The current findings suggest that the link between depression and cognitive functioning is quite complex. These results demonstrate that specific clusters of depressive symptoms and disease status, as well as gender can differentially impact neuropsychological domains. These findings point to the need to specifically examine disease status and gender when considering the effect of depressive symptoms on neuropsychological functioning as these relationships are obscured in the overall models when simply covarying for these factors.
This is further illustrated by our examination of the relationship between symptom clusters and neuropsychological performance by case status (Alzheimer's disease or normal controls). Among patients diagnosed with AD, the GDS-30 subfactors of dysphoria and apathy were the factors most consistently associated neuropsychological testing. Elevations in dysphoria and apathy were related to poorer scores in attention and executive functioning and verbal fluency and memory, respectively. It is interesting to note that among AD cases scores on the cognitive impairment subscale were significantly associated with better scores on delayed memory (visual and verbal) among the total sample. However, when the sample was split by gender, a different picture was observed as the vast majority of the findings were held only for women, which is consistent with our earlier findings [11
The cognitively normal controls presented a very different pattern of relationships between subfactors and cognition. In this sample, the subfactor of cognitive impairment was, by far, the most robust predictor of neuropsychological test performance. However, as with the AD sample, a different profile was seen when broken down by gender such that the link between depressive subfactors and cognitive status only held for women.
These findings point to the need to directly examine cohorts by demographic factors rather than covarying for them when considering the complex impact depression has on health status. Our findings suggest that the link between depression and neuropsychological functioning is most relevant for women, particularly among those without a neurodegenerative disorder. Rosenberg and colleagues [17
] in a longitudinal study of cognitively healthy older women used total GDS score and found a strong relationship between number of depressive symptoms and decline in a number of cognitive domains. In our study we found that the specific depressive symptom cluster related to concern with cognitive status was significantly related to poorer cognitive performance in cognitively normal older women. These findings directly contradict the old adage “if they present with complaints of cognitive dysfunction they are likely depressed” and suggest that such a notion may potentially prevent women who have real cognitive disturbances from receiving appropriate attention.
Our findings also demonstrate that the meaning of depressive symptoms changes once an individual has been diagnosed with a neurodegenerative dementia syndrome, such as Alzheimer's disease. Among Alzheimer's disease cases, it is feelings of apathy and dysphoria that are most important. These findings are not surprising given the wealth of the literature documenting social withdrawal and isolation as an early sign (or consequence) of AD. However, given that social withdrawal is also a symptom of depression, it is important that elders presenting with these specific symptoms of depression be referred for a comprehensive dementia examination so that treatment can be implemented early in the course of the disease when treatments are most efficacious.
The generalizability of our findings is limited by the nature of the sample. The study is cross-sectional and the TARC cohort at this time is predominately Caucasian, relatively well-educated, and, for the most part, urban. Our sample was limited to mild AD which does not allow us to evaluate the affect of these symptoms as the disease progresses. Additionally individuals with high levels of depression initially or with a diagnosis of major depression are excluded from the cohort. It is therefore difficult to determine the impact of level of depression. As can be seen from , the overall level of depressive symptoms is relatively low for both AD and controls. Even with the low number of depressive symptoms endorsed there is still a differential impact of symptom clusters on neuropsychological domains. The subscales used in this research were originally developed through factor analysis of a sample of individuals with a variety of cognitive disorders. The application of these scales to a cognitively normal sample may be a limitation although previous research on cognitively intact elderly [11
] using the subscales has shown their utility in describing the relationship of depressive symptom clusters to neuropsychological performance. Additionally the role that education may play in moderating the relationship between depressive symptoms and neuropsychological performance was not directly investigated. It is likely that the effect of education was controlled to at least some extent through the use of education-adjusted neuropsychological scales scores and the finding that education did not significantly impact any of the relationships reported.
A potential implication for the current findings is on rate of progression. Given the current findings of a differential link between specific depressive symptoms and cognitive functioning between AD cases and controls, it is possible that the depressive symptoms most important for conveying risk of cognitive decline over time will vary. This is of critical importance as the current investigators are unaware of any prior work examining if specific symptoms (or symptom clusters) of depression are most important in predicting decline over time or if these “depressive risk symptoms” differ by case or controls status. That is, are specific symptoms of depression related to the risk for developing Alzheimer's disease in the future and are a different set of depressive symptoms important for rate of decline once the disease has become clinically manifest?
The current findings may suggest a better understanding of the neurobiological consequences of AD. While there has been a wealth of literature documenting alterations in inflammatory and other neurochemical pathways in both AD and depression, no prior research has examined if the specific symptoms of depression are differentially related to these pathways. If specific symptoms of depression (i.e., apathy and dysphoria) are specifically related to modifiable pathways (e.g., oxidative stress, inflammation), a novel field of targeted therapeutics for AD would be identified based on these findings. Given that the behavioral manifestations are the most significant cause of caregiver burden in AD, this line of inquiry holds great promise.