A new meta-analysis has found that cognitive behavioural therapy for insomnia (CBT-I) significantly reduces sleep problems in people who engage in heavy drinking or have alcohol use disorder, supporting the wider use of sleep-focused therapy in addiction treatment. The findings were published in the journal Sleep Medicine Reviews.
The study reviewed eight clinical trials involving 426 adults with either problematic alcohol use or a formal diagnosis of alcohol use disorder. It found that participants who received CBT-I showed marked improvements in sleep that lasted for at least six months after therapy. Researchers measured changes using the Insomnia Severity Index and reported an average drop in sleep disturbance scores by more than five points immediately after treatment, suggesting clinically meaningful results.
“Many people with alcohol use disorder struggle with insomnia. It is a frustrating and vicious cycle. Some individuals may use alcohol as a sleep aid, which disrupts sleep and can contribute to the persistence of insomnia. Insomnia, in turn, has been shown to increase the risk of alcohol relapse,” said Cagdas Türkmen, a researcher at the University of Heidelberg.
Although CBT-I did not appear to significantly affect alcohol consumption or cravings across the trials, researchers said improvements in sleep were consistent and robust. This held true regardless of whether participants were trying to abstain from alcohol or were still actively drinking. These findings challenge previous assumptions that CBT-I should be delayed until individuals achieve sobriety, and instead support its use at earlier stages of alcohol recovery or even as a preventive intervention for high-risk drinkers.
“Although benzodiazepines are commonly prescribed and can be helpful in the short-term treatment of insomnia, they carry a high risk of dependence, especially for this group. This makes it essential to find safer, more effective treatments that are beneficial in the long run,” Türkmen added.
CBT-I includes a mix of strategies such as sleep restriction, stimulus control, and relaxation training. In all the trials reviewed, it was delivered either in person, through telehealth, or via digital platforms, with most treatments lasting five to eight weeks. Only one of the trials evaluated a digital version of CBT-I, but it showed promising results, suggesting this method could increase accessibility for people who may not have access to in-person therapy.
“Despite its effectiveness, accessibility remains limited due to long waiting times and a shortage of qualified therapists. As a result, digital CBT-I is gaining attention as a promising alternative,” said Türkmen. He noted that “online versions are already showing promising results and have the potential to reach far more people.”
While sleep benefits were clear, the results for alcohol-related outcomes were mixed. Measures such as alcohol craving, the number of heavy-drinking days, and alcohol-related psychosocial issues did not show statistically significant improvement compared to control groups. Researchers noted that variations in how these outcomes were measured and differences in participant goals, such as abstinence versus harm reduction, may have influenced these findings.
“Several factors may explain this. Many studies lacked sufficient participants to detect changes in drinking behaviour. Also, the studies included different types of people. Some were abstinent, while others were actively drinking, making it difficult to reach definitive conclusions. The timing of CBT-I and the participants’ motivation to reduce alcohol use also likely played a role,” Türkmen explained.
The study also highlighted several limitations in current research. Most of the included trials were conducted in North America and involved predominantly White and male participants, many of whom were military veterans. This raises questions about how broadly the results can be applied to other populations, especially women and ethnically diverse groups. Furthermore, there is a lack of longer-term follow-up data, making it unclear how durable the effects of CBT-I are beyond six months.
Despite these limitations, the authors say their findings strengthen the case for integrating sleep therapy into addiction services. They emphasised the need for future research to assess how improvements in sleep might lead to reductions in alcohol use over time, perhaps through indirect pathways such as improved emotional regulation or stress coping.
Türkmen concluded, “Still, the key takeaway is clear. CBT-I is effective for treating insomnia in people with alcohol use disorder and should be offered as a treatment option.”
He also encouraged more research to validate these results, particularly among individuals with alcohol use disorder. “Further research is needed, especially among individuals with AUD,” he said. “Stay tuned for updates on ongoing and future research projects from the Trondheim Sleep Group at St. Olavs Hospital in Norway.”
