IPA Bulletin Recent Advances - Volume 23, Number 4

By Dr. Robert Barber, and Professor Robert Baldwin

Depression

Post-Soviet depression

Around 10% of the population of Armenia is 65 or older and their living conditions have deteriorated since the break up of the Soviet Union Bloc. Rates of depression are apparently very high. Srapyan and colleagues (Age and Ageing, 2006; 35: 190–3) measured and compared health-related quality of life (SF-36) and the Geriatric Depression Scale (GDS 30) in older people living in households and in residential homes (n=101 and n=104 respectively) in one district. Those in the residential (retirement homes) had better physical functioning than those in households, but reported more bodily pain and perceived themselves to be in poorer health. However, only 25% of those in rest homes and 18.8% of those in households respectively did not have depressive symptoms. The authors comment that their findings “suggest a possibility that there are serious problems with the well-being of this group of the population.”

Co-morbidity and depression: Parkinson’s, diabetes and screening

Patients with Parkinson’s disease have a high rate of depression although few studies have asked whether affective disorder may be a risk factor for Parkinson’s disease. Brandt-Christensen and colleagues from Denmark (Journal of Neurology, Neurosurgery and Psychiatry, 2006; 77: 781–3) using a large case register recorded the use of antidepressants, Lithium and anti-Parkinsonian drugs. The event of interest was the purchase of anti-Parkinsonian drugs. Compared to an unexposed group, those who had taken antidepressants had a 1.79 times higher risk (CI 1.72–1.86) for the subsequent taking of anti-Parkinsonian drugs. For Lithium the comparable risk ratio was 1.88 (CI 1.6–2.2). A comparator, anti-diabetic medication, showed no association. The risks seemed highest in the first six months of exposure. The cause of this intriguing association is unclear, but the authors speculate that possibly affective symptomatology is a prodrome of Parkinson’s disease or that mood drugs upset a delicate cerebral neurotransmitter balance and trigger Parkinsonism. In fact, diabetes is linked both to depression and dementia. Qiao and colleagues from Boston, USA (Journal of the American Geriatric Society, 2006; 54: 496–501) examined 291 people aged over 60, 40% with diabetes. They assessed cognitive function with an executive test battery and depression using the Center for Epidemiologic Studies — Depression CESD (score 16 or above). After adjusting for confounders, diabetes status was significantly associated with poor performance on visuo-spatial, executive and delayed memory domains of cognitive tests. Significantly more of the diabetic group achieved the cut-score for depression. Although cross-sectional, i.e. causality cannot be assumed, this finding fits with the established view of microvascular disease of the brain.

Given high level of depression among older people with medical co-morbidity, could screening medical inpatients for depression help? Cullum and colleagues from the UK (International Journal of Geriatric Psychiatry, 2006; 21: 469–76) used the 15-item GDS to screen 50% of all consecutive acute medical admissions to one hospital in England in patients aged over 65 and over a 15 month period.  A second stage structured interview generated an International Classification of Diseases ICD 10 diagnosis. Those with severe confusion and language problems were excluded. The overall prevalence was 17.7% of ICD 10 depressive disorders but using the recommended cut off point of greater than 5 on the GDS 15, over 40% would have been classified as depressed. The authors examined various cut scores for the GDS and found seven or more on the GDS-15 was the optimum in this patient group.

Depression and mild cognitive impairment

Major depression is known to be a risk factor for dementia but what about Mild Cognitive Impairment (MCI)? Barnes and colleagues (Archives of General Psychiatry, 2006; 63: 273–9) used data (n=2,220; mean age 74 years) from the cognition area of the Cardiovascular Health Study across four sites in the United States. Depressive symptoms were evaluated with the Center for Epidemiologic Diseases CES-D scale and there was an extensive work up for vascular risks/disease using three groups: (history of vascular events, subclinical vascular disease, and neuroimaging evidence of vascular disease). Thirteen percent developed MCI at up to six years follow-up, a risk that was associated in a dose-response way with a number of depressive symptoms at baseline. For the highest symptom scores, the risk of MCI approximately doubled. This association held even controlling for gender, ethnicity, and the use of antidepressants, and was independent of vascular disease, although this too was associated with MCI. This suggests that in depression, pathways other than vascular disease may impair cognition. As the authors surmise, hippocampal damage via glucocorticoid disturbance is one possibility.

Also, of relevance are the findings of the US National Institute of Mental Health conference “Perspectives on depression, mild cognitive impairment, and cognitive decline” (Steffens et al, Archives of General Psychiatry, 2006; 63:130–8). The need to study mild cognitive impairment and depression together, rather than as separate entities is emphasized. The report summarizes the evidence that depression may be a prodrome of dementia or an independent predictor of it. Also highlighted is the fact that non-amnesic forms of MCI may be more relevant to late-life depression than amnesic MCI but the former is less researched. Instruments should capture a range of both depressive symptoms and behavioural manifestations that may occur in both depression and cognitive impairment such as irritability, apathy, thoughts about death and changes in temperament. Other relevant areas are personality, appetite change, sleep apnoea and psychiatric history. Naturally the report asks for more research but a main message is collaboration which is being practised and not merely preached because the NIMH, the National Institute on Aging and the National Institute of Neurological Disorders and Stroke have begun to share ideas.

In the same edition of Archives there is a report in which the research question posed is whether depression contributes to cognitive decline over time, independent of the occurrence of dementia. The Monongahela Valley Independent Elders Survey (MoVIES) has 12 year follow-up data (Ganguli et al, Archives of General Psychiatry, 2006; 63: 153–60). Besides a cognitive screen, participants completed a modified CES-D. All subjects were aged at least 67 years and were assessed regularly; 1,094 were dementia-free and 171 subjects eventually developed dementia, with a mean follow-up of 7.4 years and 5.9 years respectively. A total of 106 (9.7%) of the dementia free group and 22 (12.9%) of the eventual-dementia group had depression. In the no-dementia group, cognitive decline was minimal. Depressive symptoms were cross-sectionally related to cognitive function but were not associated with subsequent decline on any of the measures. In contrast, those who developed dementia had marked decline on all tests over time but without much effect attributable to baseline depressive symptoms. The authors conclude that substantial cognitive decline does not occur in the absence of incipient dementia and so cannot be attributed to depression. Practically speaking then, it may be that patients with depression who develop cognitive impairment should be closely watched for dementia rather than their symptoms assumed to be secondary to depression.

And not forgetting bipolar disorder

Although there has been much interest in cognitive impairment in depressivedisorder, less attention has been given to Bipolar Disorder in old age (BPD). Young and colleagues (Journal of Affective Disorders, 2006; 92: 125–31) conducted a literature review and report their own findings. Over a 35-year period, seven studies which assessed cognition were identified. These suggested that altered cognition was negatively associated with behavioural function and living skills in the community. The authors further support this with their own data. They compared 37 BPD patients (average age 69) with 37 comparison subjects. The former had lower Mini Mental State Examination (MMSE) and Mattis Dementia Rating Scale (DRS) scores, notably on initiation/perseveration on the DRS. As with depression, there is a concern that some BPD patients develop persistent cognitive deficits. The authors argue for further research into the putative neuro-protective effects of mood stabilisers such as Lithium.

Depression in care homes

Is depression an independent risk factor for nursing home admissions? Harris and Cooper (Journal of the American Geriatric Society, 2006; 54: 593–7) explored this using the US Health Outcome Survey, Nursing Home Minimum Data Set and the Medicare Enrollment Database. Three questions were used to identify depression: feeling sad, depressed or loss of interest in the past year for two weeks or more; feeling depressed or sad much of the time over the past year; having two years or more in life feeling depressed or sad most days. The authors found the question “In the past year have you felt sad or depressed much of the time?” was the best discriminator. It was independently associated with depression even allowing for physical function and age, which were also predictors. Interestingly, the same question was also found to be predictive of high death rate in another study. Clearly the usual problems with the direction of causality apply but the data may be of importance not only to clinicians but to insurance-funded health systems that may wish to employ depression screening questions prior to nursing home admission.

Since depression is so common in older community residents, how should prevention be targeted in a cost-effective way? Smit and colleagues from the Netherlands (Archives of General Psychiatry, 2006; 63: 290–6) used data from the Longitudinal Ageing Study of Amsterdam (LASA) to identify new cases of depression with the CES-D (20 items). From a database of 2,200, cases of depression were ascertained. One in five so identified was a new case. The main risk factors were: being female, low education, having two or more chronic illnesses, high depressive symptoms, more functional limitation and a small social network. By examining only those with both depression and functional limitation, a smaller risk profile emerged: being female, low education and chronic diseases. Together these produced a number-needed-to-treat (NNT) of 4 for the minimum effort to generate the greatest health gain. Or to put it another way, 1,785 new cases of depression might be prevented in every 1 million older adults if the intervention was 30% successful. This suggests a viable preventative strategy could be developed at a population level.

Personality

The LASA data have also been employed to understand whether personality is a predictor for becoming depressed in later life. Steunenberg and colleagues (Journal of Affective Disorders, 2006; 92: 243–51) studied neuroticism, mastery, self-efficacy and self-esteem. The hypotheses was that poorer scores on these measures will strengthen the impact of health-related variables and social situational factors on the onset of depression (n=1,511 nondepressed respondents, 255 of whom developed depressive symptoms). Low mastery and higher neuroticism were strongly related to becoming depressed, more so than physical health and social resources. There was no significant interaction between personality and age.

Antemortem differential diagnosis of DLB from AD using myocardial scintigraphy

Over the past five or so years there has been growing evidence, largely from academic centres in Japan, that measuring cardiac sympathetic activity can help to differentiate neurodegenerative diseases. The technique uses a radioactive compound, abbreviated to (123)I-MIBG, to measure cardiac sympathetic activity. Reduced uptake is thought to reflect a loss and/or dysfunction of the postganglionic cardiac sympathetic nervous system. There has been a flurry of studies reporting on this technique in 2006. As reviewed by Kashihara and Yamamoto (Journal of Neurology 2006;253 suppl 3:iii35–iii40), reduced cardiac MIBG uptake is reported to be more pronounced in patients with Parkinson’s disease compared to those with progressive supranuclear palsy (PSP), corticobasal degeneration (CBD) and multiple system atrophy (MSA). Within subjects with Lewy body disease, uptake appears to be even more compromised in those patients with dementia (DLB) compared to Parkinson’s disease (Suzuki et al, Journal of the Neurological Sciences, 2006 15;240(1–2):15–9). Not surprisingly, interest has also focused on whether the technique can be usefully applied to distinguish subjects with DLB from AD. Hanyu et al (European Journal of Nuclear Medicine and Molecular Imaging 2006 33(3):248–53) found MIBG scintigraphy enabled more accurate discrimination between DLB and AD than was possible with perfusion SPECT: MIBG uptake was significantly lower in the DLB group compared to AD, with marked reductions observed in all patients with DLB (n=19). Yoshita et al (Neurology, 2006 66(12):1850–4) found the heart-to-mediastinum uptake ratio (H/M ratio) of myocardial MIBG uptake was significantly decreased in the DLB group (n=37) vs AD (n=42) and controls (n=10). The delayed H/M ratio had a sensitivity, specificity and positive predictive value of 100%. This discrimination of DLB from AD was independent of the severity of Parkinsonism, with DLB patients without Parkinsonism (n=7) recording similar results to those with parkinsonism (n=30). Abnormal (low uptake) MIBG myocardial scintigraphy is now included as a “supportive feature” in the recently revised diagnostic criteria for DLB (McKeith et al, Neurology, 2005;65(12):1863–72).

How many older people without dementia have Alzheimer disease pathology at autopsy?

Findings from community based clinicopathological studies can have the tendency to challenge the way we see and classify diseases and disorders. Perhaps one such recent study was by Bennett et al (Neurology, 2006 66(12):1801–2). From a cohort of more than 2,000 individuals without dementia under going annual clinical evaluation, they report findings from 134 individuals who underwent post-mortem assessment. Approximately 37% of the sample met NIAReagan criteria for Alzheimer’s disease: 1.5% for a high likelihood and 35.8% for intermediate likelihood. Also, 21.6% had cerebral infarctions and 13.4% Lewy bodies. There was no difference in antemortem Mini-Mental State Exam (MMSE) scores (closest to death) between the sample meeting autopsy criteria for AD compared those not meeting the criteria (mean score 28.2 vs 28.4). A more detailed subanalysis of neuropsychological performance yielded only one difference: individuals with Alzheimer pathology had a slightly lower performance on tests of episodic memory than those without this pathology. The authors conclude that Alzheimer disease pathology can be found in older people without dementia and is related to subtle changes in episodic memory.

Fruit & vegetable juices and Alzheimer’s disease

Readers of the IPA Bulletin who like to follow a “healthy” diet may take further comfort from a recent prospective study examining the relationship between fruit and vegetable juices and incidence of Alzheimer’s disease (Dai et al American Journal of Medicine 2006 119(9):751–9). Just over 1,800 dementia-free Japanese Americans were followed up for nearly 10 years, and after adjusting for potential confounders, individuals who drank juices more than 3 times a week were less likely to develop Alzheimer’s disease than those drinking less. This effect was particularly evident among those with a higher risk of AD, most notably those with an ApoE-4 allele. The consumption of vitamins (E, C or beta-carotene) had no effect. That fruit and vegetable juices could have a positive effect could, at least in theory, stem from their high concentration of polyphenols. Polyphenols are thought to offer more neuroprotection against hydrogen peroxide mediated damage than antioxidant vitamins.

Judging pain in severe dementia

The value of effective “palliative care” in a dvancing dementia is now well recognized. Successful care depends at least in part on being able to detect, and ultimately managed, factors that can have a deleterious effect on patients, such as pain. But can symptoms like pain be reliably detected in patients with severe dementia? To examine this issue, researchers from Switzerland (Pautex et al, Journal of American Geriatric Society, 2006 54(7):1040–5) compared a number of pain self-assessment tools with an observational pain rating scale in patients with advanced dementia (MMSE < 11). Overall, they found patients were capable of reliably reporting their own pain, and indeed the observational rating scale tended to underestimate the severity of pain.

Memantine and Alzheimer’s disease

Results from a multicentre, randomized, double blind placebo controlled study of Memantine in mild to moderate Alzheimer’s disease were published in American Journal of Geriatric Psychiatry (Peskind et al, 2006 14(8):704–15). Just over 400 outpatients with Alzheimer’s disease (MMSE 10-22) were enrolled in this 24-week study, with half randomised to active treatment (Memantine 20mg/day) and half to placebo. Using standard outcome measures, patients on Memantine had better outcomes on measures of cognition, global status and behaviour. Overall the drug was well tolerated. In a separate publication, Smith et al (Alzheimer Disease Association Disorders 2006 20(3):133–7) reviewed all memantine trials completed by August 2005. At this time there were six published phase three trials; two were in mild-moderate vascular dementia, one in mild-moderate AD, two in moderate-severe AD, and one in severe dementia (both AD and VaD). In summary, they concluded that the treatment effect of Memantine in Alzheimer’s disease was small-to-medium across the main outcome domains of cognition, behaviour, functioning and global impact.

Twenty-five percent rate of conversion from MCI to Alzheimer’s disease

It is generally acknowledged that patients with mild cognitive impairment (MCI) are at higher risk of developing Alzheimer’s disease than age-matched controlled patients, but the extent of this risk varies between studies. To explore this issue further, Boyle and colleagues from Chicago, United States (Neurology, 2006 67(3):441–5) followed up 221 individuals with MCI and 565 without cognitive impairment for an average of 2.5 years. Over this period, 25.8% of those with MCI developed AD, a rate 6.7 times higher than those without cognitive impairment. Perhaps not surprisingly, the rate of cognitive decline was also more rapid in the MCI group. Findings from a separate neuropathologic study indicated Alzheimer pathology is presence in most, if not all, cases with amnestic MCI (Storandt et al, Neurology, 67(3):467–73).

Which patients with MCI are most likely to develop Alzheimer’s disease?

Predicting which patients with early cognitive difficulties will progress to dementia was the aim of a recent study by Tabert and colleagues (Archives of General Psychiatry, 2006 63(8):916–24). Over three years of the study, they found patients with mild cognitive impairment (MCI) who had memory plus other cognitive domain deficits (so called “amnestic-plus”) had a much higher conversion rate to AD (32/64 = 50%) compared to those with “pure amnestic” MCI (2/20 = 10%). In terms of knowing which specific neuropsychological measures best predicted conversion to AD, combining two tests, total immediate recall and digit symbol test coding, gave an 86% predictive accuracy of conversion by three years. Therefore, the high risk group were patients with “amnestic-plus” MCI, especially those with deficits in verbal memory and psychomotor speed/executive function. In addition, findings from Storandt et al longitudinal study of MCI (Neurology, 67(3):467–73) suggested that change within individuals provides a better way to determine who will develop AD than comparing results against group norms.

Reprinted from IPA Bulletin, Volume 23, Number 4

Previous Recent Advances