Alcoholism and alcohol misuse encompass a spectrum of injury that affects all the organs and tissues of the body [1,2]. It may represent the oldest form of injury known to mankind. Alcoholic beverages existed at least as early as 10,000 BC  and liver diseases related to its use have been recognized for almost as long . To this day, alcohol remains a major cause of diseases worldwide . From the medical point of view, alcohol leads to organ injury. Unfortunately, many individuals who misuse alcohol become symptomatic only when severe disease is already present. Patients with alcoholic problems often have coexisting dysfunction such as cardiomyopathy, skeletal muscle wasting, neuropathies, pancreatic dysfunction and parotid gland enlargement . The pharmacokinetics of alcohol determine the time course of ethanol concentration in blood after the ingestion of an alcoholic beverage and the degree of exposure of organs to its effects. The interplay between the kinetics of absorption, distribution and elimination is thus important in determining the pharmacodynamic responses to alcohol. There is a large degree of variability in alcohol absorption, distribution and metabolism, and elimination rate as a result of both genetic and environmental factors. The between-individual variation in alcohol metabolic rates is, in part, due to allelic variants of the genes encoding the alcohol metabolizing enzymes such as alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) . The ALDH2*2 and the ADH2*2 alleles, as well as the c2 allele of the cytochrome p450 2E1 (CYP2E1) gene are unique to Orientals. This prompted Sun et al.  to analyze their involvement in the drinking behavior in 322 middle-aged Japanese men. The ALDH2*2 allele showed a protective effect against a high level of alcohol consumption and problem drinking behavior, as determined by the Kurihama Alcoholism Screening Test. The ADH2*2 allele, present in 95% of individuals, also exhibited a suppressive effect on alcohol consumption. In contrast, the c2 allele of CYP2E1, present in 40% of individuals, was associated with greater alcohol consumption .
Possible factors that affect the development of alcohol-related injury include the dose, duration and type of alcohol consumption, drinking patterns, sex, and ethnicity [8,9]. In addition, there are associated risk factors such as obesity, iron overload, concomitant infections, genetic factors, as well as the interaction between therapeutics with alcohol [10,11,12,13,14]. In a study looking at the effect of alcohol consumption on isoniazid therapeutic efficacy in tuberculosis, a direct relationship between the self-reported amount of alcohol consumption and the incidence of hepatic injury was noted. This association becomes clearer when considering that CYP2E1, which metabolizes isoniazid, is induced by ethanol . Environmental and occupational hazards combined with alcohol consumption have to also be taken in consideration. In miners for example, a combination of alcohol abuse and arsenic exposure has been blamed for the occurrence of cirrhosis . In addition, other known hepatotoxins are dangerous in combination with alcohol consumption .
Geographic variability exists in the drinking patterns throughout the world [18,19]. Approximately two-thirds of adult Americans consume alcohol . The majority drinks small or moderate amounts, and do so without evidence of clinical disease [21,22,23]. A subset of drinkers consumes excessive amounts of alcohol, develops physical tolerance and withdrawal, and is diagnosed with alcohol dependence. On the other hand, alcohol abusers and problem drinkers engage in harmful alcohol consumption, defined by the development of negative social and health consequences such as unemployment, loss of family, transmission of infectious diseases, organ damage, and accidental injury [24,25,26]. The burden of alcohol-related disease is highest in the developed world, where it may account for as much as 9.2% of all disability-adjusted life years [27,28,29,30]. Even in developing regions of the world, alcohol accounts for a major portion of global disease burden, and is projected to take on increasing importance in those regions over time [26,31]. Alcohol-attributable mortality and burden of disease in Disability-Adjusted Life Years focused on estimating the burden attributable to alcohol consumption, including estimates of exposure (average volume of alcohol consumption and drinking patterns) and determination of risk relations. Moreover, using these estimates to determine alcohol-related burden of disease such as mortality, years of life lost due to premature mortality or due to disability, and disability adjusted life years are important methodologies. The spectrum of laboratory findings in individuals with alcoholic problems can delineate the specific amount and the period when alcohol was consumed. Rehm et al.  modelled the impact of alcohol dependence on mortality burden and showed that alcohol consumption can affect the available treatment.
Diagnostic Screening for Alcoholism
The relationship between alcohol consumption and biomarkers was determined in healthy volunteers who consumed controlled levels of alcohol. Reference levels of biomarkers of alcohol consumption are obtained in known teetotalers and individuals required to abstain from alcohol. Biomarkers can be correlated with alcohol consumption patterns in social drinkers. Finally, biomarkers are measured in individuals with alcohol abuse problems, and are used to monitor progress during alcohol withdrawal treatment and potential relapse. The various cut-off values used to delineate different drinking patterns must be considered in light of the specific method used in each study, based on validated methods and manufacturer recommendations for each assay.
Several screening questionnaires exist to establish the link between behavior and diagnosis of alcoholism, including the Kurihama Alcoholism Screening Test , the CAGE questionnaire [32,33], the Michigan Alcoholism Screening Test, and the Alcohol Use Disorders Identification Test (AUDIT) . The Michigan Alcoholism Screening Test is a longer test with 25 questions, making it less appealing as a screening tool [35,36]. There have also been comparisons between these methods [37,38]. The AUDIT, developed by the World Health Organization, is a shorter tool with 10 questions aimed to avoid cultural and ethnic bias. The degree to which these questionnaires have been validated varies, and their performance in selected populations also influences their accuracy . In special circumstances such as during evaluation for liver transplantation, in legal cases, or during alcohol withdrawal therapy, there is a need for specific, measurable analysis in patients who deny alcohol intake.
3. Breath Test
The incidence of alcohol-impaired driving was 1.29% in a sample of drivers receiving random breath testing around the city of Barcelona. Trends towards a higher incidence of impaired driving were noted on weekends, during the night, among men, and among drivers traveling with at least one passenger . The incidence of self-reported alcohol consumption (in the preceding 6 h) was 8.3% in a Brazilian sample. This rate was not corroborated by breath teats results due to a low proportion of drivers agreeing to this test . In the context of alcohol consumption, a breath test measures the alcohol level present in exhaled air. The breath alcohol concentration (BrAC) is then used to estimate the blood alcohol concentration (BAC). In cases in which driving offenses are committed, it is beneficial to measure the BAC at the earliest time possible. In most cases, the breath test provides the quickest results. A BAC of >0.05% will result in driver license suspension in the Canadian province of Ontario . An additional blood sample collected at the time of the offence could provide better measurements of BAC. These two techniques can be employed to complement one another . A breath test can further be used in the context of an ignition interlock device that measures the driver’s BAC in individuals with a history of driving under the influence (DUI). The vehicle will not start if the driver’s BAC, as measured by an in-car alcohol breath screening device, exceeds a pre-set limit (e.g., 20 mg alcohol/100 mL blood in the province of Ontario) [42,44]. Prevention programs are also in place. For example, trucks and public vehicles such as buses may be equipped with interlocks that require a breath sample from the operator at the beginning of travel . Moreover, a recent study has shown a poor relationship between how intoxicated many individuals perceive themselves to be and their actual BAC (assessed via breathalyzer), showing the importance of such devices .
A mean peak BrAC of 65 ± 19 mg/dL was identified between 20-35 min after alcohol ingestion in a sample of volunteer social drinkers . Among individuals involved in traffic accidents, being BAC positive (≥0.01%, measured in blood or estimated from breath test) was significantly associated with death compared to being BAC negative (p < 0.0001) . In order for blood and breath analysis to be interchangeable with each other, the relationship between BAC and BrAC needs to remain stable at all time points. However, the relationship between the two is variable during the absorption stage, and stabilizes during the post-absorptive stage, 60–90 min after alcohol ingestion .
A high correlation was found between BrAC and BAC in a sample of healthy volunteers with no history of alcoholism (r = 0.983, 97% sensitivity, 93% specificity) . However, this data is based on means obtained from a large cohort, in which the BAC/BrAC ratios for each individual show a high degree of heterogeneity. When taking into account individual cases, a poor correlation between BrAC and BAC is observed . The breath test thus represents a poor estimate of BAC in real-world situations. Of particular interest are drivers accused of DUI, in which a BAC of ≥0.08%, estimated from the results of the breath test, leads to immediate arrest. Okorocha  argues that basing BAC levels on breath test alone can result in innocent drivers getting arrested (overestimated BAC) and guilty drivers being allowed to walk free (underestimated BAC). Furthermore, Ashdown et al.  calls into question the sensitivity of some commercially available breathalysers. They measured the specificity, sensitivity, positive predictive value (PPV) and negative predictive value (NPV) of three commercially-available breathalyzer devices in a sample of adults who had consumed alcohol, and compared these results with those obtained from a reference police breathalyzer for the purpose of predicting which of them would be over the Unite Kingdom legal driving limit (BrAC 35 μg/100 mL). According to the reference police device, 18.3% of participants were at or over the legal driving limit, while the personal devices show 89.5%, 94.7% and 26.3% sensitivity, respectively, in detecting those individuals at or over the legal limit, showing that these devices vary considerably, and may falsely suggest that some individuals are safe to drive while in reality they may be too inebriated to do so safely .
A BAC/BrAC ratio of 2100:1, which is the standard ratio used in law enforcement, was achieved after 30 min in a sample of healthy volunteers, and this remained relatively stable through almost 3 h post-ingestion. This study used a novel breath analyzer that standardizes BrAC to the alveolar-air water vapour concentration . Using a breath analyzer that allows the measurement of alcohol in free exhalation in a small sample of healthy volunteers who drank 0.6 g alcohol/kg body weight, the BAC/BrAC ratio decreased over time (3318 ± 1657 at 2 min, 2514 ± 429 at 5 min, 2311 ± 225 at 10 min, 2246 ± 140 at 15 min, 2089 ± 99 from 30–167 min), with a mean ratio of 2251 ± 46 during the post-absorptive phase. A very good correlation between BrAC ×2251 and arterial BAC was observed (r = 0.998, p < 0.001), yet arterial BAC-time profiles of individual patients show a great degree of variability even beyond the 30 min mark when the BAC/BrAC ratio stabilized. In contrast, the correlation between venous BAC and BrAC was poor, with a ratio that fluctuated between 1834 and 3259 . A study conducted in 88 hospitalized patients (35 women and 53 men) shows that estimating BAC from BrAC (2100:1 ratio) would lead to underestimation of venous BAC by 26% compared to the actual measured values . Even using a conversion factor of 2260 led to underestimation of venous BAC by 15% compared to the actual measured values in a sample of drivers .
The breath test was a poor estimate of BAC in a sample of bar patrons. Generally, the breath test overestimated BAC in patrons who consumed alcohol only at the bar, and underestimated BAC in patrons who also consumed alcohol before getting to the bar . The presence of mouth alcohol was shown to contaminate BrAC readings . In a different survey, 90.9% of 227 British students attending pub crawls reported drinking prior to arriving at the bar. The median alcohol consumption was 10.0 alcohol units (80 g ethanol) at the time of interview, and it was estimated to exceed 16 by the end of the event. Median BAC among drinkers at the time of interview, measured as BrAC using breathalyser tests and then converted to BAC, was 0.10%. A high BAC was associated with not consuming food in the 4 h prior to interview (OR 1.2, p < 0.01), a longer time spent drinking (OR 1.4, p < 0.01), and the number of drinks consumed per hour (OR 1.2, p < 0.01) . Increasing levels of intoxication, as assessed by breath test, were noted late at night and in the early morning within night-time entertainment districts in an Australian study . Consuming energy drinks was found to have no influence on the subjective (self-reported) level of drunkenness or the objective (assessed using a breath test) intoxication in a sample of Dutch bar patrons. Similarly, consumption of energy drinks did not influence alcohol consumption .
An overall strong, positive correlation between the cumulative AUDIT-C score and BrAC reading (r = 0.416, p = 0.001) shows a strong correlation between a qualitative measurement of long-term hazardous drinking and current drinking . A breath test did not show sufficiently sensitivity in identifying self-reported heavy drinking in a cross-sectional sample of alcohol-dependent patients .
Relapse is routinely measured in patients under alcohol dependence treatment . Relapse was assessed by ethyl glucuronide (EtG) in urine, breath alcohol tests and self-reports in outpatients undergoing long-term alcohol dependence treatment. The percentage of patients showing alcohol relapse was 1.1% by self-report, 4.4% by breath test (range: 0.06–2.60 g alcohol/L), and 37.7% by EtG measurement (mean concentration 47.2 mg/L, range 0.2–1220 mg/L). A good agreement was observed between self-report and breath test, between self-report and EtG, and between breath test and EtG measurement. However, low discrepancies exist, and a high percentage of alcohol relapse cases (93.2%) were only identified by EtG measurement. In contrast, the breath test identified a few cases of alcohol relapse that were not identified by EtG measurement. In these cases, high alcohol concentrations (mean 1.24 g/L) were found, likely suggesting recent alcohol consumption following abstinence . This study shows that each method has its specific timeframe during which it is useful, and multiple methods should ideally be used together in order to identify immediate, short-term or long-term alcohol consumption.
3.1. Factors that Affect Breath Test Results
Variability exists in replicate breath alcohol exhalation profiles for one subject collected over a short time interval. There were no age or gender influences, while the breath exhalation volume and breath exhalation time also lacked significant associations . However, the breathing pattern seems to have an effect. These include measuring too early in the expiratory phase, shallow expiration or hyperventilation, or measuring hyperventilation under conditions of chilly ambient temperature. All of these factors give rise to underestimates of BrAC, compared to reference values, while the expired volume is kept constant .
The complex exchange of gasses and water in the blood vessels and the mucosa of the airways may further affect the BAC/BrAC ratio [50,65]. Estimating BAC from BrAC is based on the premise that exhaled air reflects the alveolar air alcohol concentration, which is thought to be in direct equilibrium with the blood in the pulmonary circulation. Several factors determine BrAC, most importantly body and lung physiology [66
A widely-used urine test, the EtG (ethyl glucuronide) test is a biomarker screening that detects the presence of ethyl glucuronide, a breakdown product of ethanol, in urine samples. It can also detect the presence of EtG in your blood, hair, and nails, although the urine test is the most widely used.
The purpose of an EtG test is to document required alcohol abstinence, but the urine test can only measure alcohol intake within the last one to three days.
How Long the EtG Test Can Detect Alcohol
You may be surprised to learn that after consuming alcohol, only about 0.5 percent to 1.5 percent of it is eliminated in your urine, and this is after undergoing a process called glucuronidation to form the breakdown product, EtG.
Even so, the EtG test is quite sensitive and can detect even low levels of alcohol. In fact, the EtG urine test can technically detect alcohol in the urine up to five days after consumption. That said, in studies of participants without alcohol use disorders, EtG has been detected in urine samples for up to 80 hours after heavy alcohol exposure, so up to three days is probably a more reasonable estimate.
Interpreting the EtG Urine Test
Due to the common use of EtG to confirm recent alcohol abstinence, the Substance Abuse and Mental Health Services Administration (SAMHSA) has suggested the following cutoff values, based on scientific research:
"High" positive EtG test (for example, >1,000ng/mL) may indicate:
- Heavy drinking on the same day or the previous day
- Light drinking on the same day as the test
"Low" positive EtG test (for example, 500 to 1,000ng/mL) may indicate:
- Heavy drinking within the last one to three days
- Light drinking within the last 24 hours
- Recent intense exposure to environmental products containing alcohol (within the last 24 hours)
"Very Low" positive EtG test (for example, 100 to 500 ng/mL) may indicate:
- Heavy drinking within the last one to three days
- Light drinking within the last 12 to 36 hours
- Recent exposure to environmental products containing alcohol
EtG Test Limitations
A problem with the EtG test is that it can produce a positive test from the mere exposure to alcohol that is present in many daily use products. Examples of environmental or home products that contain alcohol include:
- Foods prepared with or flavored with alcohol
- Cleaning products
- Breath sprays
- Hand sanitizers
- Hygiene products like antiperspirant
- Hair dye
In fact, there are hundreds of household products that contain ethanol, according to the National Library of Health's Household Products Database, and this could possibly lead to a false positive on the EtG urine test.
EtG Test Accuracy
In addition, SAMHSA lists EtG as a "highly" sensitive and specific alcohol biomarker. As a sensitive test, this means that the EtG test accurately at least 70 percent or more of the time detects a person who recently consumed alcohol. One recent study showed that for moderate to heavy drinking, this number jumps to 85 percent.
As a specific test, this means that the EtG accurately at least 70 percent or more of the time identifies people who did not recently consume alcohol. For moderate to heavy drinking, the above-mentioned showed that the specificity is 89 percent.
Application of the EtG Test
The test for EtG is widely used to detect alcohol abstinence in situations that do not allow drinking, including:
- Alcohol treatment programs
- A DUI or DWI program
- Liver transplant patients
- Schools or the military
- Professional monitoring programs (for example, airline pilots, healthcare professionals, attorneys)
- Court cases (for example, child custody)
It's important to note that the EtG test is not recommended for use in workplace testing programs as it does not measure current impairment from alcohol.
A Word From Verywell
All in all, the EtG test is considered a highly useful test for detecting recent alcohol consumption. But like any test, there is the possibility for a false positive. This is why a positive test should be confirmed either with another test or with verification from the person that he or she did indeed drink alcohol.
Hopefully, as the research on EtG and other alcohol biomarkers unfolds, a clearer cutoff value can be made in order to distinguish between true alcohol use and exposure to alcohol in environmental products.
Jastrzębska I, Zwolak A, Szczyrek M, Wawryniuk A, Skrzydło-Radomańska B, Daniluk J. Biomarkers of Alcohol Misuse: Recent Advances and Future Prospects. Przegla̜d Gastroenterologiczny. 2016;11(2):78-89. doi:10.5114/pg.2016.60252.
Jatlow P, O’Malley SS. Clinical (Non-forensic) Application Of Ethylglucuronide Measurement: Are We Ready?Alcoholism, Clinical and Experimental Research. 2010;34(6):968-975. doi:10.1111/j.1530-0277.2010.01171.x.
Shukla L, Sharma P, Ganesha S, et al. Value of Ethyl Glucuronide and Ethyl Sulfate in Serum as Biomarkers of Alcohol Consumption. Indian Journal of Psychological Medicine. 2017;39(4):481-487. doi:10.4103/IJPSYM.IJPSYM_71_17.
Substance Abuse and Mental Health Services Administration (SAMHSA). The Role of Biomarkers in the Treatment of Alcohol Use Disorders, 2012 Revision. Spring 2012;11(2).