What actually fueled our efforts to shed light on the potential utility of depressive symptoms in differentiating between in older adults with and without cognitive decline, is the fact that depressive symptoms are very, very common in older adults with cognitive decline. For instance, we know from the literature that depressive symptoms are present in more than 40% of individuals who suffer from dementia due to Alzheimer’s disease...
What actually fueled our efforts to shed light on the potential utility of depressive symptoms in differentiating between in older adults with and without cognitive decline, is the fact that depressive symptoms are very, very common in older adults with cognitive decline. For instance, we know from the literature that depressive symptoms are present in more than 40% of individuals who suffer from dementia due to Alzheimer’s disease. And depressive symptoms are also very, very common in individuals with mild cognitive impairment in this predementia stage of Alzheimer’s disease, which is characterized by very mild cognitive deficits which do not really affect activities of daily living. So depression and depressive symptoms are very common in both dementia due to Alzheimer’s disease and mild cognitive impairment. And of course, we do have depressive symptoms also in older adults without cognitive decline, but the prevalence of depressive symptoms in this category of older adults is not so common, is not so high as in individuals with mild cognitive impairment and dementia.
So this difference in the prevalence of depressive symptoms motivated us to try to shed light on the potential utility of depressive symptoms in differentiating between older adults with and without cognitive decline. What we detected through our analysis is that the total score on a commonly used depression scale like the Geriatric Depression Scale, is not useful in differentiating between older adults with and without cognitive decline. But when we analyze the data with machine learning techniques and with the help of colleagues who are computer scientists, then these algorithms, these methods help us differentiate between those with cognitive decline and those without cognitive decline. The accuracy of the method is higher than 80%. And what is even more important is that individuals can complete the questionnaire about different depressive symptoms even in the waiting room of health care facilities. So, it could be another tool for clinicians to differentiate between older adults with and without cognitive decline.
I think that these findings are just preliminary findings, even though they are findings based on two epidemiological cohorts, independent cohorts, one which is recruited in Greece and another one which is recruited in the in the federal state of Saarland in Germany. We have to replicate these findings in independent cohorts with more individuals with dementia and with mild cognitive impairment. And we do have also to replicate our findings in independent cohorts through using not only Geriatric Depression Scale but also other depression scales, depression instruments such as the Hospital Anxiety and Depression Scale or the Beck’s Depression Inventory. So our findings should be treated with caution, of course these are preliminary findings we have to replicate together with colleagues from other institutions, we have to work to replicate them in independent cohorts.
