophy.(31) Furthermore, we have identified numerous genes in which expression is altered dramatically after acute administration of 1 mg/kg of ethanol, suggesting that alcohol has direct effects on the uterus (Turner et al., unpublished results, 2000).

Studies performed in growing male(21,32–35) and female(36,37) rats indicated that chronic consumption of large doses of alcohol in the diet suppresses bone growth. The resulting relative osteopenia is caused by the failure to acquire a normal bone mass and is relevant to juvenile alcohol abusers. The present study in older rats is more relevant to adult humans and shows alcohol-induced bone loss in rats. Our failure to detect measurable longitudinal bone growth at the proximal tibial growth plate provides definitive evidence that the observed changes were not influenced by growth. The present study also differs from previous work in that it attempts to model moderate as well as abusive alcohol consumption.

There is no universal agreement as to what constitutes moderate drinking. Additionally, the level of ethanol consumption in rats in this study cannot be related directly to humans because the rates of metabolism differ between the two species. The variables that must be considered and evaluated in these studies include: total ethanol consumed, the percent caloric intake contributed by ethanol, and peak blood levels of ethanol.

The low-dose group (3% caloric intake) consumed approximately 0.4 ml ethanol/day. On a body mass basis, this would be equivalent to approximately three standard drinks by a 50-kg woman, which is on the high end of moderate alcohol consumption. However, relative to caloric intake, this level of consumption would be the equivalent of <0.5 daily drinks, which is on the low end. Blood ethanol levels may be more important than the absolute amount of alcohol consumed. Ethanol administered at 35% of caloric intake resulted in measured blood ethanol levels of 0.09–0.15%.(21,32,33,36) The measured levels are likely to underestimate peak blood alcohol levels because sequential measurements have not been performed throughout the rat feeding cycle. Nevertheless, these levels are near to or exceed the impairment level, which generally is considered to be between 0.08 and 0.10%. This high blood alcohol level contrasts with the 3% caloric intake dose rate, which results in blood alcohol levels below the assay detection limit. Taken as a whole, the data suggest that our dose range in the rat extends from the human equivalent of low-moderate to alcohol abuse.

Bone formation generally is reduced in alcoholics.(3,4,7,17) Additionally, administration of ethanol to healthy volunteers results in an acute decrease in serum osteocalcin levels.(8–10) The similarity between our results and those seen in humans indicates that in addition to being a good model for alcohol abuse in adults, the mature rat also may be predictive for the skeletal effects of moderate drinking.

The implications of a reduction in bone turnover in moderate drinkers may depend on age and other factors. On one hand, ethanol consumption by adolescents might reduce peak bone mass, which would predispose the individual to osteoporosis. On the other hand, a reduction in bone turnover is likely to reduce the risk of osteoporosis in postmenopausal women because these individuals are losing bone at a rapid rate because of, in part, elevated bone turnover. This speculation is supported by epidemiological data indicating that postmenopausal moderate drinkers have a higher bone mass than abstainers.(38–41) It also is supported by studies in ovariectomized rats, which show that ethanol does not accelerate the bone loss associated with gonadal insufficiency and may reduce osteoclast number.(22,42,43)

Neither moderate nor high doses of alcohol prevented bone loss in ovariectomized rats.(22,42,43) Ovariectomy is likely to result in a more extreme depletion of gonadal hormones than menopause. Epidemiological studies suggest that estrogen replacement accentuates the putative beneficial skeletal response to alcohol in postmenopausal women.(41) Thus, it is possible that the combined antiremodeling actions of estrogen replacement and alcohol have additive effects in women and rats.

In summary, these studies in rats show that alcohol consumption results in dose-dependent bone loss and decreased bone turnover. A no-effect dose for alcohol was not observed. The findings in rats suggest that even moderate levels of alcoholic beverage consumption in humans may reduce bone turnover and potentially have detrimental effects on the skeleton.


The authors thank Ms. Lori Rolbiecki for typing this manuscript and Ms. Peggy Backup for editorial assistance. The authors also thank Mr. Charles Rowland for his biostatistical analysis. These studies were supported by the National Institutes of Health grant AA11140 and the Department of Defense grant DAMD17-98-1-8517 as well as the Mayo Foundation.