Article Text


General Liver I
PTU-027 Lifetime alcohol consumption in heavy drinkers with and without liver disease. 1: Threshold effect and male-female differences
  1. D Gleeson1,
  2. A Ali1,
  3. J Jones1,
  4. M Bradley1,
  5. R Peck2,
  6. K McCormack3
  1. 1Liver Unit, Sheffield Teaching Hospitals, Sheffield, UK
  2. 2Radiology, Sheffield Teaching Hospitals, Sheffield, UK
  3. 3Research, Sheffield Teaching Hospitals, Sheffield, UK


Introduction It is unclear whether the risk of alcoholic liver disease (ALD) is related to amount of alcohol consumed (dose effect) or is independent of alcohol consumption above a given threshold (J Hepatol 2004;41:25). The perception that women are more susceptible to ALD than men is based partly on observations that alcohol consumption in women with ALD is less than that of men with ALD. We have previously (Am J Gastro 2008;103:3039) characterised two large cohorts of heavy drinkers (>60 Units/wk (M) or >40 Units/wk (F) for >5 years): patients with decompensated ALD and controls without serious liver disease on clinical, laboratory and ultrasound examination. Here, we aimed to compare total lifetime alcohol consumption (TLA) between these cohorts and to examine male-female differences in consumption.

Methods Subjects (328 patients, 233 male, mean age 48 yr and 237 heavy-drinking controls, 187 male, mean age 48 yr) completed a lifetime alcohol questionnaire. Total alcohol consumption was calculated and the predominant beverage group recorded, separately, at home and outside home, and during Monday–Thursday and Friday–Sunday. Data were summed over each stable drinking period in the subject's lifetime. For individual beverage analysis, we assumed that all alcohol consumed in a given circumstance was the stated predominant beverage. Body weight data were available in 216 controls and (pre-fluid overload) in 198 patients.

Results TLA was (median (IQR)) 118 (80–175) × 103 Units in patients and 131 (93–183) × 103 units in controls (p=0.041) by Mann–Whitney): 9 (6–12)-fold and 9 (7–12)-fold higher respectively than the minimum specified by the above inclusion criteria. TLA, corrected for body wt was similar in patients and controls. Females with ALD consumed less total alcohol (absolute 85 (52–11) vs 136 (94–193) × 103 units (p<0.01) and corrected for wt (p=0.04)), and also, less beer (p=0.001) but more wine (p<0.001) and a trend towards more spirits (p=0.166) over their lifetime than did males with ALD. However, female controls showed similar significant differences from male controls: lower TLA (both absolute 82 (69–119) vs 145 (108–193) × 103 Units) and corrected for body weight) and also, less beer, more wine and more spirits (all comparisons p≤0.001).

Conclusion The higher total alcohol consumption in controls than in patients is more consistent with a threshold effect then a dose effect of alcohol for development of ALD. Male-female differences in consumption are independent of the presence of liver disease and therefore, are not evidence per se of increased female susceptibility to ALD, although the data do not exclude this.

Competing interests None declared.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.