We all want to know how much alcohol our patients are drinking. Asking patients directly about their drinking often generates useful information and helps build rapport, but this tactic is not always reliable. Denial and shame as well as issues related to secondary gain can sometimes contaminate self-report. Thankfully, there are many measures that can help add objectivity to the assessment and monitoring of alcohol use, which will be reviewed below. It is important, however, to keep in mind that no current test is ideal, and that quantity of drinking is not sufficient to diagnose alcohol use disorder (AUD). Thus, all tests should be interpreted within the greater context of the patient’s history. It is also important to keep in mind the purpose of administering such a test prior to ordering it. Are you monitoring for relapse in patients with newly established abstinence? Are you identifying patients with heavy drinking? Or are you screening patients to determine if they need further investigation into their alcohol use history?

Direct alcohol testing
Although 95% of alcohol (ethanol) is metabolized to acetaldehyde via alcohol dehydrogenase, a little less than 5% is excreted unchanged in urine, breath, and sweat, and can thus be measured directly via those sources as well as blood. Because of alcohol’s short half-life, a disadvantage of direct testing is the small window of detection; these tests will only be useful for several hours following the most recent drink. Nonetheless, the tests are useful in medical contexts and in screening patients for relapse.

Blood alcohol testing
The blood alcohol concentration (BAC) of an individual roughly translates to the individual’s degree of intoxication, and there are several standards that relate BAC to specific levels of impairment (https://awareawakealive.org/educate/blood-alcohol-content).

Measurement of BAC is somewhat confusing, since many articles and textbooks refer to “percent” alcohol concentration, whereas many emergency rooms will report a simple (and usually quite large) number. These are actually the same thing—both express the amount of alcohol per unit of blood. When you see an alcohol level less than 1, it is being reported as mg of alcohol per mL of blood—hence the legal limit of 0.08% means 80 mg per 100 mL of blood. When you talk to the emergency room nurse, you hear that your patient has an alcohol level of “80,” which is 80 mg of alcohol per dL (deciliter, or 0.1 liter) of blood.

The blood alcohol level peaks an hour or two after a drink (or after many drinks in our patients). Because alcohol is metabolized in a predictable and linear way, you can reasonably predict that your patient’s BAC will drop by about 1 standard drink per hour (which is 0.02%–0.04% per hour depending on the person’s weight and body fat percentage).

Urine alcohol testing
Urine alcohol content (UAC) is frequently measured in the emergency room in addition to BAC because it is less invasive and can be assessed serially. UAC levels peak 45–60 minutes after ingestion and are about 1.3 times the BAC.

Breathalyzer
Breath testing (ie, breathalyzer) is probably the most popularized test and carries the distinct advantages of being inexpensive, noninvasive, and rapid. Breathalyzer levels are directly normed to BAC levels and, like BAC, roughly indicate degree of intoxication. Newer developments include smartphone apps with attached breathalyzers that allow patients or family members to do BAC testing (various devices are available on https://www.amazon.com for around $100).

Commonly ordered blood tests
The following common blood tests are all indirect markers of alcohol use, indicating organ damage or other secondary pathophysiological changes:

• Gamma glutamyl transferase (GGT). GGT is a glycoenzyme on endothelial cell membranes of various organs, especially the liver. Increased levels of GGT may indicate that a patient has been drinking heavily and continuously for a period of several weeks. Since GGT levels take 2–6 weeks to normalize, patients will need a decent length of sobriety before they produce a normal GGT, assuming the elevation was caused by heavy alcohol use.


• Aspartate amino transferase (AST)and alanine amino transferase (ALT). Elevations in AST and ALT may reflect liver damage caused by alcohol. An AST:ALT ratio of 2:1 is more indicative of alcohol-induced liver damage compared to other sources of liver injury. This measure is not as specific to drinking as GGT.


• Mean corpuscular volume (MCV). Chronic heavy alcohol use can cause MCV elevation, both because of alcohol’s detrimental effect on erythroblast development and as a secondary effect of vitamin deficiencies sometimes seen in severe AUD. It is not as sensitive to alcohol use as other biomarkers.

Carbohydrate-deficient transferrin (CDT)
Transferrin is an iron transport protein that in the presence of alcohol loses some of its carbohydrate side-chains. Thus, the specific “carbohydrate-deficient” transferrin indicates alcohol use. Only prolonged heavy drinking (ie, greater than 4–5 drinks a day for 2–3 weeks), however, is likely to produce an elevated CDT level. So, although this test is well utilized to identify heavy drinkers, it is not good for detecting relapse or moderate drinking. There are only a few rare causes of false positives (eg, inborn error of glycoprotein metabolism, primary biliary cirrhosis), so the test is quite specific, and is more specific to alcohol use than the GGT we discussed earlier.

Ethyl glucuronide (EtG) and ethyl sulfate (EtS): Improved urine testing for alcohol
Although the majority of alcohol is either metabolized to acetaldehyde or excreted unchanged (see above), a tiny fraction is alternatively converted to EtG and EtS. Both of these minor metabolites can be detected in the urine and have good sensitivity and a longer half-life than alcohol. Each test performs well individually, but they should be ordered together.

If a newly abstinent individual consumes even a minor amount of alcohol, EtG and EtS can detect the relapse for 1–2 days after consumption, whereas breathalyzer and other alcohol tests would likely miss this development. In fact, EtG in particular is so sensitive that it may detect incidental exposure to alcohol, such as through use of alcohol-containing mouthwash or hand sanitizer. Although the detection window for EtG and EtS is sometimes reported as 3–4 days, only high levels of alcohol consumption (ie, greater than 10 drinks) can typically be detected in that time frame.

Phosphatidylethanol (PEth)
PEth is an abnormal glycoprotein formed only after exposure to alcohol, making it extremely specific. The sensitivity is also quite good, especially as new analytic methods are improving its ability to detect low and moderate drinking (Helander A et al, Alcohol and Alcoholism 2012; 47(5):552–557). PEth can be detected in blood 1–2 hours after alcohol consumption and remains positive on average for 4 days, although greater lengths of 7–12 days from a single intake have been reported. You can think of PEth as almost a cross between EtG and CDT. It can be useful in differentiating abstinence from relapse, yet is not as sensitive as EtG. And like CDT, PEth is adept at differentiating heavy drinkers from moderate drinkers, yet it has a shorter half-life than CDT. This biomarker is not yet routinely used, but it is offered by many labs and has great potential.

CATR Verdict: Biomarkers can be a helpful source of objective data in the context of a full clinical picture. EtG/EtS is the preferred test to differentiate abstinence from relapse, and CDT is a good way to differentiate chronic heavy drinking from more moderate drinking. Breathalyzer, GGT, AST, ALT, and MCV can be used as screening tests. In the future, PEth may be the best overall option.