| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Behavioral Immunology Laboratory (A.L.M., R.S.) and the Behavioral Physiology Laboratory (K.L.P., S.A.N., J.D.F., S.B.M.), Department of Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania; the Center for Clinical Pharmacology (M.F.M.) and the Department of Psychiatry (C.R.), University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania; and the Department of Psychiatry (J.D.F.), Mount Sinai School of Medicine, New York, New York.
Address correspondence and reprint requests to Anna L. Marsland, PhD, RN, Behavioral Immunology Laboratory, Department of Psychology, 3943 O'Hara Street, Pittsburgh, PA 15260. E-mail: marsland{at}pitt.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Five hundred relatively healthy community volunteers aged 30 to 54 had blood drawn for the determination of plasma IL-6 levels and completed a battery of neuropsychological tests evaluating memory and executive function.
Results: After controlling for age, gender, race, and education, hierarchical regression analyses revealed an inverse relationship between circulating levels of IL-6 and performance on clusters of tests assessing auditory recognition memory, attention/working memory, and executive function. In contrast, there was no association between IL-6 and performance on tests of general memory. Secondary analyses demonstrated that relationships between IL-6 and auditory recognition and working memory and executive function were independent of a number of health factors, including body mass index, smoking, and hypertension.
Conclusions: These findings contribute to a growing body of evidence linking chronic inflammation to poorer cognitive functioning and extend these findings to a midlife community sample, raising the possibility that IL-6 may represent a biomarker for risk of future cognitive decline.
Key Words: interleukin-6 • cognitive performance • memory • executive function • middle age
Abbreviations: IL = interleukin; CNS = central nervous system; BMI = body mass index; WMS = Wechsler Memory Scale; BP = blood pressure.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
To date, much of the evidence that proinflammatory cytokines in the periphery are related to neurocognitive function derived from animal models. This literature shows peripheral IL-6 and IL-1 cytokine levels (whether the result of exogenous administration or in vivo immune challenge) to be associated with increased levels of cytokines in regions of the central nervous system (CNS) where IL-1 and IL-6 receptors are localized, specifically the hippocampus and prefrontal cortex (7,8). Increased levels of IL-1 have been found to interfere with long-term potentiation in the hippocampus (9) and have been associated with the development of cognitive deficits such as impairment of spatial learning and memory (10). Peripheral or central administration of specific IL-1 receptor antagonists prevents cognitive sequelae of central proinflammatory mediators (1,11), suggesting a primary role of inflammatory mediators in cognitive decline. Further support for a neurodegenerative role of brain proinflammatory cytokines comes from the study of transgenic mice that express high brain IL-6 levels and show deficits in synaptic plasticity manifested by defects in avoidance learning (12). Together these findings suggest that peripheral levels of proinflammatory cytokines are associated with activation of central inflammatory mechanisms that negatively affect cognitive processes.
It has been hypothesized that inflammatory mechanisms play a key pathogenic role in several age-associated diseases that involve impairments of memory, including Alzheimer disease (AD), vascular dementia, and age-related cognitive decline. Indeed, numerous reports document an association between AD and high levels of central and peripheral IL-1 and IL-6 (13–17), with IL-6 mRNA levels predicting the clinical progression of the disease (18). Furthermore, polymorphic variation in the IL-6 gene resulting in lower plasma IL-6 levels is also associated with a lower risk of developing AD (19). In summary, growing evidence supports a role of chronic inflammation in pathologic cognitive decline.
Peripheral markers of chronic inflammation have also been associated with mild cognitive decline in well-functioning elders. The majority of findings show an increase in plasma/serum levels of IL-6 with advancing age (reviewed by (20)), and in recent epidemiologic studies, peripheral IL-6 levels covary inversely with cognitive function in aged subjects (5,6). Moreover, high levels of IL-6 appear to predict subsequent cognitive decline among this population (5,6 but not (21)). Notably, these findings are independent of many other factors known to be associated with cognitive deterioration, including age, sex, and level of education (22,23). However, it remains unclear whether elevated peripheral cytokine levels cause neurodegeneration and cognitive impairment in the elderly or are merely a peripheral reflection of brain inflammation (24).
Accumulating evidence suggests that a subset of asymptomatic midlife adults show evidence of chronic low-grade inflammation as measured by stable elevations in IL-6 and that these individuals are at increased risk of developing a range of age-associated diseases, including cardiovascular disease, type 2 diabetes, joint inflammation, frailty, and functional decline (reviewed by (25)). This raises the possibility that any decline in learning and memory performance associated with systemic inflammation and raised levels of IL-6 may begin well before the typical cognitive decline associated with old age. Accordingly, the present study examines the potential association between peripheral IL-6 levels and poorer learning, memory, and executive function in a community sample of midlife adults. Given evidence that increased peripheral IL-6 levels predict cognitive decline in the elderly (5,6) and an animal literature suggesting that raised levels of peripheral proinflammatory cytokines interfere with memory formation in the hippocampus and possibly activity in the prefrontal cortex (reviewed by (1)), our primary hypothesis is that elevated IL-6 levels in healthy, midlife adults will be associated with poorer performance on memory and executive function tasks.
The precise source of IL-6 in peripheral circulation remains unclear. In addition to release by activated immune cells as part of the inflammatory process, IL-6 is produced by adipocytes. This raises the possibility that individuals with higher levels of body fat produce more IL-6 and may be at risk of greater associated cognitive decline. Available evidence shows a positive relationship between body mass index (BMI) and circulating IL-6 levels (26,27). Furthermore, recent findings suggest that higher BMI is associated with poorer performance on a range of cognitive function tasks (28). For this reason, secondary analyses were performed to examine whether associations between IL-6 and cognitive function exist independently of BMI and of other health factors that are known to be associated with immune parameters and cognitive function, including smoking and exercise.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Procedure
Neuropsychological assessment and the blood draw for the determination of circulating levels of IL-6 were both conducted on the same day. Participants were asked to fast overnight for 8 hours and avoid exercise for 12 hours and alcohol for 24 hours before coming into the laboratory. All sessions were scheduled in the morning. On arrival, participants were seen by the project nurse who completed a medical history and medication/dietary supplement use interview, obtained measurements of height and weight for the determination of BMI (in kilograms per meters squared), took three manual blood pressure measurements, and drew a 40-mL blood sample. The blood pressure readings and blood draw were conducted after the participant had been sitting for at least 30 minutes. After this health assessment, subjects were offered a snack and drink (regular or decaffeinated coffee and/or juice) before beginning the neuropsychological battery. Neuropsychological testing took approximately 1.5 hours and was conducted by study staff who were supervised by a licensed clinical psychologist.
Neuropsychological Tests
A number of neuropsychological tests were selected on the basis of their ability to assess domains of cognitive functioning theoretically associated with peripheral inflammation, including verbal and nonverbal learning and memory, attention, and executive function. To test the association of circulating IL-6 with attention, learning, and memory, we administered the six primary and three supplemental subtests of the Wechsler Memory Scale–third edition (WMS-III (29)). This battery of memory tests examines attention, working memory, and auditory and visual immediate and delayed memory.
Auditory Memory Subtests of the Wechsler Memory Scale–Third Edition
Verbal memory (immediate and 30-minute delayed recall) was assessed by recall of a short story using the Logical Memory I and II subtest and by a verbal associative learning test (Verbal Paired Associates I and II), in which subjects are presented with 12 unrelated word pairs across four trials and asked to recall them 30 minutes later. In addition, a delayed auditory recognition score was calculated as the sum of the delayed recognition of the Logical Memory subtest and the delayed recognition of the verbal paired associates.
Visual Memory Subtests of the Wechsler Memory Scale–Third Edition
Immediate and 30-minute delayed memory for faces was evaluated using the Faces subtest, which requires subjects to identify photographs of 24 faces that they have been shown previously from a series of 48 pictures. Visual memory was also assessed with the Family Pictures I and II subtests, in which subjects study a family photograph and a series of scenes and are asked to recall (immediate and 30-minute delayed) the family members presented in each scene, where they were, and what they were doing. Nonverbal memory was also assessed by immediate and 30-minute delayed recall of five geometric figures (Visual Reproductions I and II subscales).
Working Memory Subtests of the Wechsler Memory Scale–Third Edition
Attention and working memory were examined with the Letter–Number Sequencing subtest, in which subjects are presented with a string of alternating letters and numbers and are required to repeat the string back placing the numbers together in ascending order and the letters together in alphabetical order. The Digit Span subscale of the WMS III was also administered. In this test, random number sequences of increasing length are presented orally to subjects who are asked to repeat the number sequence either in the same or reverse order. Working memory was also evaluated with the Mental Control subtest of the WMS III. Here, subjects are asked to alternate between overlearned tasks such as saying the alphabet and less familiar tasks such as counting by 6's (e.g., 6, 12 . . .). Finally, subjects completed the Spatial Span subtest, which requires subjects to repeat spatial patterns demonstrated on a three-dimensional board both in the order demonstrated and in reverse order.
Executive Function Tests (e.g., mental flexibility, response inhibition)
In addition, participants were administered two tests of executive functioning: the Trail Making Test (30) and the Stroop Color-Word Test (31). The Trail Making Test involves two parts. Part A requires subjects to draw a line connecting randomly arrayed, consecutively numbered circles as quickly as possible and provides a test of visuomotor speed. Part B of the test requires subjects to draw a line connecting numbered and lettered circles as quickly as possible, alternating numbers and letters (e.g., 1-A-2-B-3-C . . .). To derive a measure of cognitive function that is independent of psychomotor speed, a difference score was also calculated (i.e., Part B – Part A) with higher difference scores reflecting poorer performance.
The Stroop Color-Word Test requires subjects to read aloud as quickly as possible from three pages of color word lists: page 1 requires reading a list of color names (i.e., red, green, blue . . .), page 2 requires naming the colors of the inks, and page 3 requires naming the color of the ink from a list of color names printed in incongruent colors (e.g., the word blue printed in yellow ink). Here, scores are the number of correct responses within a 45-second period with higher scores indicating better performance. In addition, an interference score was calculated indicating the participant's susceptibility to interference (i.e., difficulty inhibiting a primary verbal response). This score is derived by first calculating: (no. of items/45 seconds on page 2 x no. items/45 seconds on page 1)/(no. of items/45 seconds on page 2 + no. of items/45 seconds on page 1). This provides a predicted score for page 3, which is then subtracted from the actual score for page 3 (no. of items/45 seconds). This difference score reflects the degree of interference.
Interleukin-6 Measures
Plasma samples for the determination of IL-6 were collected in citrated plasma tubes and frozen at –80°C until analysis in batches. IL-6 levels were determined using a high-sensitivity quantitative sandwich enzyme immunoassay kit (R & D Systems). Briefly, standards, controls, and samples were added to a 96-well microplate precoated with monoclonal anti-IL-6 antibodies. Unbound substances were removed by washing and an enzyme linked polyclonal anti-IL-6 antibody was then added. This was followed by washing, incubation with a substrate solution and then with the amplifier solution. The intensity of the color that resulted was measured at 490 nm and is representative of the amount of IL-6 present in the sample. The assay standard range is 0.156 to10 pg/mL. IL-6 levels were extrapolated from a standard curve with linear regression from a log-linear curve. All samples were run in duplicate and the average coefficient of variation between samples was 5%.
Control Variables
A number of control variables were assessed that might provide alternative explanations for associations between IL-6 and neurocognitive function. These included age, sex, BMI, race, systolic and diastolic blood pressure (SBP, DBP), antihypertensive treatment (endorsed taking at least one medication to control BP versus not), years of education, smoking status (current smoker versus ex/nonsmoker), sleep volume (hours of sleep during last 7 nights = (average hours/week night x 5) + (average hours/weekend night x 2)), exercise, as measured using the Paffenbarger Physical Activity Questionnaire (32), alcohol use (average number of alcoholic drinks/week), and reported history of past head injury, including loss of consciousness for more than 10 minutes.
Data Reduction
First, all measures were examined for normality and five measures, Trails A, B and B-A, BMI, and IL-6 levels, were subjected to log normal transformation before statistical analysis. Next, to identify domains of memory function, the 20 WMS-III subscale scores were grouped into the eight Primary Indices of the WMS-III following direction in the test manual (29). The Primary Indices are listed in Table 1. To derive a single measure of executive function, a principal components analysis with varimax rotation of the two executive functioning tasks was conducted. The difference score of the Trail Making Test and the Stroop Color-Word T score and interference score loaded on a single factor (factor loadings = 0.52, 0.94, and 0.86, respectively). Thus, we created a factor scale by averaging the appropriate standardized scale scores and equally weighting each of the subscales in calculating the factor score.
|
Statistical Analyses
All analyses were performed using SPSS for Windows (version 11.5). To test the primary hypothesis that elevated circulating levels of IL-6 are associated with poorer performance on tests of memory and executive function, Pearson product moment correlations were conducted. Secondary analyses were then conducted to examine whether any associations between IL-6 and cognitive function were independent of demographic, lifestyle, and health factors that are known to be associated with immune parameters and/or cognitive function. For this purpose, a series of Pearson product moment, point-biserial, or Spearman correlations were performed to examine relationships between circulating levels of IL-6 and demographic characteristics, lifestyle, and health factors and cognitive performance. Next, hierarchical linear regression analyses were conducted examining whether IL-6 levels predicted the eight Primary Indices of the WMS-III and executive function after adjustment for demographic characteristics. For these initial analyses, age, sex, race, and years of education were entered in the first step of the equation followed by IL-6 in the second step. Similar secondary analyses were then conducted to examine whether IL-6 was associated with cognitive function independently of health practices. In the first set of these analyses, demographic characteristics were entered in step 1, smoking in step 2, and IL-6 level in step 3 of the regression equation. This hierarchical regression procedure permitted us to systematically examine the independent contribution of variables entered in each step after taking into account the effects of variables already in the model. Alcohol use, sleep, exercise, and loss of consciousness for more than 10 minutes were not associated significantly with IL-6 levels and/or cognitive function; thus, they were not included in these analyses. Expected relationships between IL-6 and BMI and between BMI and cognitive function led us to perform a further series of exploratory regression analyses with demographic factors in step 1, BMI in step 2, and IL-6 level in step 3 to examine whether any association between IL-6 and cognitive function was independent of BMI. Finally, observed relationships between IL-6 and BP and between BP and cognitive function led us to conduct a similar set of exploratory analyses entering BP and antihypertensive treatment in step 2. To provide an estimate of the effect size of each association, for the regression results, we present the corresponding partial correlation with each F statistic. The partial correlation represents the association between IL-6 and a specific outcome after adjusting for covariates.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
|
|
|
In regard to medical history, there were no associations between IL-6 levels and self-reported history of cancer (n = 13), asthma (n = 39), or thyroid disease (n = 19). There was an association between history of arthritis (n = 66) and higher levels of IL-6 (t (1,452) = –2.48, p < .01); however, arthritis was not associated with cognitive performance. No participants reported a history of heart attack, Parkinson disease, or multiple sclerosis. In regard to medications, there was no association between IL-6 and current use of hormone replacement treatment (n = 18), oral contraceptives (n = 23), or antilipemics (n = 8). The use of antihypertensives (n = 17) was associated with higher mean levels of IL-6 (t (1,458) = –3.21, p < .001). Resting BP was also positively associated with IL-6 (SBP: r = 0.16, p < .001); DBP: r = 0.14, p < .004). As a consequence, exploratory analyses were performed to examine whether IL-6 was related to cognitive function independently of BP. No participants endorsed taking nitrates, antiarrhythmics, proteases, or anti-HIV medications. Participants reported taking a range of over-the-counter medications and dietary supplements; however, there was no association between IL-6 and current use of antihistamines (n = 34), antiulcer medications (n = 12), antacids (n = 6), laxatives (n = 1), aspirin (n = 37), nonsteroidal antiinflammatory drugs (n = 36), acetaminophen (n = 37), antioxidants (n = 70), other vitamins (n = 185), dietary supplements (gingko, ginseng, 5-HTP, SAMe, melatonin, acetyl-L-carnitine, kava kava, inositol, phenylalanine, DHEA; n = 115), St. John's wart (n = 1), niacin (n = 2), fish oil (cod liver oil, salmon oil, omega 3 oil; n = 9), and short-chain omega-3 oils (flaxseed oil, essential fatty acids, gamma linolenic acid, evening primrose oil, borage oil; n = 8).
Correlations Between Cognitive Performance and Demographic and Health Practices
Next, we assessed whether the same series of control variables were associated with cognitive performance (see Table 3). Higher scores on all of the memory and executive function scales were associated with more years of education, not smoking, and being white. Higher BMI and SBP were associated with poorer cognitive performance on all measures except the visual immediate and delayed memory indices. Different patterns of cognitive performance were observed among males and females with women showing better performance on all the memory indices except working memory than men. There were no sex differences in executive function. Age was not related to performance on any of the cognitive dimensions, which likely reflects the small age range of the sample. There were no significant relationships between hours of sleep/week, exercise/week, or loss of consciousness and performance on any of the cognitive tasks. As a consequence of these analyses, it was decided to include sex, race, and years of education as standard covariates in the first step of regression analyses exploring whether the associations between IL-6 and memory and executive function were independent of demographic characteristics, smoking, BMI, and BP. In light of existing literature demonstrating variation in cognitive performance by age, age was also included as a covariate in all analyses.
Correlations Between Interleukin-6 and Cognitive Function After Controlling for Demographic Characteristics
After controlling for age, gender, race, and education, regression analyses revealed a relationship between IL-6 and poorer performance on the auditory recognition index (b = –0.12, F (5,459) = 16.34, r = –0.13, p < .007) and the working memory index (b = –0.13, F (5,459) = 23.62, r = –0.14, p < .002). Similarly adjusted regression analyses examining performance on the individual subtests revealed that individuals with higher levels of IL-6 were less able to recall (either immediately or after a 30-minute delay) the themes of short stories (all p < .02) or recognize auditory information after a 30-minute delay (p < .01) than individuals with lower IL-6. On attention/working memory tasks, IL-6 was negatively associated with ability to recall spatial sequences both forward and backward (all p < .01) and to repeat alternating letter–number sequences (p < .04). There were no associations between IL-6 and individual subtest scores on the word list, verbal paired associates, mental control, faces, family pictures, or digit span tests.
Secondary analyses of executive function tasks revealed a negative association between IL-6 and the factor score with the standard controls in the model (b = –0.13, F (5,453) = 11.35, r = –0.13, p < .004). Analysis of the individual subtests demonstrated that higher IL-6 levels were associated with greater difficulty naming the color of the ink from a list of incongruous color-words (p < .003) and with longer times to complete the Trail Making Part B test (p < .01) and a trend for a greater difference between time to complete Trails Part B and Part A (p < .09). There was also a trend for higher IL-6 to be associated with greater susceptibility to interference on the Stroop task (p < .06). There were no significant interactions between demographic or health-related risk factors and IL-6 or between age and sex in predicting cognitive function.
The Role of Smoking
Our initial zero-order correlations demonstrated an association between higher IL-6 and both current smoking and less exercise (see Table 2). Given that smoking, but not exercise, was also associated with poorer performance on most of the cognitive tasks, we next explored whether IL-6 predicted cognitive function after adjustment for smoking. Entering smoking into the second step of the regression equation with the standard demographic controls already in the model did not reduce the magnitude of the association between IL-6 and any of the cognitive factors.
The Role of Body Mass Index
In light of evidence that adipose tissue is one source of circulating IL-6 (4), we next explored whether IL-6 predicts auditory recognition, attention/working memory, and executive function independently of BMI. As expected, a positive correlation was observed between IL-6 and BMI (r = 0.32, p < .001). Furthermore, a series of regression analyses revealed that, after controlling for age, gender, race, education, and smoking status, BMI was negatively associated with the following memory indices: auditory immediate (r = –0.11, p <.02), auditory delayed (r = –0.09, p < .05), working memory (r = –0.13, p < .005), and executive function (r = –0.14, p < .004). Thus, for cognitive measures also associated with levels of IL-6 (i.e., working memory and executive function), we next explored the effect of adding BMI into the regression model after the demographic controls and smoking. Adding BMI as a covariate reduced the relationship between IL-6 and working memory from r = –0.14 to r = –0.11 and the relationship with executive functioning from r = –0.13 to r = –0.09; however, both IL6 and BMI continued to predict attention/working memory (IL-6: b = –0.10, p < .02; BMI: b = –0.12, p < .005) and executive function (IL-6: b = –0.09, p < .06; BMI: b = –0.13, p < .004) in the full model.
The Role of Hypertension
A large body of evidence links elevated BP to decrements in cognitive functioning (for review, see (33)). Initial univariate analyses revealed a positive relationship between IL-6 and SBP and DBP and inverse relationships between BP and performance in a number of the cognitive domains (see Table 3). Entering resting SBP, DBP, and use of antihypertensive medications into the second step of the regression equation with the standard demographic controls and smoking already in the model revealed no independent effect of SBP, DBP, or antihypertensive use on cognitive performance, except for a positive relationship between resting SBP and the visual immediate memory index (b = 0.14, p < .04). In contrast, IL-6 remained a predictor of scores on the auditory recognition and working memory indices and the executive function scale (b = –0.12, p < .009; b = –0.12, p < .007; b = –0.12, p < .01, respectively).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Our findings that IL-6 was inversely associated with executive function and attention/working memory are generally consistent with an extensive animal literature supporting an association between peripheral inflammation and increased levels of proinflammatory cytokines in the prefrontal cortex and hippocampus, where central IL-6 receptors are concentrated (8,34). Increased brain levels of proinflammatory cytokines have been associated with a decline in cognitive function, including the impairment of learning and memory (10). This memory impairment can be prevented by the administration of antagonists that block proinflammatory cytokine receptors (35). Further support for a relationship between IL-6 and cognitive function comes from transgenic mice that overexpress brain IL-6 and show deficits in memory and learning (36).
Studies in humans with dementia further support a relationship between inflammation and cognitive decline. When compared with age-matched control subjects, patients with AD show higher levels of peripheral proinflammatory cytokines (IL-1 and IL-6), which predict more rapid progression of the disease (18). A role for inflammation in the pathogenesis of AD is also supported by evidence that polymorphisms in the IL-6 gene that result in lower levels of plasma IL-6 are associated with decreased risk of developing the disease (19). Although these findings support a primary role of inflammation in the neurocognitive decline of dementia, the majority of studies in this literature are cross-sectional, and there is much debate about whether the raised peripheral cytokine levels are the cause or consequence of brain inflammation. Our findings support a relationship between inflammatory mediators and cognitive functioning in cognitively normal middle-aged adults, raising the possibility that lower cognitive function begins well before the appearance of clinically significant deficits and providing further support for the hypothesis that inflammation precedes, and may contribute to, neurocognitive decline.
It remains to be determined whether chronic inflammation is a marker of poorer function of specific brain regions or of a more global neurodegenerative process. The current cross-sectional findings provide some support for both possibilities. In this regard, we demonstrate an inverse relationship between IL-6 and performance on both a) executive function tasks (e.g., cognitive flexibility and set shifting), which involve the prefrontal cortex; and b) attentional capacity tasks, which involve more global neurocognitive function. It is also possible that chronic inflammation is associated with progressive, global brain changes and that poorer executive function and attention/working memory are early consequences of this process. Longitudinal studies tracking individuals from midlife are necessary to shed further light on these relationships and to explore whether chronic inflammation among midlife individuals predicts increased risk of future cognitive dysfunction.
It is widely assumed that IL-6 is a marker of chronic inflammation, although the actual source of the peripheral IL-6 measured in this study remains unknown. IL-6 is secreted by a number of different cell types, including activated macrophages and dendritic cells. Macrophages and dendritic cells are widely dispersed throughout the body and serve a central role in most inflammatory processes. Inflammatory reactions in atherosclerotic plaques are proposed as one source of circulating IL-6 (37) and peripheral IL-6 predicts coronary heart disease risk (38,39) and mortality in the elderly (40). Thus, it is possible that the association between IL-6 and cognitive function is mediated through subclinical cerebrovascular disease. Our findings are also consistent with the model proposed by Maier and Watkins (1). This model posits that peripheral proinflammatory cytokines resulting from chronic inflammation act as direct signals to the brain, crossing the blood–brain barrier (2) or activating afferent neural pathways to modulate central inflammatory mechanisms promoting neurodegeneration and cognitive impairment (1).
Another source of peripheral IL-6 is adipose tissue (4), raising the possibility that adipose tissue mass could explain higher levels of IL-6 and be associated with cognitive function. Consistent with the findings of others (26,27), we found an association between BMI and IL-6 (r = 0.32, p < .001). Our findings also support earlier work (28) in showing that individuals with higher BMI performed less well on the majority of the cognitive measures (see Table 3). However, the relationship between IL-6 and auditory recognition, attention/working memory, and executive functioning held after controlling for BMI, suggesting that the effects of IL-6 and BMI on cognitive functioning are partially independent and making it unlikely that adipose tissue is the sole source of the variability in IL-6 levels associated with cognitive function.
In addition to BMI, current smoking was associated with higher IL-6 levels and poorer performance on a number of the cognitive measures. Once again, these effects were independent of the association between IL-6 and cognitive function. However, it remains possible that other health practices not measured here may account for associations between IL-6 and cognitive function. Another potential covariate that may be related to IL-6 levels and cognitive function is depression. The current sample used rigorous exclusions to recruit a relatively healthy sample. As a consequence, there were very few individuals (eight of 500) who met criteria for depressive disorder. We reran the analyses excluding these eight individuals and there were no changes in the observed associations.
There are a number of limitations of the current study. First, its cross-sectional design prevents us from making causal inferences or evaluating the hypothesis that peripheral inflammation is associated with a decline in cognitive performance. Alternative explanations for our results include peripheral levels of IL-6 as a marker of individual differences in cognitive ability across the lifespan or the possibility that subclinical neuroinflammatory conditions exist and account for both higher levels of peripheral inflammatory cytokines and poorer cognitive function. Another limitation is the single assessment of IL-6. Although evidence suggests that levels of IL-6 are relatively stable over extended periods (41), a more accurate indicator of chronic interindividual variability in this measure would be derived from multiple assessments over time. Indeed, it is possible that the single measure of IL-6 used in this study reflects an acute event such as an infection. This possibility is reduced by our systematic exclusion of individuals with IL-6 levels greater than 10 pg/mL because our methods did not permit accurate quantification of higher levels. Although this may have resulted in our underestimation of the true relationship between IL-6 and cognitive function, it also decreased the probability that acute inflammatory activation influenced our findings. In the future, further testing is indicated in larger, age-stratified normative samples as well as longitudinal investigations beginning in early adulthood encompassing the full range of IL-6 and tracking changes in cognitive function over time. In this work, it will be important to determine whether variation in cognitive function associated with inflammatory processes in midlife adults is predictive of future cognitive decline and the onset of dementia.
Despite these shortcomings, our findings contribute to a growing body of evidence linking chronic inflammation to poorer cognitive functioning and extend these findings to a midlife community sample, raising the possibility that IL-6 may represent a biomarker for risk of future cognitive decline. The pattern of cognitive functioning associated with peripheral IL-6 seen in the present study is consistent with a large animal literature and the immune-to-brain communication model proposed by Maier and Watson (1).
The expert technical assistance of Cyndi Kravetz is gratefully acknowledged.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This study was supported by grant P01HL40962 from the National Heart Lung and Blood Institute.
DOI:10.1097/01.psy.0000238451.22174.92
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I. Soreca, C. Rosano, J. R. Jennings, L. K. Sheu, L. H. Kuller, K. A. Matthews, H. J. Aizenstein, and P. J. Gianaros Gain in Adiposity Across 15 Years is Associated With Reduced Gray Matter Volume in Healthy Women Psychosom Med, June 1, 2009; 71(5): 485 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Dickinson and P. D. Harvey Systemic Hypotheses for Generalized Cognitive Deficits in Schizophrenia: A New Take on An Old Problem Schizophr Bull, March 1, 2009; 35(2): 403 - 414. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Waldstein, S. C. Rice, J. F. Thayer, S. S. Najjar, A. Scuteri, and A. B. Zonderman Pulse Pressure and Pulse Wave Velocity Are Related to Cognitive Decline in the Baltimore Longitudinal Study of Aging Hypertension, January 1, 2008; 51(1): 99 - 104. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
