Alzheimer’s disease (AD) is a major type of dementia and poses a growing global health burden (1). Population aging and growth are projected to drive further rises, with global dementia cases expected to grow from 57 million in 2019 to 152 million by 2050 (2). As the predominant subtype of dementia (accounting for 60–80% of cases) and the most common neurodegenerative disorder, AD ranks among this century’s most lethal, costly, and debilitating diseases (3). Its core features include progressive cognitive decline (involving memory, language, and other domains), which severely impairs daily functioning, along with a spectrum of neuropsychiatric symptoms such as depression, agitation, apathy, anxiety, and psychotic features (4, 5). These multidimensional impairments not only accelerate disease progression and increase caregiver burden but also highlight the urgent need for effective therapeutic strategies.

Recent years, disease-modifying treatments such as donanemab and lecanemab for AD have been approved by the Food and Drug Administration (6). However, the effects of these treatments remain limited by multiple factors, including the stage of the disease, potential side effects, cost implications, and individual variations (7). Therefore, non-pharmacological strategies that can delay the progressive functional deterioration in AD may be a reasonable alternative, encompassing approaches such as music therapy, exercise interventions, repetitive transcranial magnetic stimulation (rTMS) and cognitive interventions (8). As a specific form of cognitive intervention, Cognitive Behavioral Therapy (CBT) focuses on dynamic interaction between cognitive processes, emotional responses, and behavioral patterns (9). CBT was initially mainly applied in the treatment of mood disorders such as depression and anxiety (10). Studies demonstrated that individuals suffering from dementia can learn and develop skills (11). This suggests that CBT, as an effective psychological intervention method, can be introduced into the treatment and care of AD (12, 13). In the context of the ongoing progress in interdisciplinary research integrating geriatrics, neurology, and psychology, the application of CBT in AD gradually entered a more clinically refined stage (14). In non-pharmacological treatment strategies for AD, CBT is often integrated with other interventions such as cognitive training, mindfulness and home-based occupational therapy, forming comprehensive and multifaceted intervention programs (15–17).

However, the empirical evidence supporting CBT’s efficacy in AD has been mixed. Systematic reviews have yielded inconclusive findings. For instance, a 2022 review pooling people living with dementia (PLWD) across etiologies found psychological therapies (often including CBT components) reduce depression and anxiety, but effects on global cognition, activities of daily living, and quality of life remain inconsistent (18). Earlier reviews of “cognitive therapy” in dementia (spanning multiple modalities beyond CBT) also judged the evidence inconclusive due to variability in interventions and outcomes (19). Therefore, the present meta-analysis aims to systematically evaluate the effects of CBT on cognition, neuropsychiatric symptoms, and quality of life in patients with clinically diagnosed Alzheimer’s disease, using data from randomized controlled trials (RCTs).

Given the established efficacy of CBT (cognitive behavioral therapy) in treating mood disorders such as depression and anxiety (10), its techniques have been progressively adapted to address the specific needs of AD patients experiencing cognitive decline. This systematic review and meta-analysis assessed the effects of CBT on global cognition, depression severity, neuropsychiatric Symptoms, and life satisfaction in elderly patients with AD. The pooled results suggest that CBT may be an effective intervention to improve cognitive decline in AD. Furthermore, subgroup analyses indicated that CBT cycles of 8–16 weeks were beneficial in improving depression and quality of life in AD patients. These findings provide important evidence for the potential of CBT to address multiple domains of impairment in AD and warrant further exploration in future clinical applications.

Regarding cognitive outcomes, CBT intervention groups showed significant cognitive improvement compared with the control group after treatment, and subgroup analyses showed that long-term CBT was particularly beneficial. When the observed SMD (0.67) is translated into MMSE points using the pooled SD, it corresponds to an estimated 2.87-point improvement, suggesting clinically meaningful benefits of CBT for cognitive function in AD patients (see Supplementary Table S6 for details). These results are consistent with previous findings from other psychosocial interventions. An umbrella review found that cognitive interventions consistently improved MMSE scores in dementia (8). Likewise, an evidence synthesis noted that enhanced brain plasticity underpins such improvements and that cognitive training showed a small but significant global cognitive gain (48). These findings are generally in line with observation of improved MMSE scores post-CBT intervention. Neuroimaging evidence shows that such training boots functional connectivity in learning and memory networks even in AD patients. For example, Behfar and colleagues found that cognitive stimulation therapy can lead to a significant improvement in MMSE scores and an increase in resting-state connectivity in memory region (49). These results prove that neuroplasticity can be harnessed to alleviate cognitive impairment in AD. In other words, repeated cognitive exercises may strengthen connections at the synapse or recruit compensatory networks, thus enhancing cognitive reserve. This plasticity-based mechanism aligns with our findings of cognitive benefit after CBT interventions. Beyond cognition, neuropsychiatric symptoms and quality of life are also important domains to consider when evaluating the impact of CBT in AD.

Meta-analyses and systematic reviews also indicate that CBT-based treatments confer modest mood and improve quality of life. A Cochrane review found that CBT which includes behavioral activation and problem-solving therapy probably slightly reduces depressive symptoms in dementia (SMD =–0.23) and modestly improves quality of life (SMD = 0.31) (18). A recent systematic review similarly found that CBT reduces depression in AD patients, although certainty is limited due to heterogeneity (50). These findings are in agreement with results from our subgroup analyses which showed medium-term CBT lowered CSDD scores and improved QoL compared to long and short-term CBT. Interventions based on CBT work by identifying and modifying maladaptive thoughts and behaviors. By challenging negative automatic beliefs and teaching problem-solving or activity-planning skills, CBT reduces depressive effect and increases engagement in adaptive behaviors. In dementia care, this means helping patients reinterpret stressors and re-engage in something more meaningful. As a result, it yields small improvements in mood and daily activities. This mechanism accounts for reductions in CSDD scores and improvement QoL in our meta-analysis of CBT.

Despite these encouraging findings, several limitations should be acknowledged when interpreting the results of this meta-analysis. First, the number of studies was small (n=15), which reduces statistical power and the robustness of conclusions. The issues include difficulty in recruitment, treatment fidelity, and compliance. In addition, the presence of several small-sample trials raises concern for small-study effects: while the MMSE funnel plot did not display clear asymmetry on visual inspection, Egger’s test indicated a statistically significant bias, suggesting that smaller studies with positive results may be overrepresented. Consequently, the magnitude of the positive effects observed could be overestimated and our results should be interpreted cautiously. Second, significant heterogeneity was observed in among trials (MMSE: I²=86.9%). This variability may stem from difference in study populations (age, gender distribution, dementia severity), intervention protocols (frequency, delivery format), control conditions, methodological quality (inadequate allocation concealment, limited blinding) and sample size variation. To better understand this, we summarized potential sources across studies ( Supplementary Table S7 ). Besides, the use of MMSE, which is a relatively crude screening tool, may have further amplified inconsistencies and contributed to potential publication bias. Third, while subgroup analyses suggested improvements in depressive symptoms, global neuropsychiatric outcomes (NPI) did not show significant changes, suggesting that CBT may have limited effects on broader neuropsychiatric symptoms in AD. One possible explanation is that CBT primarily targets mood-related symptoms such as depression and anxiety. By contrast, other domains captured by the NPI, such as delusions, hallucinations, and disinhibition, are more closely linked to neurodegenerative processes and may therefore be less responsive to psychological interventions. In addition, limitations of the NPI itself should be considered: the tool relies on caregiver reports, may lack sensitivity to subtle improvements, and its composite scoring may dilute specific benefits of CBT on mood-related symptoms. Finally, most trials failed to adequately address allocation concealment, and only two implemented participant blinding. The inability to blind participants is a common challenge in psychological intervention studies (51). Because psychological interventions are difficult to disguise, participants often recognize their group allocation, which increases the risk of expectancy bias. These challenges in participant blinding underscore the broader importance of rigorous trial design, including proper allocation concealment. In our study, both participant blinding and allocation concealment were difficult to achieve. This is particularly concerning subjective outcomes such as depression and quality of life, where empirical evidence suggests that lack of blinding and inadequate allocation concealment can lead to exaggerated intervention effect estimates (52). Collectively, these factors led to downgrading the certainty of evidence; according to the GRADE framework, the certainty of evidence ranged from low to moderate across different outcomes, and the results should therefore be interpreted with caution.

To our knowledge, systematic evaluations of CBT in AD through meta-analyses are relatively few, and subgroup analyses considering treatment duration have seldom been explored. By including a comprehensive literature search and strict selection of RCTs, this study provides preliminary evidence on the effects of CBT on cognitive and neuropsychiatric outcomes. With limited drugs available to postpone cognitive decline and manage neuropsychiatric symptoms in AD, these findings indicate that CBT may be a useful, low-risk, and scalable psychological treatment for AD. Importantly, the blinding of outcome assessment was reasonably maintained. In addition, both attrition bias and selective reporting appeared limited. Despite the presence of significant statistical heterogeneity, sensitivity analyses demonstrated that the findings for all outcomes remained stable, which collectively supports the reliability of these results. It is reasonable to consider that incorporating structured CBT programs into standard care may provide clinicians with a practical approach to enhance patients’ cognitive engagement and emotional well-being. Given remaining limitations in the evidence base, future research should prioritize larger, methodologically rigorous trials with extended follow-up periods. These studies are needed to more conclusively determine the optimal type and duration of CBT interventions for improving or maintaining cognitive function, emotional well-being, and quality of life in patients with AD.

Early cognitive decline is the hallmark of AD, which imposes a substantial emotional and physical burden on patients and caregivers. This meta-analysis provides preliminary evidence that CBT may be effective in slowing cognitive decline in patients with AD. Subgroup analysis suggests that CBT interventions lasting 8–16 weeks may also help improve depressive symptoms and quality of life. However, the findings should be interpreted with caution due to the limited number of studies, small sample sizes, and considerable heterogeneity in population, such as age and gender distribution. Additionally, the potential for publication bias and unclear blinding procedures in several trials may have influenced the outcomes. Despite these limitations, CBT appears to be a promising, low-risk non-pharmacological approach for managing cognitive and emotional symptoms in AD. Future high-quality, large-scale randomized controlled trials using standardized CBT protocols are needed to confirm these findings and inform clinical application.