Both depression and dementia are common conditions in old age. Around 7% of those over 65 have dementia (Livingston et. al., 2017) and 17% have depression (Luppa et. al., 2012). The two conditions commonly coexist, with around a third of people with dementia also having significant depressive symptoms (van der Linde et. al., 2016), which are also often persistent. As well as being distressing for the person affected, depression in dementia is linked to worse disease outcomes and higher stress for family members.
However, treating depression in people with dementia is challenging for clinicians, as there are few reliably effective interventions. This is partly because depression in dementia has several different faces. Some of those affected have had depression during their lifetimes which persists once dementia develops. Others are highly distressed by the gradually progressive symptoms that characterise the onset of dementia, and then further dismayed by these symptoms being laid bare by cognitive testing in clinic and receiving the diagnosis. Moreover, in other people the depressive symptoms are likely to result from the direct neuropathological damage caused by Alzheimer’s disease or other dementia pathologies. The challenge for the clinician is to decipher whether the depressive symptoms displayed by individual patients represent treatable disease or intractable dementia symptomatology.
There is therefore little clear consensus for treatment. In clinical practice, use of medication for depression and other psychological symptoms in dementia is common. UK primary care data shows that 37% of people with dementia in 2015 were prescribed an antidepressant, which had increased from 28% in 2005 (Donegan et. al., 2017). Other effective treatments (such as collaborative care) for depression in older people without dementia may be avoided because of the understandable concern that people with dementia may not be able to meaningfully engage. However, concern about potential adverse-effects of antidepressants (Mueller et. al., 2017) and their ineffectiveness – a 2018 Cochrane review (Dudas et. al., 2018) concluded they have little or no effect – has renewed the search for effective treatments.
A recent study by Jennifer Watt and colleagues (Watt et. al., 2021) used a network meta-analysis with the aim to identify the most effective interventions.
The researchers searched five research databases for randomised controlled trials assessing any drug or non-drug intervention compared to another intervention or usual care or placebo. Study participants had to have dementia and either comorbid major depressive disorder or have their depressive symptoms quantified by an appropriate scale.
A network meta-analysis (NMA) was then performed on the relevant studies to synthesise the results together. The NMA allows comparison of the effect of multiple different treatments in a single analysis, even if a direct comparison has never been undertaken. The principle is that if we know the difference in effect on an outcome of treatment A and B, and of treatment B and C, then we can use this information to deduce how treatment A would compare to C.
Watt and colleagues found 256 trials of 61 different treatments, the majority of which were compared to usual care or placebo. They linked these together in a series of analyses aiming to find which treatments had meaningful effect on depressive symptoms, or on the incidence of major depression. Several sensitivity analyses were designed to consider whether different study designs or populations affected the results.
Treatments for depressive symptoms in dementia
From the NMA of 235 studies examining the effect of interventions on individuals with dementia who had symptoms of depression but not major depression, ten interventions were considered effective compared to usual care. These included:
- Cognitive stimulation combined with a cholinesterase inhibitor
- Cognitive stimulation combined with social interaction and exercise
- Massage and touch therapy
- Cognitive stimulation
- Multidisciplinary care
- Occupational therapy
- Reminiscence therapy
- Animal therapy
- Psychotherapy combined with reminiscence therapy and environmental modification
Overall, the interventions found to be most effective were cognitive stimulation combined with exercise and social interaction, cognitive stimulation combined with a cholinesterase inhibitor, and massage and touch therapy. For example, cognitive stimulation therapy with exercise and social interaction had a standardised mean difference compared to usual care of -2.43 (-3.73 to -1.05) for depressive symptoms. These three treatments were also found to be more effective at relieving depressive symptoms than antidepressants.
Treatments for depressive disorder in dementia
Fewer studies (22) examined major depression in people with dementia and this small number precluded an NMA for this patient group, so the authors summarised in a brief narrative synthesis the rather mixed evidence from these studies.
The authors concluded:
Non-drug interventions were as, or more, efficacious than drug interventions for reducing symptoms of depression in people with dementia and without a diagnosis of a major depressive disorder.
Few clear conclusions could be drawn on the effect of interventions on major depressive disorder in dementia.
Strengths and limitations
The results of this review are highly credible due to the rigour of its systematic search, which is likely to have found the most relevant papers within the more than 22,000 studies that were first identified, and the meticulous NMA approach. The authors published in advance the study protocol (Watt et. al., 2017) describing the plan for this study and a related one on treatments for agitation in dementia (Watt et. al. 2019), and the analysis conducted appears to have adhered to the pre-specified plan.
However, meta-analyses are only as good as the studies that can be included within them. Between the 250 included papers, there will be substantial variation in the interventions and participants included. For example, ‘social care’ was treated in this study as a single entity, yet in practice can be hugely variable. Therefore, there is value in considering individual studies in specific populations, rather than simply relying on the headline results.
Study quality also varies and may have affected this review’s findings. The authors considered that study factors may affect results and, as recommended, they examined several of these in sub-analyses with fairly consistent results. But although around one third of studies may have been biased by lack of blinding, the authors were unable due to insufficient power to examine whether this affected results. Study effects are inflated if blinding is insufficient (Savović et. al., 2018) and this effect is likely to be more pronounced for non-drug treatments, which may partly explain the finding that these outperformed medications.
Finally, we cannot be sure about how long these interventions may provide benefit as the follow-up duration is variable. Treatments for other psychological symptoms in dementia are effective immediately, but have no lasting effect once the intervention is withdrawn (Livingston et. al., 2014), so the lasting effect of, for example, massage therapy remains unclear.
Implications for practice
This study gives strong support for several non-drug treatments including multidisciplinary team care, cognitive stimulation therapy, and encouraging social interaction. Access to these treatments, which are already considered to be the cornerstones of good dementia care, should be extended. Faced by a person with dementia who is troubled by depressive symptoms, I will continue to consider their individual likelihood of a good response to medication and discuss this and the risks involved with them and their families, but will seek other treatment options and hope that these are available.
We are some way from an ideal ‘personalised medicine’ approach where we can identify the source of depression for each individual and use a targeted treatment with confidence. The treatments with strongest apparent benefit combine several approaches which may reflect the variation in depression causes and symptoms. Perhaps having a full toolkit available is better when we cannot be sure exactly what it is we are aiming to fix.
Statement of interests
No relevant conflicts of interest.
Watt J, Goodarzi Z, Veroniki A, et al. Comparative efficacy of interventions for reducing symptoms of depression in people with dementia: systematic review and network meta-analysis. BMJ 2021;372:n532
Dudas R, Malouf R, McCleery J and Dening T, 2018. Antidepressants for treating depression in dementia. Cochrane Database of Systematic Reviews, (8).
Fox N, Black N, Livingston G, et al. Trends in diagnosis and treatment for people with dementia in the UK from 2005 to 2015: a longitudinal retrospective cohort study. The Lancet Public Health 2017, 2(3):e149-e156
Livingston G, Kelly L, Lewis-Holmes E, Baio G, Morris S, Patel N, Omar RZ, Katona C, Cooper C. Non-pharmacological interventions for agitation in dementia: systematic review of randomised controlled trials. The British Journal of Psychiatry. 2014 Dec;205(6):436-42.
Livingston G, Sommerlad A, Orgeta V, et al. The Lancet International Commission for Dementia Prevention and Care.The Lancet 2017;390(10113):2673-2734
Luppa M, Sikorski C, Luck T, et al. Age- and gender-specific prevalence of depression in latest-life – Systematic review and meta-analysis. Journal of Affective Disorders 2012;136(3):212-221
Mueller C, Huntley J, Stubbs B, et al. Associations of neuropsychiatric symptoms and antidepressant prescription with survival in Alzheimer’s disease. Journal of the American Medical Directors Association. 2017;18(12):1076-81[Pubmed abstract]
Savović J, Turner RM, Mawdsley D, et al. 2018. Association between risk-of-bias assessments and results of randomized trials in Cochrane reviews: the ROBES meta-epidemiologic study. American Journal of Epidemiology, 187(5), pp.1113-1122.
van der Linde RM, Dening T, Stephan BC, et al. Longitudinal course of behavioural and psychological symptoms of dementia: systematic review. The British Journal of Psychiatry 2016;209:366-377
Watt JA, Goodarzi Z, Veroniki AA, et al. 2019. Comparative efficacy of interventions for aggressive and agitated behaviors in dementia: a systematic review and network meta-analysis. Annals of internal medicine, 171(9), pp.633-642.
Watt JA, Goodarzi Z, Tricco AC, et al. 2017. Comparative safety and efficacy of pharmacological and non-pharmacological interventions for the behavioral and psychological symptoms of dementia: protocol for a systematic review and network meta-analysis. Systematic reviews, 6(1), pp.1-9.