Citalopram and escitalopram have been used in the treatment of alcohol dependence especially when co-morbid with major depressive disorder. Treatment outcomes with other SSRIs have not been consistent [3, 15, 26, 50–53]. In our study, both memantine and escitalopram patients reported reduced alcohol craving and consumption, and patient compliance was good. However, the two study groups did not differ in alcohol craving, obsessive thoughts of drinking, compulsive drinking, alcohol consumption, maintaining abstinence, and number of abstinent days per week. Our study corroborates a recent study by Krupitsky et al. , who reported that memantine reduced alcohol cue-induced craving in recovering alcoholics.
Few possible predictive elements for the treatment of alcohol dependence comorbid with major depressive disorder with either escitalopram or memantine were observed. The abstinence at the beginning of the treatment predicted more likely the continuing of the treatment, especially in the memantine group. Evans et al.  found no effect of memantine in patients who were actively drinking at the beginning of treatment, which may explain our findings that such patients had a higher dropout rate than those who were abstinent at the start of treatment. Other predictor observed was the early age at onset of the first depressive episode, which leads to poor treatment outcome with escitalopram but not with memantine. However, due to multiple testing without a priori hypothesis, this difference should be interpreted with caution.
Our study has several limitations. It is a comparative study of two medications and is limited by the absence of a placebo group. Both treatment groups improved significantly and a placebo-effect could be significant in both groups, as has been observed in earlier studies . Spontaneous recovery and intermittent periods of lower alcohol intake are of typical in alcohol dependence [54, 55]. Therefore, we cannot determine whether the overall improvement in the present study was due to the medications, and our interpretations are limited to comparisons between memantine and escitalopram. Detoxification and a certain period of abstinence could influence the results.
Another limitation is the rather small number of patients, which may have been too low to detect a significant difference between the treatments. Socio-demographic indicators correspond well to those generally found among patients treated at Finnish A-clinics , suggesting that the present material represents a relatively unbiased sample. The only difference observed was the higher percentage of women, which can probably be attributed to the inclusion criterion of major depression. We did not attempt to distinguish between patients with primary depressive disorders and substance-induced depression. This situation corresponds to that at the onset of treatment; it is when the clinician has to decide which medications to prescribe. Our patients were treatment-seeking, so the option of providing no active medication was not accepted by either the treating professionals or the patients.
Our finding may suggest that memantine could be useful treatment for type one alcoholics (early onset) comorbid with depression. The finding that the early onset of the first depressive episode is a negative predictor for escitalopram treatment in alcohol dependence confirms our previous finding on treatment of this comorbidity regarding major depressive disorder in patients with this dual diagnosis .
It may be concluded that memantine seems comparable to escitalopram and could be used as well as escitalopram in the treatment of alcohol dependence comorbid with major depression. The results, therefore, warrant further studies of memantine in patients with alcohol dependence comorbid with major depression as well as of predictive signs of treatment outcome. However, because of the small number of study patients and the conventional statistics, the results should be taken with caution.