A randomized, placebo-controlled(RCT), double-blinded trial studied whether citalopram was effective intreating alcohol dependence in subjects over a 12-week period. Subjects wereincluded if they were between the ages of 18 and 65 and diagnosed with alcoholabuse or dependence by the Diagnostic andStatistical Manual of Mental Disorders (DSM-IV).
The exclusion criteria was females that werepregnant or breastfeeding, history of serious adverse reactions/intolerance toSSRIs, taking other psychiatric medications (including SSRIs), abuse of asecond substance (not nicotine) and had a psychotic/organic brain disorder, andif psychiatric administration/inpatient detoxification was required.1The primary objective was tocompare citalopram 20 mg per day, titrated up to 40 mg per day after 2 weeks,and placebo, and how they affected a subject’s number of heavy drinking days, abstinencelength, and mean drinking days. Their secondary objective was to track subjects’mental health by evaluating a subject’s depression via the Beck DepressionInventory (BDI) and anxiety via Beck Anxiety Inventory (BAI). Patients werefurther categorized as complete, partial, or no response.1 The sample consisted of 265 (n =265) subjects, of which only 141 (53%) completed the trial and 14(5%) werepulled out due to either severe adverse events or psychiatric symptoms. Thenumber of heavy drinking days in the placebo group and citalopram group were4.78 ±0.68 and 7.
60 ±0.85 (p = 0.007) respectively.
The mean number of drinking days inplacebo and citalopram groups were 3.60 ±0.55 and 5.
37 ±0.72 (p = 0.030) respectively.
The percentdecrease in the number of drinking days in the placebo and citalopram groupswere -74.72 ±3.75 and -57.92 ±6.
14 respectively (p = 0.016. The percent of subjects that were abstinent at the endof the trial in placebo and citalopram groups were 56.6% and 50.7% respectively(p = 0.416).
There was a decrease indepression (BDI = 18.60±0.69 vs. 11.
65±1.12) and anxiety (BAI = 18.99 ± 0.77vs. 11.00 ± 1.
41), but they were not statistically significant since theirp-values were 0.841 and 0.800 respectively. Overall, the authors concluded thatnot only was citalopram ineffective in treating alcohol dependence, but it alsoled to worse outcomes than placebo.1 All studies have their limitations and strengths. The majority ofthe subjects that received the diazepam therapy for alcohol detoxification werein the citalopram group.
Since diazepam has prolonged action in obese patients andit can interact with alcohol, this is a possible cause of adverse events sinceBMI was not calculated.2 The sample also consisted of 92%Caucasians, majority of whom were male (69.8%). Both factors decrease externalvalidity. Perhaps the biggest limitation of this trial was that the authorsencouraged, but did not require, their subjects to attend Alcoholics Anonymous.
Since some subjects received additional behavioral interventions that were notbeing tracked, it turns into a confounding variable. Overall, the trial waswell designed since it was a double-blinded RCT, which decreases the possibilityof bias and generates high-level of evidence. The use of validated andreputable tests for assessments and analysis made the results relevant. Theauthors also allowed subjects to concurrently take their other medicationsduring the trial, aiding to eliminate the confounding variable of worsening concurrentdisease states. Ultimately, it is not clinically useful to recommend citalopramfor the treatment of alcohol dependence.A separate randomized, placebo-controlled, single-blinded trialstudied the variation of effects of citalopram on both alcohol dependence inmen and women over a 12-week period. Subjects were included if they drank atleast 28 drinks per week for at least the past 3 months, were socially stable,medically stable, and were diagnosed with mild to moderate alcohol dependencevia the Alcohol Dependence Scale (ADS) and the Diagnostic and Statistical Manual of Mental Disorders, 3rdedition, revised (DSM-III-R). If a subject was diagnosed for an anxietydisorder, clinical depression, other drug/substance dependence, any otherpsychiatric condition that required treatment, or if they responded to theplacebo during the 2-week baseline, they were excluded from the trial.
3The primary objective was to compare citalopram 40 mg per day withbrief psychosocial interventions (BPI) and a daily placebo with BPI, and howeach arm affected a subject’s alcohol dependence based on sex. The primaryoutcomes measured were the percent change in drinks per day, drinks perdrinking day, and total percent of abstinent days. The secondary objective wasto determine the differences between gender groups, which were done via a 2-wayANOVA and Duncan’s post hoc.3 A sample of 99 subjects (n = 99) was randomized, of which only 61(61.6%) completed the trial and 16% discontinued due to adverse reactions.Women and men in the citalopram group observed decreases of 27.42% and 43.99% respectively(p < 0.
05) in percent decrease indrinks per day, decreases of 20.66% and 31.15 respectively (p < 0.10) inpercent decrease in drinks per drinking day, and increases of 34.92% and 39.92%respectively in total percent increase of abstinent days.
In the placebo group,the percent decrease in drinks per day for women and men was 39.87% and 38.04%respectively, the percent decrease in drinks per drinking day for women and menwas 28.87% and 23.20% respectively, and the total percent increase in abstinentdays for women and men was 67.30% and 58.52% respectively. Men in the citalopramgroup observed larger reductions in drinks per day compared to women (p < 0.
05).The authors concluded that the use of citalopram for alcohol dependence is lesseffective than BPI alone, but has a bigger impact on men than women.3There are some limitations to consider. First off, this was asingle-blind trial, which introduces the possibility of bias from theresearcher side, thus decreasing internal validity. The authors also startedall subjects on 40 mg of citalopram from day 1 instead of starting at 20 mginitially and then titrating up to 40 mg, which is what was done in the othertrial and mentioned in the guidelines.1, 4 This is a possible sourceof adverse events and also decreases internal validity.
The study was supportedby a grant from Lundbeck, a manufacturer of citalopram, which introduced apossibility of bias, thus decreases internal validity.5 Despitethese limitations, the trial was a RCT, which provided high-level evidence. Theauthors conducted proper medical/lab tests before starting the trial in orderto make sure that each subject was eligible to take citalopram, thus increasinginternal validity. The study used validated tests, which made their resultsrelevant to the public.
Based on the inconclusive results of this trial, it isnot clinically useful to recommend citalopram for alcohol dependence in eithermen or women.