The aging process is defined by a slow disruption of immunologic homeostasis with a shift toward a proinflammatory state. This neuroinflammation is thought to compromise the brain’s functioning by increasing its vulnerability to neurotoxicity and impairing neurogenerativity. Our results support this theory in that both encoding and recall of memories were negatively associated with serum IL-6 levels in depressed and healthy elderly adults. As IL-6 levels increased, recall ability decreased for information learned from 5 to 20 minutes earlier as did the ability to encode new information. In addition, the encoding association with IL-6 was modulated by sex, with women showing a greater sensitivity to higher levels of IL-6 than men. Contrary to our hypothesis, patients and controls did not differ in the strength of the relationship between IL-6 and encoding or recall performance. Also contrary to our hypothesis, IL-6 levels did not correlate negatively with executive function or attention/processing, and CRP was not associated with any cognitive function assessed. The relationship between IL-6 and both memory functions currently assessed is in addition to significant contributions from age, MMSE scores, literacy level and depression diagnosis.
The effect sizes (standardized betas) from the multiple regression equation for Recall indicate that MMSE and diagnosis are the strongest predictors of recall performance, yet IL-6 makes approximately the same contribution as age or sex. Thus, it is an important and potentially modifiable condition for patients who complain of changes in recall ability. Elderly patients recovering from illness may also experience transient problems with memory with elevated IL-6 levels.
On the other hand, the multiple regression equation for Encoding indicates that IL-6 and the IL-6-by-sex interaction are the strongest predictors of performance and are stronger than age, literacy level, diagnosis or MMSE score. Unfortunately, poor encoding performance can be mistaken for mild cognitive impairment or prodromal dementia because recall does not improve with the use of cues. When being tested for memory function, cues facilitate recall performance if encoding occurred successfully. When encoding does not occur, cues are not helpful and suggest possible damage in the hippocampal and entorhinal complexes and indicate need of further investigation. Thus, immunologic imbalance could lead to a misrepresentation of an elderly person’s, particularly a woman’s, memory problems.
Moderately depressed patients and controls did not differ in their levels of IL-6, yet depressed individuals showed a deficit in memory performance relative to controls. This difference may result from an enhanced sensitivity to the presence of IL-6 among depressed patients if they have experienced earlier episodes or have subtle increases in IL-6 that do not reach statistical significance but which are physiologically important. Trzonkowski and colleagues (2004) reported that older depressed individuals showed elevated levels of TNF-α and IL-6 after receiving an anti-influenza vaccination together with insufficient production of the anti-inflammatory IL-10 and natural killer cell cytotoxicity compared to age and sex-matched controls.(48
) Although the differences disappeared when they corrected for anti-CMV antibodies, Trznokowski and colleagues concluded that the depressed patients demonstrated a proinflammatory profile suspected of resulting from prior exposure to elevated levels of IL-6 and deficiency of suppressive IL-10+ cells.
Our results are consistent with Trznokowski and colleagues’ findings. Although depressed and healthy individuals were comparable in IL-6 levels, the difference in cognitive performance may indicate that moderately depressed patients develop a sensitivity based on prior periods of increases in IL-6 that is independent of absolute IL-6 levels. In our model, depression status could potentially represent this enhanced sensitivity to IL-6 as it was an independent predictor of recall and encoding ability after accounting for IL-6 levels. A recent review by Miller et al (2009) outlines how depressed individuals exhibit features of inflammation including elevations in relevant inflammatory cytokines and their soluble receptors in peripheral blood and cerebrospinal fluid.(49
) Peripheral blood concentrations of acute phase proteins, chemokines, adhesion molecules, and inflammatory mediators such as prostaglandins show concurrent elevations during depressed episodes.(14
) Thus, depression status may be associated with a broad immunologic disruption of which IL-6 is one factor. Subgroups of depressed patients may experience one or multiple changes while healthy elderly have a more limited response.
The link between memory ability and cytokines may occur at the molecular level via cytokines’ role in neurogenesis, memory consolidation and synaptic plasticity. (50
) Transgenic mice that over-express IL-6 have a 63% reduction in neurogenesis in the dentate gyrus of the hippocampus, a medial temporal structure critical for memory consolidation.(51
) Inflammation resulting from cranial radiation therapy among humans causes cognitive loss that appears associated with impaired neurogenesis in the dentate gyrus via stem cell dysfunction.(52
) Additionally, the hippocampus and the hypothalamus have the highest expression of inflammatory cytokine receptors for IL-6 and IL-1β in the brain.(53
) Synaptic plasticity may be modulated when peripheral cytokine IL-1 elevation induces nuclear factor κB (NF-κB) in the brain.(54
) This induction in the brain is thought to contribute to changes in neuronal growth and survival, especially through the further induction of nitric oxide and, ultimately, oxidative stress, which has been shown to alter promoter function for several genes central to synaptic plasticity. (55
Among elderly depressed patients, hippocampal surface contractions have been observed by our research team among elderly depressed patients, particularly among late-onset patients, and the contractions correlate with decreased performance in nonverbal memory (encoding and recall were not tested separately).(57
) A postmortem study of brains from depressed individuals reported condensation of neuronal architecture in the hippocampus, which is consistent with neuronal dearborization rather than apoptosis in hippocampal tissue suggestive of a pathodegenerative process (58
) such as developing Alzheimer’s disease. Furthermore, increased levels of IL-1β disrupt long-term potentiation,(59
) a cellular mechanism believed to be important for recall, one type of memory assessed in this study. In mice, treatment with lipopolysaccharide (LPS), which causes an inflammatory response, produces increased IL=1β, IL-6 and TNF in brain areas including the hippocampus.(61
) Pugh et al (62
) demonstrated that injections of LPS or IL-1β directly into the hippocampus disrupted hippocampal-dependent memory consolidation. In mice, the Morris water maze, a hippocampal-dependent test of spatial learning, has been shown to be sensitive to disruption by infection or peripheral immune activation and subsequent upregulation of proinflammatory cytokines.(63
) Data from the Longitudinal Aging Study Amsterdam indicates that in persons with a high level of IL-6, there is a negative association between higher homocysteine at baseline and immediate recall.(64
) Consequently, people with higher steady state levels of IL-6 appear more vulnerable to current and future cognitive deficits. Future research is needed that explores the links between hippocampal neuronal dearborization, pro-inflammatory response patterns, decreased cognitive performance and the onset and severity of mood disorders.
Future research also needs to examine the relationship between IL-6 and IL-1β in depressed and healthy elderly because the relationship of IL-1β to memory functions is variable. At basal levels, IL-1β appears to support memory mechanisms, but changes in the basal level, whether increase or decrease, appear to disrupt memory consolidation and may result in cognitive impairment.(63
) Baune et al (65
) reported an association between IL-1β, IL-6 and TNF and cognitive performance among healthy elderly humans in the general population, but the TNF was thought to be protective of processing speed.(66
) Finally, longitudinal research is needed to address the question of whether elevated IL-6 predisposes one to developing depressive syndrome, or does the onset of the depressive episode from another, as-yet-unknown cause, precipitate either an increased immune response or the failure of the suppression response that causes a cascade of brain changes that have functional implications.
In conclusion, results suggest that influences on memory performance for elderly individuals and depressed patients are multifactorial, but one factor is the serum level of IL-6. Elevated levels of IL-6 are associated with some of the variation in recall performance for all elderly individuals, but encoding ability may be singularly sensitive to fluctuations in IL-6 levels for older women.