We are students at the Equator Publication School in Oxford and have spent the last week learning to use guidelines to improve study reporting. Today we have also spent 3 hours working with André from the Mental Elf to write a blog about a recent systematic review looking at the use of antidepressants to treat depression in people with Alzheimer’s Disease. Our blog highlights some issues with the review and how it was reported.
Depression is one of the most common neuropsychiatric symptoms within Alzheimer’s Disease and up to half of people with Alzheimer’s experience depression or clinically significant depressive symptoms. Antidepressants are recommended by NICE as a treatment for depression in people with Alzheimer’s Disease, but NICE also recommend a number of psychosocial interventions including cognitive behavioural therapy (with the active participation of their carers), reminiscence therapy, multisensory stimulation, animal‑assisted therapy and exercise.
Previous reviews of the effectiveness of antidepressants in Alzheimer’s Disease have been inconclusive. The Cochrane review has not been recently updated (Bains et al, 2002), so this systematic review attempts to provide more reliable evidence to help.
Through this blog we would like to engage with a wider audience of clinicians and members of the public about their own experiences of depression in Alzheimer’s Disease and its management. Please share your thoughts with us on Twitter or via the comments below.
The researchers searched the major medical databases and major trials registers for published and ongoing research on depression in Alzheimer’s Disease dementia. They also searched the reference lists of included studies and reviews of depression in Alzheimer’s Disease and other dementias.
Studies were eligible if they were:
- Randomised trials comparing any single antidepressant with placebo
- Participants had a diagnosis of Alzheimer’s Disease
- The Alzheimer’s Disease patients also had a diagnosis of major or minor depression confirmed by a validated measure
- Response to treatment (dichotomous)
- Mean depression scores
- Number of drop outs
- Number of adverse events
The quality of the included studies was assessed and summarised using the Cochrane Risk of Bias tool and the GRADE approach.
From the 3,822 records identified from the database and additional searches, the screening and assessment process identified 7 completed and 1 ongoing study eligible for inclusion. There were 3 other eligible trials identified. One had been terminated and 2 were completed but had not published the results.
Response to treatment at 6-13 weeks and 24-39 weeks (dichotomous outcome)
- Six studies (520 patients) contributed data on short term response with no significant difference in the odds of response between the antidepressant and placebo groups. The odds ratio (OR) was 1.95 (confidence interval (CI) 0.97 to 3.92). There was significant heterogeneity between the studies.
- Only two studies (359 patients) looked at longer term response and also showed no difference between the treatment and placebo groups.
Mean depression scores at 6-13 weeks and 39 weeks (continuous outcome)
- Five studies (311 patients) contributed data at 6-13 weeks and there was no difference between the standardised means of the treatment and placebo groups -0.13 (CI -0.49 to 0.24). There was moderate heterogeneity between the studies.
- One study looked at the depression scores at 39 weeks and also found no difference -0.00 (CI -1.77 to 1.77).
- For cognition measured using MMSE scores five studies (251 patients) contributed data. There was no difference between the treatment and placebo group at 6-13 weeks 0.14 (CI -1.65 to 1.93)
- The number of drop outs was similar between the treatment and placebo groups
- More adverse events were reported by patients in the intervention group with all seven studies (632 patients) contributing data. However, there was no significant difference overall.
Quality of the evidence
The authors judged that bias is more likely to have occurred by inadequate sequence generation and blinding, and judged the overall quality of the evidence using the GRADE approach as “moderate”.
The reviewers concluded:
Despite the importance of depression in people with AD [Alzheimer’s Disease], few RCTs are available on efficacy of antidepressants, limiting clear conclusions of their potential role. There is a need for further high quality RCTs.
Strengths and limitations
This is a good systematic review, but as a group we feel that there are a number of important limitations, both of the primary literature and the review itself:
- We couldn’t find any subgroup analysis of settings, types of drugs, levels of depression and compliance. If these were done, they don’t seem to have been reported.
- Given the amount of heterogeneity in the clinical trials included in the review, further exploration of subgroups would be useful. We also wondered if meta-regression might be a useful tool for this analysis.
- Diagnostic tools for depression have not all been validated in people with Alzheimer’s Disease, and this might be a problem in different settings. We speculated about how difficult it must be to reliably measure depressive symptoms in people with Alzheimer’s Disease, particularly as the illness progresses.
- It’s not completely clear from the paper how different outcome measurements have been dealt with. Detail that may be available in a data supplement.
- We struggled with the way the paper was reported, which was perhaps down to a word limit in the journal. In any case, some of these gaps in information could have been solved by using the PRISMA statement for reporting systematic reviews.
Overall though, we agree with the reviewers that further randomised clinical trials are needed with larger samples of dementia patients, before we can be confident about prescribing antidepressants in this population.
Implications for practice
In our group, we had a considerable amount of experience of caring for family members with dementia and depression. Of course we brought this to the task of writing this blog. We appreciate how difficult it is to diagnose and treat people with this combination of illnesses and we were interested to see the list of interventions recommended by NICE for people with comorbid emotional disorders and dementia.
There certainly does not seem to be convincing evidence that antidepressants are safe and effective for this population when compared with placebo. Polypharmacy must be considered when adding one more medication, especially one that may have limited benefits and potential harms.
There are non-pharmacological approaches that could be considered as an alternative to antidepressants, although the availability of these may be limited for some patients, and of course not all patients will find these psychosocial options acceptable.
We were reminded in our conversation about this paper of the importance of quality of life in people affected by dementia. It’s striking that there is relatively little high quality evidence published about the ways in which we can improve the daily lives of people with dementia. A case of mistaken priorities perhaps?
People interested in exploring these issues further may like to watch the Mental Health Question Time event from November 2016, featuring amongst others Prof Rob Howard from UCL; one of the authors of this systematic review.
Thanks to the Equator Publication School students: Patricia Ayala, Brigid Hickey, Helena VonVille, Jackie McRae, Jiae Choi, Antony Palmer, Mimi Ko, Patricia Logullo, Paula Dhiman, Raja Jayaram, Regina Sit, Wendy Wong, Yu Uneno. Thanks also to Caroline Struthers and Jennifer De Beyer from the Equator Network.
Orgeta V, Tabet N, Nilforooshan R, Howard R. (2017) Efficacy of Antidepressants for Depression in Alzheimer’s Disease: Systematic Review and Meta-Analysis. Journal of Alzheimer’s Disease 58 (2017) 725–733 DOI 10.3233/JAD-161247
Antidepressants for treating depression in dementia. Cochrane Database of Systematic Reviews 2002, Issue 4. Art. No.: CD003944. DOI: 10.1002/14651858.CD003944., , . (2002)