Comparing usual quantity/frequency and graduated frequency scales to assess yearly alcohol consumption: results from the 1990 US National Alcohol Survey

Data were used from the 1990 US national alcohol survey to compare two volume measures of alcohol consumption: usual quantity/frequency (QF) and graduated frequency (GF). The QF measure obtained global estimates of average alcohol intake per day; the GF measure is a series of questions on the frequency of consuming specific numbers of drinks which ranges from the most ever consumed in the last year to 1–2 drinks per occasion. The GF measure provided higher estimates of alcohol use. Respondents whose reports were inconsistent between the measures were more likely to be male, never married, and report getting drunk at least monthly and drinking five or more drinks on one occasion at least monthly during the last year.     

What did you drink yesterday? Public health relevance of a recent recall method used in the 2004 Australian National Drug Strategy Household Survey

AIM: To (i) compare the Yesterday method with other methods of assessing alcohol use applied in the 2004 Australian National Drug Strategy Household Survey (NDSHS) in terms of extent of under-reporting of actual consumption assessed from sales data; and (ii) illustrate applications of the Yesterday method as a means of variously measuring the size of an Australian 'standard drink', the extent of risky/high-risk alcohol use, unrecorded alcohol consumption and beverage-specific patterns of risk in the general population. SETTING: The homes of respondents who were eligible and willing to participate. PARTICIPANTS: A total of 24 109 Australians aged 12 years and over. DESIGN: The 2004 NDSHS assessed drug use, experiences and attitudes using a 'drop and collect' self-completion questionnaire with random sampling and geographic (State and Territory) and demographic (age and gender) stratification. MEASURES: Self-completion questionnaire using quantity-frequency (QF) and graduated-frequency (GF) methods plus two questions about consumption 'yesterday': one in standard drinks, another with empirically based estimates of drink size and strength. RESULTS: The Yesterday method yielded an estimate of 12.8 g as the amount of ethanol in a typical Australian standard drink (versus the official 10 g). Estimated coverage of the 2003-04 age 12+ years per-capita alcohol consumption in Australia (9.33 ml of ethanol) was 69.17% for GF and 64.63% for the QF when assuming a 12.8 g standard drink. Highest coverage of 80.71% was achieved by the detailed Yesterday method. The detailed Yesterday method found that 60.1% of Australian alcohol consumption was above low-risk guidelines; 81.5% for 12-17-year-olds, 84.8% for 18-24-year-olds and 88.8% for Indigenous respondents. Spirit-based drinks and regular strength beer were most likely to be drunk in this way, low- and mid-strength beer least likely. CONCLUSIONS: Compared to more widely used methods, the Yesterday method minimizes under-reporting of overall consumption and provides unique data of public health significance. It also provides an empirical basis for taxing alcoholic beverages in accordance with their contributions to harm and can be used to complement individual-level measures such as QF and GF.     

Measuring average alcohol consumption: the impact of including binge drinks in quantity-frequency calculations

AIMS: Average daily alcohol consumption is usually calculated based on self-reports of the quantity (number of drinks consumed per drinking-day) and frequency (number of drinking-days) of alcohol consumption within a given time period. However, this method may underestimate average daily alcohol consumption (and in turn, the prevalence of heavy drinking), because studies demonstrate that respondents do not typically include binge drinking occasions in estimates of their 'usual' or 'average' daily alcohol consumption. DESIGN: We used the Behavioral Risk Factor Surveillance System (BRFSS), an annual random-digit telephone survey of US adults aged 18 years or older, to estimate average daily alcohol consumption using standard quantity-frequency questions, and then recalculated this measure by including self-reports of binge drinking. The proportion of respondents who met a standard, sex-specific definition of heavy drinking based on average daily alcohol consumption was then assessed nationally and for each state. FINDINGS: Compared to standard quantity-frequency methods, including binge drinks in calculations of average daily alcohol consumption increased the relative prevalence of heavy drinking among all adults by 19% to 42% (depending on the method used to estimate the number of drinks per binge). Among binge drinkers, the overall prevalence of heavy drinking increased 53% relative to standard quantity-frequency methods. As a result, half of women binge drinkers and half of binge drinkers aged 55 or older met criteria for heavy drinking. CONCLUSIONS: Including binge drinks (especially the application of age- and sex-specific estimates of binge drinks) in the calculation of average daily alcohol consumption can improve the accuracy of prevalence estimates for heavy drinking among US adults, and should be considered to increase the usefulness of this measure for alcohol surveillance.     

Quantity-frequency measures of alcohol consumption: beverage-specific vs global questions

Estimates of alcohol consumption (ounces of absolute alcohol per day, AA) based on beverage-specific and global quantity-frequency (QF) questions were compared in a survey representative of the adult drinking population in New York State. Beverage-specific AA estimates were higher than global (0.72, 95% confidence intervals=0.68, 0.76) compared to 0.49 (95% confidence intervals=0.47, 0.51), although estimates were highly correlated (r=0.75). Discrepancies between beverage specific and global AA estimates increased as the number of beverages and the amount drunk increased. Sociodemographic characteristics were not significantly related to differences between beverage-specific and global AA estimates after adjusting for the amount drunk; however, drinking patterns did influence the differences. It was concluded that beverage-specific QF questions are probably more valid measures of alcohol consumption than global QF questions, but that the global questions provide useful information. Parallel analyses of variant global QF questions employed in the first US Health and Nutrition Examination Survey obtained similar results.     

The Relationship Between Self‐Reported Drinking and BAC Level in Emergency Room Injury Cases: Is it a Straight Line?

Background: While the validity of self‐reported consumption based on blood alcohol concentration (BAC) has been found to be high in emergency room (ER) samples, little research exists on the estimated number of drinks consumed given a BAC level. Such data would be useful in establishing a dose–response relationship between drinking and risk (e.g., of injury) in those studies for which the number of drinks consumed is not available but BAC is. Methods: Several methods were used to estimate the number of drinks consumed in the 6 hours prior to injury based on BAC obtained at the time of ER admission of n = 1,953 patients who self‐reported any drinking 6 hours prior to their injury and who arrived to the ER within 6 hours of the event, from the merged Emergency Room Collaborative Alcohol Analysis Project (ERCAAP) and the World Health Organization Collaborative Study on Alcohol and Injury across 16 countries. Results: The relationship between self‐reported consumption and averaged BAC within each consumption level appeared to be fairly linear up to about 7 drinks and a BAC of approximately 100 mg/dl. Above about 7 reported drinks, BAC appeared to have no relationship with drinking, possibly representing longer consumption periods than only the 6 hours before injury for those reporting higher quantities consumed. Both the volume estimate from the bivariate BAC to self‐report relationship as well as from a Widmark calculation using BAC and time from last drink to arrival to the ER indicated a somewhat weak relationship to actual number of self‐reported drinks. Conclusions: Future studies may benefit from investigating the factors suspected to be driving the weak relationships between these measures, including the actual time over which the reported alcohol was consumed and pattern of drinking over the consumption period.     

Alcohol consumption during pregnancy in nonindigenous west Australian women

BACKGROUND: High alcohol intake in pregnancy has been linked to abnormal fetal development. There are limited published data in Australia on standard drinks of alcohol consumed on a typical occasion during the periconceptional period or pregnancy. METHODS: During 1995 to 1997, a 10% random sample of all nonindigenous women giving birth in Western Australia was surveyed 12 weeks after delivery (N=4,839). Women were asked questions about alcohol consumption in each of the 4 time periods: the 3 months before pregnancy and each trimester of pregnancy. Questions were framed to measure volume, frequency, and type of alcoholic beverage. RESULTS: 46.7% of the women had not planned their pregnancy. Most women (79.8%) reported drinking alcohol in the 3 months before pregnancy, with 58.7% drinking alcohol in at least 1 trimester of pregnancy. The proportion of women consuming 1 to 2 drinks on a typical occasion did not change much during pregnancy, but the number of occasions declined. Although the proportion of women consuming more than 2 standard drinks on a typical occasion declined after the first trimester, 19.0% of women consumed this amount in at least 1 trimester of pregnancy and 4.3% of women consumed 5 or more standard drinks on a typical occasion in at least 1 trimester of pregnancy. In the first trimester of pregnancy, 14.8% of women drank outside the current Australian guideline for alcohol consumption in pregnancy, decreasing to 10% in the second and third trimesters. CONCLUSIONS: Women generally reduced their average alcohol consumption and the number of standard drinks on a typical occasion as their pregnancy progressed, although 10 to 14% were drinking outside current guidelines for pregnancy. It is important that all women of child-bearing age are aware, well before they consider pregnancy, of the risks of drinking alcohol during pregnancy so they can make informed decisions about their alcohol consumption in pregnancy.     

Variations in quantity-frequency measures of alcohol consumption from a general population survey

This paper examines differences in quantity–frequency (QF) measures of alcohol consumption from the 1988 US National Health Interview Survey. Three methods—global QF, beverage-specific QF, and beverage-specific QF with drink size (QFS)—were used to estimate the average daily ethanol consumption (ADC) of current drinkers. These ADC estimates then were used to categorize drinkers into light, moderate or heavier drinking levels. Total prevalence estimates of heavier drinking were not significantly different among men, but were significantly higher with the QFS measure among women. All mean ADC scores were significantly different for both sexes. The global QF showed the lowest mean consumption, followed by the higher beverage-specific QF and QFS measures. Adding beverage type and drink size to the QF measures increased mean ADC scores for both men and women. However, moderately high correlations (0.84 for men and 0.88 for women) were found with ADC scores from the beverage-specific QF and QFS measures.     

National and state estimates of the mean ethanol content of beer sold in the US and their impact on per capita consumption estimates: 1988 to 2001

BACKGROUND: The average ethanol content of the beer sold in the US is a key factor in determining the per capita consumption of ethanol, the standard measure of alcohol use in aggregate-level research. To address the lack of empirically based estimates of beer ethanol content, we have calculated national estimates for the years 1988 to 2001 and state-specific estimates for 1993 to 2001. METHODS: These estimates are based on the ethanol content by volume of leading brands in each year, the national market share of each leading brand by type, and state-specific market shares of each beer type. RESULTS: The national mean ethanol content of beer was higher than the 4.5% figure typically used, ranging from 4.58% in 1993 to 4.75% in 1996. State-specific mean ethanol content estimates were also found to vary by state and over time. Application of mean ethanol content estimates to the per capita consumption of beer led to higher consumption estimates than those with the 4.5% conversion. For example, in 2000, the national estimate indicates that nearly 10 more drinks (containing 0.6 oz of ethanol) were consumed per person aged 14 years and older during that year. CONCLUSIONS: This may indicate that a larger than previously estimated share of the alcohol consumed in the US is in the form of beer. However, the results also indicate that empirically based estimates of wine and spirits mean that ethanol content may modify their consumption estimates as well.     

Effectiveness of the derived Alcohol Use Disorders Identification Test (AUDIT-C) in screening for alcohol use disorders and risk drinking in the US general population

BACKGROUND: The three consumption questions from the Alcohol Use Disorders Identification Test (AUDIT-C) are increasingly used as a screener for alcohol use disorders (AUDs) and risk drinking. METHODS: In a representative sample of US adults 18 years of age and older, AUDIT-C scores (derived from consumption questions embedded in a large national survey) were used to estimate sensitivity, specificity, and areas under receiver operator characteristic curves (AUROCs) for alcohol dependence, any AUD, and risk drinking. AUDs were defined according to DSM-IV criteria. For men, risk drinking was defined as consuming >14 drinks per week or >4 drinks in a single day at least once a month; for women, the weekly and daily limits were >7 drinks and >3 drinks, respectively. The derived AUDIT-C was evaluated among past-year drinkers (n = 26,946), within the total population (n = 43,093), in groups defined by age, sex, and race/ethnicity, and among pregnant women, persons attending an emergency room, and college students. RESULTS: For past-year drinkers, the AUROCs for the derived AUDIT-C were 0.887 for alcohol dependence, 0.860 for any AUD, and 0.966 for risk drinking. Scores were higher in the total population, 0.931, 0.917, and 0.981, respectively. The derived AUDIT-C performed slightly better in screening for dependence among women than men. Screening for risk drinking was better among men, probably because the third AUDIT-C question directly mirrors one of the definitions of risk drinking for men but not for women. Performance in pregnant women, past-year emergency room patients, and college students was on a par with performance in the general population. CONCLUSIONS: The derived AUDIT-C performs well in screening for AUDs and risk drinking. The use of variable cut points for men and women improves its sensitivity and specificity. Validation in a realistic screening situation, in which the AUDIT-C questions are asked as stand-alone and not embedded items, is a critical future step.     

Distribution of alcohol consumption and expenditures and the impact of improved measurement on coverage of alcohol sales in the 2000 National Alcohol Survey

BACKGROUND: To validate improved survey estimates of alcohol volume and new expenditures questions, these measures were aggregated and evaluated through comparison to sales data. Using the new measures, we examined their distributions by estimating the proportion of mean intake, heavy drinking days, and alcohol expenditures among drinkers grouped by volume. METHODS: The 2000 National Alcohol Survey is a random digit dialed telephone survey of the United States with 7,612 respondents including 323 who were recontacted for drink ethanol measurement. Among drinkers, we utilized improved drink ethanol content estimates and beverage-specific graduated frequency measures to assess alcohol consumption and past month beverage-specific spending reports to estimate expenditures. RESULTS: Coverage of alcohol sales by the new measures was estimated to be 52.3% for consumption and 59.3% for expenditures. Coverage was best for wine at 92.1% of sales, but improved most for spirits from 37.2% to 55.2%, when empirical drink ethanol content was applied. Distribution estimates showed that the top 10% of drinkers drank 55.3% of the total alcohol consumed, accounted for 61.6% of all 5+ and nearly 80% of all 12+ drinking days. Spirits consumption was the most concentrated with the top decile consuming 62.9% of the total for this beverage. This decile accounted for 33% of total expenditures, even though its mean expenditure per drink was considerably lower ($0.79) than the bottom 50% of drinkers ($4.75). CONCLUSIONS: The distributions of mean alcohol intake and heavy drinking days are highly concentrated in the U.S. population. Lower expenditures per drink by the heaviest drinkers suggest substantial downward quality substitution, drinking in cheaper contexts or other bargain pricing strategies. Empirical drink ethanol estimates improved survey coverage of sales particularly for spirits, but significant under-coverage remains, highlighting need for further self-report measurement improvement.     

The Role of the Flushing Response in the Relationship Between Alcohol Consumption and Insulin Resistance

Background: Facial flushing responses to drinking, because of intolerance to alcohol, are observed in some people, especially Asians. This study examined the role of flushing responses in the relationship between alcohol consumption and insulin resistance (IR). Methods: Participants in this cross‐sectional analysis included 624 Korean men (80 nondrinkers, 306 nonflushing drinkers, and 238 flushing drinkers) who were free of cardiovascular disease and diabetes. Data on the flushing response to drinking and alcohol consumption were collected from medical records. IR was estimated using the Homeostasis Model Assessment (HOMA_IR). On the basis of comparisons with nondrinkers, the risk of IR according to the quantity of alcohol consumed per week was analyzed among nonflushers and flushers. Results: After adjusting for age, exercise status, smoking status, BMI, waist circumference, blood pressure, high‐density lipoprotein cholesterol, and triglycerides using a logistic regression model, we found a low risk of IR among nonflushers who consumed ≤4 drinks (1 drink = 14 g of alcohol) per week (OR = 0.3). In contrast, a higher risk of IR was associated with nonflushers who consumed >20 drinks per week (OR = 3.5). On the other hand, only a higher risk of IR was associated with flushers who consumed >12 drinks per week (>12 to 20 drinks: OR = 4.7; >20 drinks: OR = 3.5). Conclusions: The amount of drinking associated with the development of IR in flushers was lower than in nonflushers. Additionally, no positive effect of moderate drinking on IR was observed in flushers. The findings support acetaldehyde‐derived mechanisms in the development of alcohol‐related IR.     

Phosphatidylethanol and Alcohol Consumption in Reproductive Age Women

Background: Fetal alcohol disorders are preventable, but self‐reported alcohol consumption can be misleading and impede effective treatment. Biomarkers represent an alternative method for assessing alcohol use, and this study evaluated the relationship between blood phosphatidylethanol (PEth) and alcohol use in a sample of reproductive age women. Methods: Alcohol use was estimated by validated self‐report methods in 80 nonpregnant women ages 18 to 35. PEth was measured by a contracted laboratory using a liquid chromatography‐tandem mass spectrometry assay. Regression methods appropriate for the distribution of PEth were used to define its relationship to alcohol consumption during the prior 2 weeks and explore the effects of drinking patterns on this association. Receiver operating characteristic analysis was used to estimate the sensitivity of PEth for various drinking levels at 95% specific cutoffs. Results: PEth had a positive linear association with grams of alcohol consumed (p < 0.001), and was detectable in 93% of subjects consuming an average of 2 or more drinks per day. The relationship between total alcohol consumption and PEth may be stronger in women with recent heavy drinking days. The relationship between drinking and PEth varied considerably between individuals, and sensitivity for a certain amount of drinking was low at a highly specific cutoff concentration. Conclusions: PEth is a highly sensitive indicator of moderate and heavy alcohol consumption in reproductive age women and may complement the use of self‐report alcohol screens when additional objective markers of alcohol use are desirable. However, choosing a highly valid cutoff concentration for PEth to differentiate various levels of alcohol consumption may not be feasible.     

Prospective study of moderate alcohol consumption and risk of hypertension in young women

BACKGROUND: Heavy alcohol consumption is associated with an increased risk of hypertension. However, the effect of moderate alcohol consumption; the specific effects of wine, beer, and liquor; and the pattern of drinking in relation to risk of hypertension among young women are unclear. METHODS: We prospectively examined the association between alcohol consumption and subsequent risk of hypertension among 70 891 women 25 to 42 years of age. RESULTS: During the 8 years of follow-up, 4188 cases (5.9%) of incident hypertension were reported. After adjustment for multiple covariates, the association between alcohol consumption and risk of hypertension followed a J-shaped curve. Compared with nondrinkers, the risk of developing hypertension according to average number of drinks consumed per day was as follows: 0.25 or less, 0.96 (95% confidence interval [CI], 0.89-1.03); 0.26 to 0.50, 0.86 (95% CI, 0.75-0.98); 0.51 to 1.00, 0.92 (95% CI, 0.82-1.04); 1.01 to 1.50, 1.00 (95% CI, 0.80-1.24); 1.51 to 2.00, 1.20 (95% CI, 0.92-1.58); and more than 2.0 drinks, 1.31 (95% CI, 1.02-1.68). Exclusion of past drinkers yielded similar results. Among women in the highest category of alcohol consumption, there was a suggestion that the increased risk of hypertension was present regardless of the specific beverage consumed (beer, wine, or liquor). Episodic drinking, defined as consumption of more than 10.5 drinks over 3 or fewer days per week, was not associated with increased risk of hypertension (relative risk, 0.80; 95% CI, 0.51-1.23). CONCLUSIONS: The association between alcohol consumption and risk of chronic hypertension in young women follows a J-shaped curve, with light drinkers demonstrating a modest decrease in risk and more regular heavy drinkers demonstrating an increase in risk.     

A statistical approach to an alcoholic drinking history

The drinking history of a middle-aged male was analyzed statistically on the basis of eight and a half years of notes on the number of drinks consumed per day. During the period his average number of drinks per drinking day increased from about 7 in 1974 to a peak of about 16 in 1980 while the number of abstinent days varied between 23% and 54% with no clear trend. These figures are of the same magnitude as published reports on drinking among alcoholics. Time-series models of intake or drinking frequency could not describe adequately the time-structure of annual or monthly consumption. Occurrence of drinking was analyzed as a random series of events. The time-structure of the series was highly irregular and deviated greatly from the Poisson hypothesis which assumes that each day has an equal probability of becoming a drinking day indepedently of previous days. Instead, drinking days were clustered into sprees with an average length of 7 days, high variance and a very skew distribution, separated by abstinence periods with an average length of 4 days and a similarly shaped distribution. The entire history could be partitioned into 286 alternating drinking and abstinence intervals, one day intervals included. The drinking rhythm was very stable: no significant trends in the lengths of either type of interval could be found. The main findings are the surprising stability of the drinking rhythm, its independence of the growing amounts consumed, and the independently varying abstinence interval lengths. Even in the absence of reporting and memory bias, such a pattern of drinking may produce very inaccurate recall of the actual long-term alcohol intake, if the recall period is short. The results suggest that periods shorter than one month should be avoided when asking questions about alcohol intake, for example, in research on the effects of treatment on alcoholism or alcohol intake on health.     

A comparison between brand-specific and traditional alcohol surveillance methods to assess underage drinkers’ reported alcohol use

Background: Adolescent alcohol consumption remains common and is associated with many negative health outcomes. Unfortunately, common alcohol surveillance methods often underestimate consumption. Improved alcohol use measures are needed to characterize the landscape of youth drinking. Objectives: We aimed to compare a standard quantity-frequency measure of youth alcohol consumption to a novel brand-specific measure. Methods: We recruited a sample of 1031 respondents across the United States to complete an online survey. Analyses included 833 male and female underage drinkers ages 13–20. Respondents reported on how many of the past 30 days they consumed alcohol, and the number of drinks consumed on an average drinking day. Using our brand-specific measure, respondents identified which brands they consumed, how many days they consumed each brand, and how many drinks per brand they usually had. Results: Youth reported consuming significantly more alcohol (on average, 11 drinks more per month) when responding to the brand-specific versus the standard measure (p?p?p?Conclusion: This study contributes to the field of alcohol and adolescent research first by investigating a potentially more accurate alcohol surveillance method, and secondly by promoting the assessment of alcohol use among adolescents vulnerable to risky alcohol use. Finally, our survey addresses the potential impact of alcohol marketing on youth and their subsequent alcohol brand preferences and consumption.     

Estimates of harm associated with changes in Swedish alcohol policy: results from past and present estimates

AIMS: (i) To compare actual developments of alcohol-related harm in Sweden with estimates derived prior to major policy changes in 1995 and (ii) to estimate the effects on consumption and alcohol-related harm of reducing alcohol prices in Sweden. DESIGN: Alcohol effect parameters expressing the strength of the relationship between overall alcohol consumption and different alcohol-related harms were obtained from ARIMA (Auto Regressive Integrated Moving Average) time-series analyses. MEASUREMENTS: Measures of Swedish alcohol-related mortality (liver cirrhosis, alcoholic psychosis, alcoholism and alcohol poisoning), accident mortality, suicide, homicide, assaults and sickness absence from 1950 to 1995. FINDINGS: Previous estimates of alcohol-related harm based on changes in alcohol consumption for the period 1994-2002 for Sweden were, in some cases (e.g. violent assaults and accidents), relatively close to the actual harm levels, whereas in other cases (e.g. homicides, alcohol-related mortality and suicide) they diverged from observed harm levels. A tax cut by 40% on spirits and by 15% on wine is estimated to increase total per capita alcohol consumption by 0.35 litre. This increase is estimated to cause 289 additional deaths, 1627 additional assaults and 1.6 million additional sickness absence days. CONCLUSIONS: The estimates of future changes in harm based upon even relatively modest increases in alcohol consumption produce considerable negative effects, with large economic consequences for the Swedish economy. The additional alcohol-related deaths, for instance, amount to more than half the number of yearly traffic fatalities in Sweden.     

Estimation of Alcohol Consumption from the Health and Nutrition Examination Survey

The National Health and Nutrition Examination Survey of 1971–1973 (HANES) includes four general questions on alcohol intake, whereas historically surveys of drinking have employed detailed questions on the quantity and frequency of drinking wine, beer, and liquor, with and without a measure of variability (QF/QFV); methods have been developed to express both QF and QFV in terms of absolute alcohol intake per day (AA). In order to compare the HANES data on alcohol consumption with data based on these more well-established measures, it is necessary to develop a measure of AA based on the HANES questions, HANESAA, that is comparable to AA based on the QFV questions (AAQFV). In constructing HANESAA, a subset of data from a Western New York State Survey of drinking which employed the QFV questions was used to estimate the responses to the HANES questions which might have been made if these questions had been asked in the same survey. Alcohol intake calculated from this “HANES subset” of data, AAHANES, was found to correlate highly with AAQFV based on the full set of QFV questions (r =. 93). Regressions of AAQFV on AAHANES were done to develop equations, the parameters of which were then used to predict HANESAA from the HANES data set. HANES data on drinking, expressed in terms of HANESAA, are compared with data from surveys conducted by Harris in 1972–1974 using QF measures of AA.     

Is Pregnancy the Time to Change Alcohol Consumption Habits in France?

Background:  Although it is well known that France has a cultural history of alcohol use, no recent French data on alcohol consumption during pregnancy in a large sample are available.
Methods:  To determine the alcohol consumption patterns among pregnant women in France, we analyzed data from a 1-year multicenter self-survey. Sociodemographic profile, obstetrical history, neonatal data, and a self-report for assessing drinking patterns during pregnancy including AUDIT were recorded from women who delivered recently. Cases of fetal alcohol syndrome (FAS) were also reported.
Results:  A total of 837 pregnant women have described all parameters. The mean age at delivery of our sample was 29.7 years (SD = 4.8 years). A total of 52.2% of women indicated that they had consumed alcohol at least once during their pregnancy, and among abstainers 54.5% had a positive AUDIT score. Of the pregnant women who consumed alcohol, 13.7% reported at least one binge drinking episode (5 or more drinks on 1 occasion) during pregnancy. Binge drinking is significantly more frequent than regular alcohol consumption (at least 1 drink more than 1 time per week) during pregnancy. A prevalence rate of FAS of 1.8 per 1,000 live births was observed.
Conclusions:  There is a large population of women who still drink alcohol during pregnancy, particularly in binge drinking episodes. This underlines the need to clearly inform women of childbearing age about the dangers of alcohol during pregnancy as related to all types of consumption. Moreover, acting to prevent alcohol consumption prior to pregnancy may also greatly influence prenatal drinking.     

Drinking Patterns and Myocardial Infarction: A Linear Dose–Response Model

Background: The relation of alcohol intake to cardiovascular health is complex, involving both protective and harmful effects, depending on the amount and pattern of consumption. Interpretation of data available on the nature of these relations is limited by lack of well‐specified, mathematical models relating drinking patterns to alcohol‐related consequences. Here we present such a model and apply it to data on myocardial infarction (MI). Methods: The dose–response model derived assumes: (1) each instance of alcohol use has an effect that either increases or decreases the likelihood of an alcohol‐related consequence, and (2) greater quantities of alcohol consumed on any drinking day add linearly to these increases or decreases in risk. Risk was reduced algebraically to a function of drinking frequency and dosage (volume minus frequency, a measure of the extent to which drinkers have more than 1 drink on days when they drink). In addition to estimating the joint impact of frequency and dosage, the model provides a method for calculating the point at which risk related to alcohol consumption is equal to background risk from other causes. A bootstrapped logistic regression based on the dose–response model was conducted using data from a case‐control study to obtain the predicted probability of MI associated with current drinking patterns, controlling for covariates. Results: MI risk decreased with increasing frequency of drinking, but increased as drinking dosage increased. Rates of increasing MI risk associated with drinking dosage were twice as high among women as they were among men. Relative to controls, lower MI risk was associated with consuming < 4.55 drinks per drinking day for men (95% CI: 2.77 to 7.18) and < 3.08 drinks per drinking day for women (95% CI: 1.35 to 5.16), increasing after these cross‐over points were exceeded. Conclusions: Use of a well‐specified mathematical dose–response model provided precise estimates for the first time of how drinking frequency and dosage each contribute linearly to the overall impact of a given drinking pattern on MI risk in men and women.