Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Aliment Pharmacol Ther. Author manuscript; available in PMC 2011 November 20.
Published in final edited form as:
PMCID: PMC3220929

Synergistic association between alcohol intake and body mass index with serum alanine and aspartate aminotransferase levels in older adults: the Rancho Bernardo Study



Body-mass-index (BMI) and alcohol consumption predict elevated serum alanine (ALT) and aspartate (AST) aminotransferase levels in young adults. It is unclear if alcohol intake and BMI and their joint association have a differential effect on ALT and AST levels in older adults.


To determine the association between alcohol, BMI, and their combined effect with serum ALT and AST in older community-dwelling adults in the United States.

Design and Setting

A cross-sectional, population-based study in older adults


(n=2364) from the Rancho Bernardo Study (54% women; mean age: 70 years, BMI: 25 kg/m2, alcohol users: 63%) who attended a research visit in 1984-87. BMI was recorded by a trained nurse and alcohol use ascertained by a validated questionnaire.


Multiply adjusted odds-ratio (OR) and 95% confidence intervals (CI) of elevated serum ALT and AST (defined as ≥ 30 U/L in men and ≥ 19 U/L in women) were calculated for alcohol and BMI separately and their joint exposure using multivariate logistic regression models adjusted for age, body mass index, total cholesterol, serum triglycerides, fasting plasma glucose, systolic blood pressure, and diabetes mellitus.


In multivariate-adjusted logistic regression models, obesity independently increased the odds of elevated ALT in this cohort of older men and women by 3.0 (95% CI, 1.7-5.3) and 1.8 (95%CI, 1.1-2.7), respectively. Joint effects of consuming > 3 alcoholic drinks/day and obesity raised the odds of elevated ALT by 8.9 (95%CI, 2.4-33.1) and AST by 21-fold (95%CI, 2.6-170.1), demonstrating synergism. Obese had higher odds of elevated ALT even at 0 ≤ 1 drink/day.


In older men and women, combination of obesity with alcohol is synergistic in increasing the risk of liver injury. Older obese should restrict their alcohol intake as the risk of liver injury is higher.

Keywords: Obesity, body mass index, alanine aminotransferase, aspartate aminotransferase, nonalcoholic fatty liver disease, nonalcoholic steatohepatitis, population-based, sex-based, epidemiology


Obesity and alcohol are important risk factors affecting disease and health in the general population (1). Both obesity and excess alcohol intake have been linked to increased morbidity and mortality in large epidemiologic studies worldwide (1-3). Obesity- and alcohol-related disease states are increasing in the United States, and perhaps worldwide (3, 4). Understanding the health-related effects of alcohol use in this era of obesity epidemic poses a global public health challenge (5). There is considerable uncertainty regarding the threshold of body mass index (BMI) (or degree of obesity) and alcohol consumption that may predispose to elevated serum serum alanine (ALT) and aspartate (AST) aminotransferase levels, a marker for liver injury in general population. Elevated ALT and AST levels reliably identify individuals who have acute or chronic liver disease (6). Based on the National Health and Nutritional and Examination Survey (NHANES), it is estimated that 7-9% of the adult US population may have elevated ALT and AST levels(7).

Fatty liver disease in now considered to be the one of the most common causes of elevated aminotransferase levels in the United States (8, 9). There is strong evidence to suggest that fatty liver disease is associated with obesity, alcohol use, diabetes, and cardio-metabolic risk factors (10, 11). Because both obesity and excess alcohol use are potentially modifiable risk factors, understanding their association with each other and with serum ALT and AST might help in understanding the pathogenesis of fatty liver disease and also provide opportunities for intervention (12).

Although low levels of alcohol intake may prolong life, prevent diabetes, and have other health benefits, excess alcohol intake is harmful and can cause fatty liver disease (13, 14). Therefore, it is critical to better understand the level of alcohol and overweight or obesity together or independently modify levels of serum ALT and AST in community-dwelling older adults whose pattern of drinking is primarily social and whose income permits adequate nutrition in the face of regular alcohol use. The National Institute on Alcohol Abuse and Alcoholism's strategic five year plan (2008-13) underscores the need to examine the effect of alcohol and obesity.

The American Gastroenterological Association future trends committee report on aging and geriatric gastroenterology suggests that elderly may be more susceptible to liver injury (15). Therefore, the combined effect of obesity and alcohol consumption on the development of elevated serum ALT and AST needs further elucidation especially in older adults. In addition, there is limited data on the threshold of alcohol intake associated with elevated serum ALT and AST, or whether this varies by sex or body size in older adults.

In this study, we used data from a large, well-characterized community-dwelling cohort of older men and women to test the following hypotheses: 1) Men and women have different drinks-per-day thresholds for elevating serum ALT and AST (separate models); we used logistic regression analyses to predict odds of elevated serum ALT and AST in individuals stratified by alcohol dose in drinks/day compared to non-drinkers. 2) Men and women have a different association of overweight and obesity with elevated serum ALT and AST levels; to evaluate the sex-specific relationship between obesity, and serum ALT and AST levels, we investigated whether odds of elevated serum ALT and AST (separate models) differ in men and women in overweight and obese participants as compared to normal weight individuals. 3) Obesity and alcohol use have a synergistic association with elevated serum ALT and AST in either obese drinkers or overweight drinkers vs. normal weight non-drinker; to test this hypothesis, we examined the joint effects of alcohol dose and BMI dose (normal weight, overweight and obese individuals) on odds of elevated serum ALT and AST levels separately. 4) To examine if the interaction between alcohol use and obesity with serum aminotransferase is synergistic; to test this hypothesis, we conducted the Wald test for interaction and also performed qualitative graphical analyses and combination index analyses. STROBE guidelines for reporting cross-sectional epidemiologic studies were followed (16).


Study Design and Setting

This is a cross-sectional study of participants of the Rancho Bernardo Study (RBS), a prospective population-based study in older adult residents of a suburban Southern California community. The RBS cohort was established in 1972, when 82% of residents of a geographically defined community were recruited. The cohort is almost entirely white, most with at least a college education, and mostly white-collar workers. Participants were followed by annual mailed questionnaires and returned for research clinic visits approximately every four years.

Study Participants

The current study cohort was derived from 2,466 residents who attended the clinic between 1984-87, approximately 80% of surviving community-dwelling adults aged 30 or more. During this visit, participants provided a medical and medication history. Weight and height were measured in the clinic with participants wearing light clothing and no shoes by a trained investigator. Alcohol use was self-reported to a trained interviewer who asked about frequency, amount, and type of alcohol. A serum sample for serum ALT and AST and alcohol questionnaire were collected the same day. For the present analysis, alcohol was considered as number of alcoholic drinks consumed in a day, whether beer, wine, or spirits. One alcoholic drink is equivalent to 10 grams of alcohol. Reported alcohol intake was indirectly validated by showing a strong association with HDL cholesterol (17), and by showing a similar response to a nutritionist interviewer who obtained alcohol intake as part of a food-frequency questionnaire. The study protocol was approved by the institutional review board of the University of California, San Diego. All participants provided written informed consent. The details of the cohort, selection criteria, and purpose of the RBS have been published previously (18, 19).

Of the 2466 participants, aged 30 years or more, who attended examination in the l984-7 clinic visit, we excluded 36 who did not have serum ALT data, and 66 who had missing data on alcohol use, weight and height, lipids, fasting plasma glucose, smoking, and/or the presence or absence of a history of diabetes, leaving 2364 participants for these analyses.

Outcome variables

Fasting concentrations of serum ALT and AST, the outcome variables for all analyses, were measured by spectrophotometry. We defined elevated ALT using the cut-point of serum ALT ≥ 30 IU/L in men and ≥19 IU/L in women as suggested by Prati et al (20, 21). We used the same cut-off level for serum AST as previously used in a population-based cohort study(22). These cut-points, recommended by experts (6) were decided, a priori, before conducting the analyses. We also conducted a sensitivity analyses using a cut-point of elevated AST as >36 U/L (clinical laboratory cut point).

Covariate Assessment

In all sex-specific analyses we adjusted for the following variables based upon previously published data (6): age, sex, body mass index, fasting serum triglyceride, total cholesterol, systolic blood pressure, alcohol drinks per/day (when applicable), fasting plasma glucose levels, and history of diabetes. A trained investigator recorded two morning blood pressure readings in seated participants using a mercury sphygmomanometer, according to the Hypertension Detection and Follow-up Program protocol (23). Plasma glucose was measured using the glucose-oxidase method. Fasting plasma total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglyceride levels were measured using enzymatic methods in a laboratory certified by the Centers for Disease Control (24). Participants who had a systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 (recorded by a study investigator) or receiving medications for the treatment of hypertension were classified as hypertensive. Diabetes was defined as a fasting plasma glucose ≥ 7 mmol/L (126 mg/dl) or treatment with either insulin or an oral hypoglycemic medication.

Exposure categories

Alcohol use

Participants who reported 1) no alcohol use ever, or 2) no alcohol use in the last year, or 3) no alcohol use in the last week and no alcohol in an average week in the last year were classified as nondrinkers, and served as the referent group. Those who reported any alcohol in the previous year were further classified regarding their pattern of alcohol intake, into four exposure categories (selected a priori, prior to looking at the data): 0≤ 1 drink/day, 1 ≤ 2 drinks/day, 2 ≤ 3 drinks/day, > 3 drinks/day. Weekly consumption of alcohol use was recorded by a trained investigator and recall of alcohol intake in the previous 24 hours was ascertained by a certified dietician in a diet questionnaire with quantities assessed by use of containers and food models as previously described (17, 25-27).

BMI categories

Body mass index (BMI) was defined as weight (kilograms) divided by square of the height (meters). Participants were classified, a priori, into three body size categories, based on the World Health Organization (WHO) definitions of normal weight (18.5 to less than 25) as referent, overweight (BMI 25 to less than 30), and obesity (BMI 30 or more). Individuals below the BMI of 18.5 (n=42) were excluded for these analyses.

Analysis plan

Serum ALT and AST were logarithmically transformed to fulfill conditions of normality as their distribution was skewed. Sex-specific unadjusted followed by multiply adjusted logistic regression models were constructed to separately examine the odds of elevated serum ALT and AST in individuals who consumed 0 ≤ 1 drink/day, 1 ≤ 2 drinks/day, 2 ≤ 3 drinks/day, > 3 drinks/day, respectively, using individuals who consumed no alcohol as the referent group. We then performed unadjusted followed by multiply-adjusted logistic regression models to analyze the association between odds of elevated serum ALT and AST (separate models) with overweight and obesity compared to normal BMI as the referent group. Logistic regression analyses were undertaken to understand the joint association of alcohol dose and BMI categories on multiply adjusted odds for elevated serum ALT and AST levels (separate models). Sex-specific analyses were not conducted for joint effects models because alcohol dose response and BMI dose response models did not differ with sex. However, all logistic regression analyses were conducted using sex-specific cut-points for elevated serum ALT and AST as previously stated. A sensitivity analyses was conducted using a cut-point of elevated AST as >36 U/L. We tested the hypotheses by adjusting for age, BMI, systolic blood pressure, fasting plasma glucose, total cholesterol, triglycerides, total alcohol consumption (drinks/week) (when applicable), and prevalent diabetes for all models (when applicable). These variables were chosen as they are independent predictor of elevated serum aminotransferase levels in this cohort, and have been shown to be associated with higher odds of elevated serum aminotransferase levels based upon previously published studies(8). P-value for trend was estimated using the Cochran-Armitage test. Interaction between categories of BMI and alcohol intake was calculated using the Wald test for interaction. As the Wald test underestimates interaction especially when not significant therefore we also analyzed this interaction graphically. To determine whether the interaction was synergistic or additive, combination index analysis was used. A combination index equal to 1 would suggest additive interaction and a combination index < 1 would be suggestive of synergistic or multiplicative interaction between alcohol and obesity as proposed by Berenbaum (28).

Additionally, the statistical significance of differences in ALT and AST values among the alcohol dose exposure categories and BMI exposure categories was determined by comparing least square of means of the log-transformed ALT or AST levels.

A separate analysis was conducted excluding individuals (n = 81) with a history of alcohol abuse to assess whether the observed findings were independent of alcohol abuse. Alcohol abusers were identified using CAGE questionnaire criteria; they answered yes to at least 2 of the 4 CAGE questions.

All results were reported as odd ratios (ORs) and 95% confidence intervals (95% CI) except least square of means analyses. Either an odds ratio (95%CI) that did not include null or a two-sided p-value < 0.05 was considered statistically significant. SAS version 8.1 (SAS Institute, Cary NC) was used for all analyses.


Baseline characteristics of the 2364 participants based upon alcohol and BMI categories are shown in online Table A and B, respectively. The 2364 participants had a mean (± SD) age of 69.6 (± 10.5) years with an average BMI of 24.9 (± 3.7) kg/m2; 96 % were aged 50 or older, and 54% were women. 48.8% of men and 27.9% of women were overweight, and 9.2 % men and 8.3 % women were obese. Overall 36.8 % were classified as nondrinkers. Among those who did drink alcohol (1495 out of 2364), the distribution of alcohol consumption was as follows: 739 (55.6 % women) participants reported ≤ 1 drink/day, 414 (50.3% women) participants reported 1 ≤ 2 drinks/day, 233 (44.6% women) participants reported 2 ≤ 3 drinks/day, and 109 (26.6% women) participants reported > 3 drinks/day. Average (± SD) serum ALT and AST was 20.2 (± 15.5) IU/ml and 26.7 (± 13.7) IU/ml, respectively.

Association between Alcohol Dose and Serum ALT (Table 1.)

71 % of men and 57 % of women reported that they drink alcohol. Men (mean ± SD; 1.3 ± 1.3 drinks/day) consumed slightly more alcohol than women (mean ± SD; 0.8 ± 1.0 drinks/day). Similarly, younger (below median [72 years]age) men and women (mean ± SD; 1.2 ± 1.3 drinks/day) consumed more alcohol than older (above median age) men and women (mean ± SD; 0.9 ± 1.1 drinks/day).

Contrary to our hypothesis, no sex-specific differences between the alcohol dose and the odds of elevated serum ALT were observed. The odds of elevated serum ALT were higher for both men and women who consumed greater than 3 drinks of alcohol daily, but not statistically significant in individuals who consume 3 or less drinks/day as compared to nondrinkers, suggestive of a threshold effect at this level in both men and women. Un-adjusted odds of elevated serum ALT in men and women who consumed > 3 drinks/day vs. no alcohol was 3.2 (95% CI, 1.8-5.6), and 2.6 (95% CI, 1.2-5.8), respectively. The findings were not statistically significant at the lower level of alcohol consumption in both men and women. Multiply adjusted odds of elevated serum ALT in men and women who consume >3 drinks/day vs. no alcohol was 2.5 (95% CI, 1.3-4.7) and 2.9 (95% CI, 1.3-6.3), respectively. Additionally, a statistically significant linear association between alcohol consumption and odds of elevated serum ALT in men (p-value for trend <0.005) but not in women (p-value for trend 0.2) was also observed. Unadjusted and multiply adjusted least square of means of log-transformed ALT was statistically significant only in the >3 drinks/day category as compared to nondrinkers (online Table C).

Association between Alcohol Dose and Serum AST (Table 1.)

Consistent with these results, we observed similar associations between alcohol dose (> 3 drinks/day) and the odds of elevated serum AST in both men and women. Unadjusted OR for men and women who drank > 3 drinks/day vs. no alcohol was 2.5 (95% CI, 1.5-4.1) and 7.0 (95% CI, 1.0-52.2), respectively. Multiply-adjusted OR for men and women who consume > 3 drinks/day (no alcohol use as referent group) was 2.4 (95% CI, 1.4-4.1) and 7.7 (95% CI, 1.0-57.5), respectively. The confidence intervals were wide for women due to smaller number of women in > 3 drinks/day group (n=29). None of the other exposure categories achieved statistical significance in either model. Similar to ALT, a linear association was observed between alcohol dose and serum AST in men (p-value for trend < 0.005), and a nonsignificant association was seen in women (p-value for trend 0.08). The results remained consistent with AST cut point of >36 U/L for either sex.

The mean ± SD alcohol consumption in the highest alcohol use category, i.e., >3 drinks/day, did not differ (4.3 ± 1.4 drinks/day in men vs. 4.0 ± 0.8 drinks/day in women; p-value for sex difference 0.2).

Association of BMI Dose and Serum ALT and AST (Table 2.)

In both men and women, increasing BMI was associated with higher serum ALT levels. Obesity had a nonsignificantly greater effect on ALT in men than women, with a multiply adjusted OR of elevated serum ALT in obese men and women (normal weight as referent group) of 2.9 (95% CI, 1.6-5.2) and 1.7 (95% CI, 1.1-2.6), respectively. Statistically significant linear associations between BMI categories and serum ALT were seen in both sex (p-value for trend in men and women separately <0.001). In women, no statistically significant association between serum AST and BMI was noted in any logistic regression models with either AST cut point. In men, odds of elevated AST, using a cut point of > 36 U/L, were higher in obese than normal weight. When the serum ALT and AST were analyzed as a continuous variable, the least square of means of ALT, the results remained consistent (online Table D).

Secondary Analyses

In another model excluding individuals with alcohol abuse by CAGE criteria, the multiply adjusted odds of elevated serum ALT in men and women who consume >3 drinks/day vs. no alcohol was 1.7 (95% CI, 1.1-2.6) and 2.9 (95% CI, 1.2-6.8), respectively. Consistent with the primary analysis, the findings were not statistically significant at the lower level of alcohol consumption in both men and women, suggestive of a threshold effect at >3 drinks/day of alcohol consumption. The results across all analyses in this subset of participants were not materially different (data not shown).

Association Between High Alcohol Intake, Obesity, and Elevated Serum ALT and AST

In both unadjusted and multiply adjusted analyses, in men obesity had higher odds of predicting elevated serum ALT than high (> 3 drinks/day) alcohol intake. However, in women > 3 drinks/day had higher odds of predicting serum ALT than obesity. High alcohol consumption, but not obesity, was an independent predictor of elevated serum AST in both sexes.

The joint association of alcohol dose and obesity with serum ALT (figure 1) and AST (figure 2) shows that increase in body size and alcohol intake increase the odds of elevated serum ALT and AST. Obese individuals have higher odds of elevated serum ALT even at lower levels of alcohol consumption. Multiply adjusted OR of joint effects of alcohol and BMI for elevated serum ALT (OR; 8.9 [95%CI; 2.4-33.1]) and AST (OR; 21.0 [95%CI; 2.6-170.7]) was highest in obese participants who consumed greater than 3 drinks per day. Although the Wald test for interaction between alcohol use and BMI did not reach statistical significance (p-value 0.6), qualitative graphical analysis suggests a multiplicative interaction between obesity and alcohol intake (Online figure 1). Furthermore, the combination index (CI) for interaction between alcohol and BMI was 0.39, which is strongly suggestive of a multiplicative interaction (CI of 1 is additive and CI > 1 is antagonistic and CI < 1 is multiplicative(28)).

Figure 1
Joint association of alcohol consumption and body mass index on multiply-adjusted odds of elevated serum ALT levels
Figure 2
Joint association of alcohol consumption and body mass index on multiply-adjusted odds of elevated serum AST levels.


Principal Findings

In this community-based sample of white older adults, alcohol use of up to three drinks/day did not raise serum ALT or AST in either sex. The results were consistent when ALT and AST were analyzed both quantitatively (least square of means) and qualitatively (elevated versus normal). These findings suggest that threshold for elevating serum ALT and AST was similar in older men and women and appeared to be at >3 drink/day of alcohol consumption. Obesity was more strongly associated with serum ALT than AST in both sexes. This study shows that obesity and high alcohol consumption synergistically raised serum ALT and AST in separate analyses, and the interaction appeared to have a strong multiplicative effect in older adults.

Obesity was more strongly associated with elevated serum ALT than > 3 drinks/day in men. BMI had a linear association with serum ALT in either sex but with AST only in men. High alcohol consumption but not obesity was an independent predictor of elevated serum AST in both men and women.

This study extends current literature that obesity and alcohol jointly increase serum aminotransferase not only in young adults but also in older adults (29). Our study provides new data suggesting a multiplicative rather than additive interaction of obesity and alcohol with elevated serum ALT and AST in separate models. In addition, these data suggest that obese individuals have elevated serum ALT at lower alcohol consumption levels than nonobese individuals. Therefore, obese individuals may be more susceptible to alcohol-induced liver injury at a much lower dose of alcohol intake. These data may have important public health implications in the current obesity epidemic.

In the Context of Current Literature

Previous studies have suggested that both alcohol and obesity increase the odds of elevated serum aminotransferase levels (10, 29), but, to our knowledge, this is the first study to examine joint effects in older adults with median age of 72 years in a United States population-based cohort. In a large epidemiologic study in Japanese men aged between 20 and 59 years, Suzuki and colleagues reported that light (70 to 140 grams per week, which is equivalent to 1-2 drinks/day) to moderate alcohol consumption (140 to 280 grams per week of alcohol, which is equivalent to approximately 2-4 drinks per day) appeared to protect against development of elevated serum aminotransferase levels in younger men, but in older men the protection was limited to the light alcohol consumption group only (30). Our study sample is significantly older, Caucasian only, and includes both men and women and is more representative of a suburban U.S. population.

The association between BMI dose and odds of elevated serum aminotransferase levels has been reported by several groups. Our findings are in agreement with cross-sectional data based upon NHANES reported by Ruhl and colleagues that overweight and obese younger individuals (mean age 44 years) have higher odds of elevated serum aminotransferase levels at lower levels of alcohol consumption (29). They did not examine serum ALT and AST separately and average age of participants in that study was 44 years. Our cohort had an average age of 70 years and we examined serum ALT, a cytosolic enzyme, and AST, a mitochondrial enzyme, in separate models. Our results suggests that ALT differs from AST as ALT correlates better with overweight and obesity, and is increased at even a 1-2 drinks/day of alcohol use in obese. However, no increase in odds of elevated AST was noted in overweight or obese at 1-2 drinks/day. Odds of serum AST in obese who drink more than 3 drinks/day were 21 times whereas odds of serum ALT were raised to 8.9 suggesting that obese individuals who drink > 3 drinks/day may not only be sensitive to hepatocyte injury leading to release of ALT but mitochondrial injury resulting in release of serum AST. Our data suggest that obesity is a stronger predictor of elevated serum ALT than alcohol consumption in men, results similar to those of Lee and co-authors based on a study in 6846 Korean men; they also suggest that body weight may be a stronger determinant of serum ALT than alcohol (31). However, our study suggests that the reverse is true in women.

Alcohol use is typically self-reported in epidemiologic studies, and may be under- reported in most studies. It is possible that under-reporting of alcohol use in previous studies precluded meaningful analysis of the high end of the dose response of alcohol consumption, which was recorded here. We cannot be certain of the exact threshold level in the present study for the same reason. It is possible that there is less under-reporting in older adults, such as this cohort with mean age of 70 years, because they are usually retired with less concern about adverse consequences. Some evidence for the validity of increasing dose, if not the exact amount, comes from indirect validation--a strong association of reported intake with HDL cholesterol, as previously reported in this cohort (17); there was also a similar amount of alcohol consumption reported using the interviewer-administered questionnaire and data obtained by a self-administered food-frequency questionnaire (26, 27).


Previous studies that have suggested an integral role of CYP2E1 in the pathogenesis of fatty liver disease due to alcohol and obesity (32). It is well accepted that both diet-induced obesity and increased alcohol consumption lead to induction of CYP2E1 in the liver, which may explain the synergistic effect of these two factors in causing liver injury or elevated serum ALT and AST (33). Several groups have shown that mechanism of liver injury in nonalcoholic fatty liver disease and alcoholic liver disease may be related to certain common metabolic pathways such as induction of CYP2E1 activity, defect in β-oxidation in fatty acid synthesis, increased deposition of iron, elevation in serum ferritin levels, and depletion of S-adenosoyl methionine (SAMe) and glutathione (GSH) leading to oxidative stress (33-36). In addition, TNF-α has been implicated in the pathogenesis of alcoholic-induced steatohepatitis as well as obesity-related nonalcoholic fatty liver injury in humans as well as in animals (37, 38). All the above studies suggest that reactive oxygen species may be the underlying mechanism causing release of serum ALT and AST in response to stress due to alcohol and/or obesity.

We found that serum ALT, and not AST, is more closely associated with obesity and overweight. This pattern (serum ALT higher than AST) is commonly seen in nonalcoholic fatty liver and is consistent with previous studies conducted in patients with nonalcoholic fatty liver (8).

Strengths and Limitations

The strengths of this study include the use of a well-characterized large population-based sample of community-dwelling older adults. To our knowledge, this is the first study to report joint effects of alcohol and BMI dose response in older adults in a community-based cohort in the United States. Our study provides detailed data on sex-specific differences between serum ALT and AST by alcohol dose and body size in separate models. However, the study has some limitations: the sample is almost entirely Caucasian and largely middle to upper-middle class, consistent with the demographics of defined suburbs at the time the study was initiated. Generalizability of these findings to other race/ethnic groups or social classes or younger adults is unclear. Information on other liver diseases, such as viral hepatitis, or autoimmune or metabolic liver disease was not available, and therefore these conditions could not be excluded as a cause of ALT elevations. But the cohort is relatively healthy and these conditions are not so common that they would be expected to influence the consistent study results across analyses. Finally, the age of the cohort means that these are healthy survivors, and liver-safe levels of alcohol may differ in survivors than in the general population of older adults. The average consumption of alcohol in younger (less than median age of 72 years) individuals was higher in both men and women in our study suggesting that alcohol intake decreases with age. The participants were classified into five alcohol dose categories, a priori, before conducting data analyses. The average alcohol consumption in participants in >3 drinks/day category was 4 drinks/day in both men and women, suggesting that our results were not driven by extreme values in this group. As there were only a small number of individuals consuming higher levels of alcohol, further analyses at higher doses (>4 or 5 drinks/day) were not feasible. However, we could safely conclude that alcohol use of up to three drinks/day did not raise serum ALT or AST in either sex.

Implications for Further Research

Our findings suggest that threshold for elevation in serum ALT and AST is at >3 drinks per day in older white men and women. For the reasons stated above, further data are needed to confirm these findings in other cohorts. Future studies may consider using separate models for serum ALT and AST, especially when the aim is to understand the effect of obesity and other metabolic variables. Our study provides data that odds of elevated serum AST is more closely related to alcohol consumption in a dose-dependent manner but not related to similar changes in increase in BMI. Moreover, obesity may increase the odds of serum AST only in heavy daily drinkers (> 3 drinks/day). As the US and world population is rapidly aging, greater understanding of metabolic effects of alcohol and obesity on the health of older adults is needed.


The threshold for elevating ALT and AST is similar in older men and women and appears to be > 3 drink/day of alcohol consumption; alcohol use of up to three drinks/day did not raise serum ALT or AST in either sex. Obesity is more strongly associated with serum ALT than AST in both sexes. In men, obesity is more strongly associated with elevated serum ALT than alcohol intake; in women, alcohol intake rather than obesity is a stronger predictor of elevated serum ALT. Obese individuals have higher odds of elevated serum ALT even at lower levels of alcohol consumption. Obesity and high alcohol consumption synergistically raise serum ALT and AST, and the interaction appears to have a multiplicative effect. If confirmed that obesity plus daily alcohol have multiplicative effects increasing serum ALT and AST, these observations have important clinical implications, as both obesity and excessive alcohol intake are epidemic, and are potentially modifiable risk factors for development of elevated serum ALT and AST, which is associated with fatty liver disease. We repeat that these data do not endorse increasing the alcohol consumption to three drinks/day. Further research is needed to examine if obesity modifies the effect of moderate alcohol use on cardio-metabolic risk factors such as the future risk of development of diabetes or CAD and whether the beneficial effect of moderate alcohol use are limited to normal weight.

Supplementary Material

Online figure 1. Interaction between alcohol consumption >3 drinks/day and obesity

Footnote: The two lines represent the prevalence of elevated serum ALT that cross each other when the exposure dose increases it is suggestive of an interaction.

Online Table A. Baseline characteristics of the participants based upon alcohol consumption (drinks/day) status.

Online Table B. Baseline characteristics of the participants stratified by body mass index

Online Table C. Least squares of mean (geometric) and standard errors of serum ALT and AST stratified by alcohol dose

Online Table D. Least squares mean (geometric) and standard error of mean of serum ALT and AST stratified by body mass index


Authors would like to thank Dr. David Brenner for providing insightful comments on the study results, and Drs. Naga Chalasani and Joel Lavine for critical review of the manuscript, and Dr. Hashem El-Serag for reviewing the abstract and providing helpful suggestions to improve the clarity of analyses. This work was presented in part as an abstract at the 59th annual meetings of the American Association for the Study of Liver Diseases in San Francisco in November 2008.

Funding Support: This work is supported in part by the American Gastroenterological Association (AGA) Foundation – Sucampo – ASP Designated Research Award in Geriatric Gastroenterology and by a T. Franklin Williams Scholarship Award; Funding provided by: Atlantic Philanthropies, Inc, the John A. Hartford Foundation, the Association of Specialty Professors, and the American Gastroenterological Association to Rohit Loomba, MD, MHSc. This research was funded in part with the support of the UCSD Digestive Diseases Research Development Center, U.S. PHS grant #DK080506. This work was supported in part by the National Institute of Health grants RO1AG28507, R37AG007181, and RO1DK31801 to Elizabeth Barrett-Connor, MD.

Role of Funding Agencies: Funding agencies did not have any role in the design and conduct of the study, collection, management, analysis and interpretation of the data; preparation, review, or approval of the manuscript. There is no potential conflict of interest.


alanine aminotransferase
aspartate aminotransferase
body mass index
nonalcoholic fatty liver disease
nonalcoholic steatohepatitis
odds ratio
standard deviation


Conflict of interest: None disclosed.


1. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States, 2000. Jama. 2004;291:1238–1245. [PubMed]
2. Banks J, Marmot M, Oldfield Z, Smith JP. Disease and disadvantage in the United States and in England. Jama. 2006;295:2037–2045. [PubMed]
3. Olshansky SJ, Passaro DJ, Hershow RC, Layden J, Carnes BA, Brody J, Hayflick L, et al. A potential decline in life expectancy in the United States in the 21st century. N Engl J Med. 2005;352:1138–1145. [PubMed]
4. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. Jama. 2006;295:1549–1555. [PubMed]
5. Sturm R. The effects of obesity, smoking, and drinking on medical problems and costs. Health Aff (Millwood) 2002;21:245–253. [PubMed]
6. Kim WR, Flamm SL, Di Bisceglie AM, Bodenheimer HC. Serum activity of alanine aminotransferase (ALT) as an indicator of health and disease. Hepatology. 2008;47:1363–1370. [PubMed]
7. Clark JM, Brancati FL, Diehl AM. The prevalence and etiology of elevated aminotransferase levels in the United States. Am J Gastroenterol. 2003;98:960–967. [PubMed]
8. Angulo P. GI epidemiology: nonalcoholic fatty liver disease. Aliment Pharmacol Ther. 2007;25:883–889. [PubMed]
9. Ioannou GN, Boyko EJ, Lee SP. The prevalence and predictors of elevated serum aminotransferase activity in the United States in 1999-2002. Am J Gastroenterol. 2006;101:76–82. [PubMed]
10. Bellentani S, Saccoccio G, Masutti F, Croce LS, Brandi G, Sasso F, Cristanini G, et al. Prevalence of and risk factors for hepatic steatosis in Northern Italy. Ann Intern Med. 2000;132:112–117. [PubMed]
11. Ioannou GN, Weiss NS, Boyko EJ, Kahn SE, Lee SP. Contribution of metabolic factors to alanine aminotransferase activity in persons with other causes of liver disease. Gastroenterology. 2005;128:627–635. [PubMed]
12. Sanyal AJ. AGA technical review on nonalcoholic fatty liver disease. Gastroenterology. 2002;123:1705–1725. [PubMed]
13. Fuchs CS, Stampfer MJ, Colditz GA, Giovannucci EL, Manson JE, Kawachi I, Hunter DJ, et al. Alcohol consumption and mortality among women. N Engl J Med. 1995;332:1245–1250. [PubMed]
14. Jackson R, Beaglehole R. Alcohol consumption guidelines: relative safety vs absolute risks and benefits. Lancet. 1995;346:716. [PubMed]
15. Hall KE, Proctor DD, Fisher L, Rose S. American gastroenterological association future trends committee report: effects of aging of the population on gastroenterology practice, education, and research. Gastroenterology. 2005;129:1305–1338. [PubMed]
16. Vandenbroucke JP, von Elm E, Altman DG, Gotzsche PC, Mulrow CD, Pocock SJ, Poole C, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. PLoS Med. 2007;4:e297. [PubMed]
17. Barrett-Connor E, Suarez L. A community study of alcohol and other factors associated with the distribution of high density lipoprotein cholesterol in older vs. younger men. Am J Epidemiol. 1982;115:888–893. [PubMed]
18. Barrett-Connor E, Wingard DL, Criqui MH. Postmenopausal estrogen use and heart disease risk factors in the 1980. Rancho Bernardo, Calif, revisited. Jama. 1989;261:2095–2100. [PubMed]
19. Barrett-Connor EL, Cohn BA, Wingard DL, Edelstein SL. Why is diabetes mellitus a stronger risk factor for fatal ischemic heart disease in women than in men? The Rancho Bernardo Study. Jama. 1991;265:627–631. [PubMed]
20. Prati D, Taioli E, Zanella A, Della Torre E, Butelli S, Del Vecchio E, Vianello L, et al. Updated definitions of healthy ranges for serum alanine aminotransferase levels. Ann Intern Med. 2002;137:1–10. [PubMed]
21. Prati D, Colli A, Conte D, Colombo M. Spectrum of NAFLD and diagnostic implications of the proposed new normal range for serum ALT in obese women. Hepatology. 2005;42:1460–1461. author reply 1461. [PubMed]
22. Loomba R, Hwang SJ, O'Donnell CJ, Ellison RC, Vasan RS, D'Agostino RB, Sr, Liang TJ, et al. Parental obesity and offspring serum alanine and aspartate aminotransferase levels: the framingham heart study. Gastroenterology. 2008;134:953–959. [PMC free article] [PubMed]
23. The hypertension detection and follow-up program: Hypertension detection and follow-up program cooperative group. Prev Med. 1976;5:207–215. [PubMed]
24. Lipid Research Clinics Program Manual of Laboratory Operations: Vol 1: Lipid and Lipoprotein Analysis. 2nd. US Government Printing Office, US Dept of Health, Education, and Welfare; Washington, DC: 1974. NIH publication No. 75-628. In.
25. Jones BR, Barrett-Connor E, Criqui MH, Holdbrook MJ. A community study of calorie and nutrient intake in drinkers and nondrinkers of alcohol. Am J Clin Nutr. 1982;35:135–139. [PubMed]
26. Holbrook TL, Barrett-Connor E, Wingard DL. A prospective population-based study of alcohol use and non-insulin-dependent diabetes mellitus. Am J Epidemiol. 1990;132:902–909. [PubMed]
27. Holbrook TL, Barrett-Connor E. A prospective study of alcohol consumption and bone mineral density. BMJ. 1993;306:1506–1509. [PMC free article] [PubMed]
28. Berenbaum MC. Synergy, additivism and antagonism in immunosuppression. A critical review. Clin Exp Immunol. 1977;28:1–18. [PubMed]
29. Ruhl CE, Everhart JE. Joint effects of body weight and alcohol on elevated serum alanine aminotransferase in the United States population. Clin Gastroenterol Hepatol. 2005;3:1260–1268. [PubMed]
30. Suzuki A, Angulo P, St Sauver J, Muto A, Okada T, Lindor K. Light to moderate alcohol consumption is associated with lower frequency of hypertransaminasemia. Am J Gastroenterol. 2007;102:1912–1919. [PubMed]
31. Lee DH, Ha MH, Christiani DC. Body weight, alcohol consumption and liver enzyme activity--a 4-year follow-up study. Int J Epidemiol. 2001;30:766–770. [PubMed]
32. Nagata K, Suzuki H, Sakaguchi S. Common pathogenic mechanism in development progression of liver injury caused by non-alcoholic or alcoholic steatohepatitis. J Toxicol Sci. 2007;32:453–468. [PubMed]
33. Lieber CS. Alcoholic fatty liver: its pathogenesis and mechanism of progression to inflammation and fibrosis. Alcohol. 2004;34:9–19. [PubMed]
34. You M, Crabb DW. Recent advances in alcoholic liver disease II. Minireview: molecular mechanisms of alcoholic fatty liver. Am J Physiol Gastrointest Liver Physiol. 2004;287:G1–6. [PubMed]
35. Chalasani N, Deeg MA, Crabb DW. Systemic levels of lipid peroxidation and its metabolic and dietary correlates in patients with nonalcoholic steatohepatitis. Am J Gastroenterol. 2004;99:1497–1502. [PubMed]
36. Schattenberg JM, Wang Y, Singh R, Rigoli RM, Czaja MJ. Hepatocyte CYP2E1 overexpression and steatohepatitis lead to impaired hepatic insulin signaling. J Biol Chem. 2005;280:9887–9894. [PubMed]
37. Koteish A, Yang S, Lin H, Huang X, Diehl AM. Chronic ethanol exposure potentiates lipopolysaccharide liver injury despite inhibiting Jun N-terminal kinase and caspase 3 activation. J Biol Chem. 2002;277:13037–13044. [PubMed]
38. Diehl AM. Cytokines and the molecular mechanisms of alcoholic liver disease. Alcohol Clin Exp Res. 1999;23:1419–1424. [PubMed]