The Course of Depression in Parkinson Disease
A better understanding of depressive symptoms in Parkinson disease, their progression, and risk factors may benefit the management of Parkinson patients.
D
epression and Parkinson disease (PD) both occur frequently in later life1 2. Unsurprisingly therefore, the two conditions exist together in a substantial number of patients, with up to 35% of PD patients suffering with depression. However, evidence has suggested that depression occurs more frequently in PD patients than it does in the general elderly population, or even in those with other disabling diseases such as diabetes or osteoarthritis3.
Of course, whether depression is a risk factor for the development of disease or rather a PD associated condition is important; one study4 looked to see whether PD patients were more likely to have been diagnosed with depression prior to PD, when compared to disease free individuals. Results suggested that depression and anxiety may be early symptoms during the prodromal phase of PD, rather than a risk factor for the disease itself. However, further research into general risk factors not related to PD needs to be done in order to assess whether the two diseases simply share risk factors that predispose to both conditions. This aside, it's clear that the appropriate management of PD patients with depression is important; those with symptomatic depression are more likely to be more disabled and have a lesser quality of life, with adequate treatment improving this5.
With a view to improve management by gaining a better understanding of depressive symptoms in PD, their progression and risk factors, Larson et al.6 have carried out a 7 year longitudinal cohort study. Ideally, the prospective nature of this study reduces the potential sources of bias and confounding that may otherwise be seen in retrospective studies.
187 PD patients and 166 age and sex matched controls were recruited with average follow-up time at 6.0 and 5.9 years respectively. The Montgomery and Aasberg Depression Rating Scale (MADRS) was used at time of diagnosis, inclusion in the study and after 1, 3, 5 and 7 years of follow-up. A threefold higher depressive load in patients was observed when compared to controls, with the mean MADRS scores at 4.2 and 1.3 for patients and controls respectively. Unfortunately, the use of depression rating scales rather than diagnostic interviews appears to be the authors biggest downfall in study design. Of course, as the authors suggest, the MADRS rating scale is valid and reliable however more up-to-date studies9 suggest that other scales, such as HAMD-17, might be better suited.
MADRS scores steadily rose for the duration of the study, with the exception of a fall between the initial visit and year 1. The mean score at year 7 higher was than at year 1; importantly the confidence intervals do not overlap, allowing us to assume a 'true' difference. In contrast, the MADRS scores among controls remained low and stable during follow-up.
An astute design element to this study was the inclusion of the various sub dimensions (vegetative symptoms, dysphoria and retardation) of the MADRS score. This was important to ascertain whether an increase in vegetative symptoms, which overlap with PD symptoms, were not the only variable correlating with an increase in MADRS score. Once this had been confirmed, the depression scores could be analysed with respect to PD-specific and non-specific risk factors.
This broad multifactorial approach to causes of depression is one that has been used in previous successful studies7, and allows the authors to study the true relevance of risk factors specific to PD. Perhaps not surprisingly, the severity of PD correlated with higher depression scores in patients. However, when including PD non-specific risk factors, sex was the only variable not influenced by the severity of disease. It could have been pertinent to include additional truly non-specific risk factors, as used in other studies7 (see figure 1), such as family or personal history of conditions such as depression or anxiety.
Moreover, the observation that 12% of PD patients at the beginning of the study are using antidepressants should have been investigated in greater detail. The potential for antidepressant use as a confounding factor in the progression of disease is likely, where some studies have shown that antidepressant use, specifically selective serotonin re-uptake inhibitors (SSRIs), significantly improved depression among PD patients10. This would help provide further basis to possible management strategies in these patients.
The study authors briefly mention that, as well as having studied the natural progression of depression on a group level, they have also looked at its development in individuals. This data is not presented throughout the paper, which is unfortunate as they make the interesting point that depressive symptoms in some patients regresses; a salient observation when thinking about the management of these patients. This was done in a study by Ravina et al., who concluded that patients with severe symptoms had a reduced likelihood of remission, suggesting that early intervention would be an appropriate approach to management in these patients8. However, as was done in this study, subdividing the cohort was useful to assess the varying regression rates among patients who may have begun the study with different levels of depression.
This paper represents an advance in the method of study of depressive symptoms in PD patients, where most have relied on a retrospective cross sectional design. An evident link between depressive symptoms and PD has been demonstrated, and issues surrounding confounding factors between the measurement of the two conditions have been addressed. Moreover, an understanding of depressive sub dimensions in PD patients may help guide treatment options, however, future research needs to focus more on the general risk factors, not associated with PD that could be causing the levels of depression seen in these patients. For example, risk factors in other conditions that share a similarly high level of depression could be compared with those studied in PD to look for any correlation.