Older age is associated with an increased risk of depression . A systematic review revealed six consistent factors associated with depression in older people. Excepting female gender, physical health and functional impairment were the strongest predictors of late-life depression . Increased rates of depression have been found in older patients with a range of illnesses [3–6]. Indeed, one study demonstrated 83% of those over 75 had more than one health problem, and a psychological and health problem was the second most common co-morbidity . Of even more concern, research confirms that physical illness also increases the risk of suicide amongst depressed older adults .
Despite a large number of trials of psychological therapies in young depressed patients , there are far fewer studies in older people. The number of trials in the literature has been increasing, which has given rise to numerous reviews of the efficacy of psychotherapeutic interventions for late-life depression [10–14]. Although those reviews varied in which interventions were found to be definitely efficacious, there is some general consensus.
Cognitive Behavioural Therapy (CBT), a therapy based on changing unhelpful beliefs and increasing activity as a treatment for depression, is the most well researched therapy in the treatment of late-life depression . In their Cochrane Review, Wilson and colleagues  concluded that CBT was the only definitely efficacious therapy for late-life depression. They used very stringent criteria for trial inclusion, but nonetheless, on this point there is considerable consistency between reviews. That is, CBT is a proven effective treatment for late-life depression [10–14].
In the Cochrane review, Problem Solving Therapy (PST) was considered under the umbrella of CBT [10, 12]. However, there are important differences. CBT focuses on restructuring and challenging beliefs that are associated with low mood and behavioural activation, whereas PST focuses on dealing with the problems associated with daily life and shifting negative cognitions into goal directed activity . One review of psychological treatments in older adults  that distinguished between PST and CBT, concluded that PST was probably efficacious. However, since that review, there have been a number of additional, well controlled trials that have evaluated PST as a treatment for late-life depression, each of which has supported the efficacy of PST [16–21]. Hence, PST can be regarded as meeting the criteria for a definitely efficacious treatment . Furthermore, in a meta-analysis comparing psychotherapies for adult depression, PST was found to have the lowest drop out rates of any of the therapies, making it a highly acceptable intervention among people suffering from depression .
Although there is good evidence to support the efficacy of CBT and PST for the treatment of later life depression, there is a need for further research . Which of these treatments is most efficacious remains unclear. To date, no trials have compared CBT with PST . Research comparing those treatments shown to be most efficacious for the treatment of late life depression is needed in order to ensure that patients are offered the most efficacious intervention. Large trials comparing these treatments can also investigate whether different patients benefit from different treatments. If the different treatments are suited to patients with different characteristics, there may be ways in which interventions can be matched to those patients most likely to benefit from them.
Similarly, there is a pressing need for outcome studies for depressed patients with complex needs. In their meta-analysis, Pinquart and colleagues  found that effect sizes were smaller in studies that included people with a co-morbid medical illness. However, there are few trials of the treatment of late-life depression for people with a range of co-morbid physical health problems and patients with multi-morbidity (i.e. two or more concurrent physical and/or psychological disorders are often excluded from the published trials). Only the PROSPECT  and IMPACT  trials have studied treatment of depression in older medical patients, however, the interventions studied were not exclusively psychological. In these studies, patients were offered a choice of either medication or behavioral treatment. There are, however, some studies in patients with specific illnesses that show that CBT is effective in reducing depressive symptomatology in patients with chronic illnesses often seen in older adults. For example, CBT is successful in reducing depressive symptoms in arthritis and preventing the development of new “cases” of depression over 18 months [25, 26]. Similarly, recent research shows that CBT for Chronic Obstructive Pulmonary Disease (COPD) is successful in improving depressive symptoms .
Despite these encouraging results, neither study specifically included patients who met criteria for major or minor depression. Indeed, there appear to be no studies of CBT in late life depression where patients were included on the basis of a diagnosis of clinical depression and a comorbid health problem. Two studies investigated the efficacy of CBT for patients scoring above a cut-off score on questionnaires. Both studies (a small trial in stroke survivors  and a larger trial in Chronic Obstructive Pulmonary Disease patients ) failed to find a benefit of CBT. However, in the stroke study, patients were recruited on the basis of their scores on a screening test (the Beck Depression Inventory), known to be inflated in physically ill populations. Of those allocated to CBT, 69% met criteria for a depressive disorder, in contrast to only 43% of the control group. These initial differences could have masked any treatment effects. There are also problems with the Chronic Obstructive Pulmonary Disease study. Kunik et al.  included patients with heightened levels of depression OR anxiety and attempted in an 8 x 1 h program to address both symptoms in groups of 10 patients. The cognitive component (i.e. cognitive restructuring) involved only two sessions, and this may not be sufficient, particularly in older adults, to have conferred benefit . While these studies have some problems, the null findings highlight the importance of trials to confirm that the efficacy of CBT extends to those with comorbid medical problems and late-life depression.
There is more research about the efficacy of PST in the treatment of late life depression in physically ill patients. One early trial in older adults with medical illness found positive results , but was uncontrolled. There have since been a number of large trials in the context of particular health conditions. Areán and her colleagues have shown that PST can improve depression in patients with arthritis  and in older disabled people with cognitive dysfunction [16, 32]. This provides good evidence that at least in some disorders, PST can ameliorate depressive illness even where complicated by the presence of one serious, chronic health problem. Additionally, PST has been found to be effective in disabled older adults (many of whom have chronic illnesses) [33, 34] and reduces the risk of post stroke depression , as well as reducing depression symptoms in cancer patients . However, the question of whether PST can be applied across the range of health problems common in later life, particularly in those with multi-morbidity, and whether it is as effective as CBT, remains unclear.
Aims and hypotheses
Although there have been numerous trials of psychological interventions for patients with late-life depression in the past decade, there continue to be major gaps in the literature. Firstly, few studies aim to treat depression in the context of ill health, despite the fact that many older adults have both depression and one or more chronic physical illnesses. There are also few studies that compare different psychotherapies for late-life depression. The PROMISE-D trial aims to bridge these gaps. Specifically, the aim of the study is to evaluate two psychological therapies (PST and CBT), both of which are definitely efficacious for the treatment of late-life depression, for patients with one or more co-morbid physical illness.
We hypothesise that PST and CBT will be superior to a wait-list control group in the treatment of depression for patients with co-morbid health problems. We also aim to determine what baseline characteristics might best predict outcome in each treatment condition. Based on the hypothesised mechanisms of PST, we expect that patients with executive dysfunction (poor score on the Stroop Interference Task ) will respond better to PST than CBT . In contrast, we expect patients with excessive rumination (measured by the Ruminative Response Scale; RRS ) and patients with affective dysregulation (measured by the Emotion Regulation Questionnaire; ERQ ) to respond better to CBT than PST [39, 40].