Abstract
In the previous OCTO longitudinal study, we identified an immune risk phenotype (IRP) of high CD8 and low CD4 numbers and poor proliferative response. We also demonstrated that cognitive impairment constitutes a major predictor of nonsurvival. In the present NONA longitudinal study, we simultaneously examine in a model of allostatic load IRP and compromised cognition in 4-year survival in a population-based sample (n = 138, 86–94 years). Immune system measurements consisted of determinations of T-cell subsets, plasma interleukin 6 and cytomegalovirus and Epstein–Barr virus serology. Interleukin 2 responsiveness to concanavalin A, using data from the previous OCTO (octogenarians) immune study, hereafter OCTO immune, was also examined. Cognitive status was rated using a battery of neuropsychological tests. Logistic regression indicated that the IRP and cognitive impairment together predicted 58% of observed deaths. IRP was associated with late differentiated CD8+CD28−CD27− cells (p <.001), decreased interleukin 2 responsiveness (p <.05) and persistent viral infection (p <.01). Cognitive impairment was associated with increased plasma interleukin 6 (p <.001). IRP individuals with cognitive impairment were all deceased at the follow-up, indicating an allostatic overload.
INCREASED mortality rates are observed in elderly individuals with a compromised immune system and among those with cognitive impairment. Results from the present NONA longitudinal study and reevaluation of results from the previous OCTO longitudinal study examine the coexistence of the age-related immune system and cognition changes in relation to survival in a sample of very old humans.
The observed effects of immune and central nervous system dysregulation can be integrated into a general model of allostatic load, which suggests that cumulative dysregulations across multiple systems may have additive or multiple impacts on overall health and survival (1). The concept of allostasis is derived from homeostasis, but focuses more specifically on the challenges of regulatory mechanisms and the need of continuous evaluation and adaptation of our biological systems. The neuroendocrine, nervous, and immune systems constitute important mediators of adaptation to various challenges in maintaining stability through change (2). Given the complexity of influences from multiple systems, a longitudinal study offers unique opportunities to evaluate the effects of compromised cognition and immune function in relation to subsequent morbidity and mortality.
There is considerable evidence of age-associated changes in immune capabilities, including increased incidence of morbidity and mortality associated with altered immune function (3–5). An immune risk phenotype (IRP), consisting of high CD8 and low CD4 numbers, and poor proliferative response to concanavalin A (Con A), was initially identified using a cluster analysis approach (6). Subsequent analysis indicated that IRP could be defined using only the inverted CD4/CD8 ratio, because this sole marker is significantly associated with the IRP (7). This finding was also confirmed in a study showing that an inverted CD4/CD8 ratio predicted survival in a U.K. sample of elderly people (8). Recent studies (9,10) have extended these results to also indicate that seropositivity to cytomegalovirus (CMV) and an increase in the number of lately differentiated CD8+CD28− effector cells are likely to be associated with these changes.
On antigenic stimulation, CD8 T cells differentiate from naive cells to memory cells that are able to respond to reinfection or reactivation and to produce effector cells that secrete cytokines and have cytolytic capabilities that assist in control of infection (11–13). In healthy individuals, persistent viral agents are under control of virus-specific T cells. Latent viral agents, including CMV and Epstein–Barr virus (EBV), are able to coexist with their host only to be activated under certain conditions that involve immunosuppression (14). In HIV-infected individuals and in other immunosuppressed individuals, however, there is control failure with an increased incidence of clinically manifested disease associated with these latent viruses (14). This failure occurs in the presence of increases of virus-specific T-cell subpopulations (11–14).
Our understanding of virus-specific T-cell subpopulations has increased with the use of major histocompatibility (MHC) class I tetrameric complexes that stain antigen-specific T cells (15). It has been suggested that the determination of the expression of CD45RO, and the costimulatory receptors CD27 and CD28, provides a route to examine sequential pathways for differentiation of CD8 T cells (11–14,16). Using phenotypic markers, a sequential model has been proposed that starts with CD28+CD27+ representing early differentiated cells that finally progress to fully differentiated CD28−CD27− CD8 cells with shorter telomere length (16). This model has been used to examine the relationships of tetrameric complexes and the phenotype of virus-specific CD8 cells in the acute and chronic phases of persistent viral infections such as CMV and EBV (16).
Prevalence and incidence of cognitive impairment and dementia become substantial with age (17). It is also suggested that compromised cognition is significantly related to proximity to death (18). This pattern of terminal cognitive decline is further supported by findings of a substantially increased mortality risk among demented octogenarians (19) and nonagenarians (20). In the former study the observed risk was twofold, and in the latter they reported a hazard ratio of 1.24 for mildly impaired and 1.73 for severely cognitively impaired.
Evidence suggests interactions between the nervous and the innate immune systems, in which cytokines have a central role as mediators (21). Recent studies (22) suggest that cytokines operating in the central nervous system are significantly associated with cognitive functioning. Systemic inflammation in the brain tissue due to cytokine-mediated interactions between neurons and glial cells is suggested as a marker across the continuum of cognitive impairment and clinically manifested dementia (22). In particular, increased levels of the pro-inflammatory cytokines interleukin 1 (IL-1) and the multifunctional interleukin 6 (IL-6) seem to be associated with compromised cognition (22). In pathological conditions such as cerebral ischemia and Alzheimer's disease, microglia cells respond to injury by producing and secreting cytokines and cytotoxic factors to fight invading microorganisms and malfunctioning cells. The microglia response may also include a disconnection of synapses from injured neuronal cell bodies as part of a regenerative process with the purpose of facilitating reorganization of synaptic networks. The neurodegeneration in late life may in this context be seen as a chronic inflammatory condition that stimulates microglial activation (23).
The aim of the present study was to examine the relative importance of the IRP, here defined by the inversion of the CD4/CD8 ratio, and cognitive impairment in relation to 4-year mortality in a population-based sample of very old individuals. Specific aims include the analyses of associations between these conditions and persistent CMV and EBV viral infection, CD8+ T-cell differentiation, levels of plasma IL-6, and IL-2 production, using data from both the OCTO and NONA longitudinal studies.
Methods
Participants
The OCTO immune and NONA immune samples were recruited among participants in the NONA and OCTO Longitudinal Studies, in which population-based samples of oldest-old individuals are investigated with a broad-based battery for health and biobehavioral functioning. The samples were drawn in the municipality of Jönköping, located in South-Central Sweden. The present investigation examines primarily the NONA immune sample but includes also some reanalyses of data in the previously investigated OCTO immune sample.
Potential participants in the OCTO immune study were included if individuals aged 88–92 years were noninstitutionalized, had normal or only mild cognitive dysfunction according to neuropsychological tests (24), and were not on a drug regimen that might have influenced the immune system. The OCTO immune sample consisted at baseline of 102 individuals with a mean age of 88.3 years, all residing in ordinary or sheltered housing, with women constituting 66%. The sample has been described in more detail elsewhere (25). The sampling frame of the NONA immune study did not exclude individuals because of compromised health, and was based on available census information in September 1999 on which a nonproportional sampling procedure was used that included all individuals permanently residing in the municipality. The goal was to have an equal number of individuals aged 86, 90, and 94.
In the NONA immune study, blood was drawn at baseline in 138 individuals (42 belonged to the oldest birth cohort, 47 were 90 years old, and 49 were 86 years old). After 4 years, 68 (49%) were deceased. The mean age at baseline was 89.8, with women constituting 70%. Sixty-six percent resided in ordinary housing (often with support from the home help system), whereas 34% were in sheltered housing or in institutions.
In both studies, participants were examined in their place of residence by trained registered nurses (RNs) having extensive work experience with elderly persons. Blood samples were drawn between 9:00 am and 10:00 am. The health and biobehavioral functioning battery took about 3 hours to administer, including breaks, for individuals who were able to participate in all parts. The battery encompassed an interview and biomedical and behavioral assessments and ratings. Health status was defined based on information in medical records and from clinical chemistry, supplemented with information gathered in a health interview that focused on diagnosed illnesses, current symptoms, and medications. An informant interview was performed in all cases in which the individual was unable to participate in full due to compromised health and cognitive ability. Information from all these sources defined health status according to criteria for each individual (26).
Preparation of Plasma and Peripheral Blood Mononuclear Cells
Plasma was prepared from 40 ml of fresh blood drawn in EDTA tubes by centrifugation at 2500 rpm for 10 minutes, removed, and stored at −80°C. Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation on Lymphoprep (Nycomed Diagnostika, Oslo, Norway). The PBMC were washed three times in Dulbecco's phosphate-buffered saline (D-PBS) and resuspended in complete medium of RPMI 1640 (Invitrogen Life Technologies, Stockholm, Sweden), supplemented with 10% inactivated (56°C, 30 minutes) fetal calf serum (In Vitro Sweden AB, Stockholm), 2 mM l-glutamine, penicillin at 50 U/ml, and streptomycin at 50 μg/ml (Gibco, Stockholm, Sweden). Isolated PBMC counts, differential counts, white blood cell counts, complete blood cell counts, and examination of the whole blood were performed as previously described (6).
Mitogen Stimulation, IL-2 Production, and Plasma IL-6 Assays
Triplicate cultures of 5 × 104 PBMC in complete medium were placed in a total volume of 200 μl in 96-well flat-bottom microtiter culture plates (Costar, Cambridge, MA). Cells were stimulated with Con A at 5 and 20 μg/ml (Sigma, Stockholm, Sweden).
Mitogen response was determined by incubating cultures for 48 hours in a humidified 5% CO2 incubator at 37°C. Each well was pulsed for 18 hours with 0.55 μCi of methyl-3H-thymidine (Sigma). Radioactivity was measured in a Beckman Liquid Scintillation Counter (Beckman Coulter AB, Bromma, Sweden) and expressed as mean disintegrations per minute.
IL-2 production was evaluated by assaying the proliferative effect of supernatants from cultures of PBMC stimulated with Con A at 5 and 20 μg/ml as described above. Serial dilutions of the supernatants were incubated at 37°C for 18 hours with CTLL-2 [murine cytotoxic T cells responsive to IL-2 but not to interleukin 4 (IL-4)] cells in microtiter plates and pulsed the last 8 hours before harvest with 1 μCi of methyl-3H-thymidine. The Biological Response Modifiers Program (BRMP) standard was included in every assay. By using the Allfit software program provided by J. L. Rossio (Cancer Institute, Frederick, MD), we expressed the results in BRMP units per milliliter. Control cells produced nondetectable amounts of IL-2.
Levels of plasma IL-6 were analyzed by the use of commercial enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Stockholm, Sweden). The assay was performed as outlined in the kit protocols. All samples were measured in duplicate, and the IL-6 concentration was calculated using a Dynex MRXII spectrophotometer (Dynex Technologies, Ashford, Middlesex, U.K.).
Flow Cytometric Analysis
Monoclonal antibodies, including appropriate isotype controls, were purchased from BD Biosciences (Stockholm, Sweden). The data were acquired using FACScan (BD Immunocytometry Systems, Jönköping, Sweden) and analyzed using CellQuest software (BD Biosciences). The staining protocol for three-color staining included the CD3Per-CP/CD4FITC/CD8PE, CD3Per-CP/CD8FITC/CD28PE, and CD8Per-CP/CD27PE/CD45RAFITC combinations in tubes 4–6. In tube 1, control IgG1FITC/IgG1PE/IgG1Per-CP was used to create an analysis gate to include lymphocytes in forward-scatter light versus side-scatter light and to set the fluorescence quadrant markers on the fluorescent light channel 1 versus fluorescent light channel 2 to detect the presence of any nonantigen-specific antibody binding (nonspecific staining). In tube 2, CD3 FITC/CD4PE/CD8Per-CP was used to create a gate set on the CD3-positive lymphocyte fraction and to adjust compensation on FL1 versus FL2. In tube 3, CD45RAFITC/CD45ROPE/CD8Per-CP was used to create a gate set on the CD8-positive lymphocyte fraction. Quality control was performed using daily CaliBRITE beads and FACSComp software (both from BD Biosciences) for setting the photomultiplier tube (PMT) voltages and the fluorescence compensation, and for checking instrument sensitivity prior to use. Internal quality control was performed to check consistency for CD markers included in more than one tube and resulted in a variation coefficient of 5%–7%.
CMV and EBV Serology
Immunoassay (MEIA; Abbot, Stockholm, Sweden) was used to detect anti-CMV IgG antibodies in plasma. The procedure followed the manufacturer's instruction. Immunoassay (NOVITEC, EBNA-1 IgG antibody Test, HiSS Diagnostic Gmbh, Freiburg, Germany) was also used to detect anti-EBV IgG antibodies in plasma according to the manufacturer's instructions.
CD4/CD8 Ratio
A CD4/CD8 ratio less than 1.00 was used to define the IRP in the present study. Cluster analysis used in the previous OCTO immune study was not used, because the mitogen response to Con A and CD19 immune components necessary for replication of clusters were not investigated. In calculating the CD4/CD8 ratio, numbers of CD4 and CD8 cells were used. Use of CD4 and CD8 percentages resulted in identical CD4/CD8 ratio categories. Eight cases were found to be close to the cutoff border in the range 0.90–1.10.
Health, Cognitive Status, and CD4/CD8 Ratio
The overall health status at baseline of the NONA immune sample (26) indicated that only 13 individuals (9.4%) were rated as healthy, according to the European SENIEUR protocol criteria often used in studies of aging and immunology. Thirty-eight (27.5%) participants, defined as moderately healthy, met the following criteria used in the OCTO Immune Study (26): not residing in an institution, not being demented, and not using medication known to effect the immune system. The remaining sample (63.0%) comprised frail individuals not meeting the above health criteria (26). The percentages of very old individuals with an inverted CD4/CD8 ratio were 15.3%, 10.8%, and 18.4%, rated as healthy, moderately healthy, and frail, respectively (not statistically significant).
The brief neuropsychological battery included The Mini-Mental State Examination (MMSE) (27) and the Memory-in-Reality (MIR) tests (28). The MMSE is a screening device extensively used in epidemiological studies to identify cognitive impairment. The MIR-test comprises, first, a naming condition for 10 common real-life objects. Next, a three-dimensional model of an apartment is shown, and the participant is asked to place the objects in the different rooms according to personal preferences. Following a distraction, a recall test is administered, followed by a recognition task for items not recalled. Finally, participants are instructed to place the objects in the same locations as they were placed initially. Individuals were also asked to rate their own memory and cognitive ability. In addition, the RN rated the individuals' overall comprehension and understanding, and their degree of reliability in accurately responding to the interview.
A rating of cognitive status was conducted (by B.J.) using the above information and by taking overall health and functioning into account, including sensory and motor handicaps. In this respect, the rating of cognitive status represents a clinical evaluation based on multiple sources of information. In the present study we use the following cognitive status categories: Cognitively intact, mild cognitive dysfunction or questionable cases [MCD, evidence of compromised memory and/or cognition, not fully meeting Diagnostic and Statistical Manual of Mental Disorders–Fourth Edition (DSM-IV) criteria for dementia], and dementia (according to DSM-IV criteria).
Among those individuals from whom a blood sample was drawn, 20 (14.5%) were diagnosed with dementia [of these 20 cases, 15 were diagnosed with Alzheimer's disease according to criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS/ADRDA)]. Another 20 individuals (14.5%) were rated as MCD cases at baseline. These two diagnostic categories for cognitive status were pooled under the label of cognitive impairment and compared with those rated as cognitively intact in subsequent analyses. The percentages of individuals with cognitive impairment were 15.3% (2 of 13), 5.3% (2 of 38), and 41.4% (36 of 87), in the overall health categories healthy, moderately healthy, and frail, respectively (p <.01). The four individuals with compromised cognition in the categories of healthy and moderately healthy were all MCD cases.
Data Analysis
Statistical analyses were conducted using SPSS 11 for Windows (SPSS, Chicago, IL). Parametric analyses of variance (ANOVAs), Student t tests, nonparametric Kruskal–Wallis tests, and Mann–Whitney U tests were used for comparisons of independent groups. Correlation analysis was performed using the Pearson correlation coefficient. The chi-square test, logistic and linear regression, and Kaplan–Meier survival analysis were used to analyze 4-year survival or nonsurvival in categories of individuals by CD4/CD8 ratio (less than 1.00 and greater than 1.00) and by cognitive status (intact or impairment). Logistic regression was conducted with CD4/CD8 ratio and cognitive status as independent variables and with 4-year survival (survival or deceased) as dependent variable. Also, an interaction term (CD4/CD8 ratio × cognitive status) was included in the model. Age, sex, cardiovascular disease, chronic obstructive pulmonary disease, all malignancy, pernicious anemia, rheumatoid arthritis, diabetes mellitus, and hypothyroidism were entered as covariates.
Results
NONA Immune Study
Mortality, CD4/CD8 ratio, and cognitive status
Four years after baseline, 68 (49%) individuals were deceased. Table 1 shows that the individuals with an inverted CD4/CD8 ratio had significantly higher relative 4-year mortality (77%; 17 of 22) than those with a ratio greater than one (43%; 50 of 115). Cognitively impaired individuals showed a significantly higher relative 4-year mortality (75%; 30 of 40) than did cognitively intact individuals (39%; 38 of 98). The 4-year mortality was 90% among the 20 individuals diagnosed with dementia and 60% among the 20 individuals diagnosed with questionable cognitive dysfunction.
Among the 22 individuals with a CD4/CD8 ratio less than one, 4 had dementia (18%) and 4 had MCD (18%). Of the 116 with a CD4/CD8 ratio greater than one, 16 had dementia (14%) and 16 had MCD (14%), which was an identical relative number of dementia cases relative to MCD as for individuals with a CD4/CD8 ratio less than one.
Logistic regression analysis revealed a significant main effect for CD4/CD8 ratio (B = 1.51, SE(B) = 0.56, Wald = 7.2; p <.01) and cognitive status (B = 1.60, SE(B) = 0.43, Wald = 13.7; p <.001) with no significant interaction effect between CD4/CD8 ratio and cognitive status. The outcome was not significantly affected by entering the covariates age, sex, cardiovascular disease, chronic obstructive pulmonary disease, malignancy, anemia, rheumatoid arthritis, diabetes, and hypothyroidism. The independent main effects predicted 39 (58.2%) of 67 nonsurvivors and 55 (78.6%) of 70 survivors observed 4 years after baseline, corresponding to an overall percentage of correctly classified survivors/nonsurvivors of 68.6.
Subgroups defined by their combination of CD4/CD8 ratios and cognitive status were compared for 4-year mortality, i.e., individuals with a CD4/CD8 ratio less than or greater than one combined with those with cognitive impairment and those without (Table 2). The results indicated a significantly higher mortality (100%) in individuals with a CD4/CD8 ratio less than one and cognitive impairment (n = 8) compared with the other subgroups. Individuals with inverted CD4/CD8 and no cognitive impairment (n = 14), and those with a CD4/CD8 ratio greater than one and cognitive impairment (n = 32), had intermediate mortality (64% and 69%, respectively). Cognitively intact individuals with a CD4/CD8 ratio greater than one (n = 83) showed the lowest mortality (34%).
Kaplan–Meier survival analysis with survival curves for the subgroups defined by their combination of CD4/CD8 ratios and cognitive status for the 4-year study period is shown in Figure 1. In a linear regression analysis, the annual mortality was estimated for these subgroups. The mortality was 8.5% for those participants with CD4/CD8 > 1 and intact cognition, and 15% both in the CD4/CD8 < 1 and intact cognition subgroup and among those participants with CD4/CD8 > 1 and cognitive impairment. The corresponding annual mortality rate in the subgroup with CD4/CD8 < 1 and cognitive impairment was 42%.
CD4/CD8 ratio: T-cell subsets, plasma IL-6, CMV IgG, and EBV IgG antibodies
Individuals with a CD4/CD8 ratio less than one and those with a ratio greater than one differed significantly in their numbers of CD3+CD8−CD4+ cells (537 and 866 cells/mm3, respectively; p <.05). Also, a profound increase was found in the number of CD3+CD8+CD4− cells in individuals with a CD4/CD8 ratio less than one (866 and 345 cells/mm3, respectively; p <.001).
The increase in CD3+CD8+CD4− cells occurred particularly for the CD3+CD8+CD28− and CD8+CD27− subsets (Table 3). Correlation coefficients indicated a close association between the CD8+CD27+ and the CD3+CD8+CD28+ numbers (r = 0.90, p <.001) and the CD8+CD27− and CD3+CD8+CD28− numbers (r = 0.94, p <.001).
There were no significant differences in plasma IL-6 between individuals with a CD4/CD8 ratio less than and greater than one (median values, 3.1 and 3.2 pg/ml, respectively; not statistically significant).
All 22 individuals with a CD4/CD8 ratio less than one were CMV+EBV+, whereas 23 of 115 individuals (20%) were CMV− (n = 18) or EBV− (n = 5) among those with a ratio greater than one (p <.01).
One-way ANOVAs were performed at baseline for CD3CD8CD28 and CD8CD27CD45RA subset numbers relative to the CMV and EBV antibody status. The ANOVAs indicated that the CD3+CD8+CD28−, CD8+CD27−CD45RA−, and CD8+CD27−CD45RA+ numbers increased significantly across subgroups in the order CMV−EBV+, CMV+EBV−, CMV+EBV+ (Table 4). A post hoc Scheffe test indicated significantly higher numbers for CD3+CD8+CD28−, CD8+CD27−CD45RA−, and CD8+CD27−CD45RA+ cells in the CMV+EBV+ individuals compared to those who were CMV−EBV+ (Table 4).
Cognitive status: T-cell subsets, plasma IL-6, CMV IgG, and EBV IgG antibodies
Similar numbers of CD3CD8CD28 and CD8CD27CD45RA subsets were found in individuals with cognitive impairment compared with cognitively intact (no data shown). There were no relationships between cognitive status and the prevalence of CMV or EBV IgG antibodies (no data shown).
Mann–Whitney U tests of the amount of plasma IL-6 showed significant (p <.001) increases for individuals with cognitive impairment (mean rank 89.7, median 6.3 pg/ml, n = 40) compared with those cognitively intact (mean rank 61.2, median 2.5 pg/ml, n = 98). No significant differences were found comparing those with dementia (mean rank 91.5, median 6.3 pg/ml, n = 20) with MCD (mean rank 88.6, median 6.1 pg/ml, n = 20).
CD4/CD8 ratio and cognitive status: CD8CD27CD45RA subsets and plasma IL-6
A one-way ANOVA of the number of cells in CD8CD27CD45RA subsets in subgroups of CD4/CD8 ratio and cognitive function is presented in Table 5. The ANOVA indicated significant differences across subgroups for the CD8+CD27−CD45RA− and CD8+CD27−CD45RA+ subsets. A post hoc Scheffe test indicated significantly higher numbers of CD8+CD27−CD45RA− cells in individuals with inverted CD4/CD8 ratio combined with cognitive impairment compared with subgroups with a CD4/CD8 ratio greater than one. A Scheffe test also indicated significantly higher numbers of CD8+CD27−CD45RA+ cells in individuals with an inverted CD4/CD8 ratio combined with no cognitive impairment compared to those participants with a CD4/CD8 ratio greater than one (Table 5).
A Kruskal–Wallis test indicated significant differences in plasma IL-6 across the subgroups of individuals created by combining CD4/CD8 ratio with cognitive status (Table 6). Subgroups with cognitive impairment indicated significantly higher level of plasma IL-6 compared with subgroups without. Individuals with the combination of CD4/CD8 less than one and cognitive impairment demonstrated the highest levels compared with all other subgroups, even when compared with those with cognitive dysfunction and a CD4/CD8 ratio greater than one (p <.05).
OCTO Immune Study
Mortality, CD4/CD8 ratio, cognitive status, mitogen, and IL-2 production
We reanalyzed previously collected OCTO data on mortality, cognitive status, IL-2, and mitogen response to Con A in the light of the NONA findings presented above. Individuals with cognitive impairment severe enough to meet the criteria for dementia were not included in the previous OCTO immune study at baseline because of applied exclusion criteria. Individuals with MCD, however, participated. IL-2 and mitogen responsiveness were not investigated in the present NONA Immune Study.
Our reanalysis of the OCTO immune sample indicated that of the 14 IRP individuals identified at baseline by cluster analysis (6), 13 had a CD4/CD8 ratio less than one and 1 CD4/CD8 ratio equal to one was allocated to the IRP group. This finding indicated a close association between IRP and an inverted CD4/CD8 ratio. Another two individuals with a CD4/CD8 ratio less than one were missing cases in the previous cluster analysis, which used several parameters to form clusters (6).
Four years after baseline, 43 of 95 (45%) individuals were deceased (3 individuals dropped out in the analysis). Individuals with an inverted CD4/CD8 ratio (n = 16) had significantly (p <.05) higher 4-year mortality (69%), compared with those with a ratio greater than one (40%; 31 of 78, 1 dropout). MCD individuals (n = 31) showed a significantly (p <.01) higher 4-year mortality (68%) compared with cognitively intact individuals (34%; 22 of 64).
Among the 16 individuals with a CD4/CD8 ratio less than one, 4 had MCD and 12 were cognitively intact. Of the 79 individuals with a CD4/CD8 ratio greater than one, 26 had MCD (33%) and 52 were cognitively intact (66%, 1 dropout). Individuals with a CD4/CD8 ratio less than one and MCD (n = 4) were all deceased after 4 years. In cognitively intact individuals with a CD4/CD8 ratio less than one (n = 12) and in those participants with a ratio greater than one and MCD (n = 26), intermediate mortality was found (58% and 61%, respectively). Cognitively intact individuals with a ratio greater than one (n = 52) showed the lowest mortality (29%). These results confirm those reported above for the NONA sample in Table 2.
The levels of IL-2 production on stimulation with the mitogen Con A (20 μg/ml) is shown in Figure 2 in subgroups of individuals defined by their combination of CD4/CD8 ratio and cognitive status. One-way ANOVA, with post hoc Scheffe test, indicated that the subgroups with a CD4/CD8 ratio less than one produced a significantly lower amount of IL-2 compared to individuals in the subgroups with a ratio greater than one (p <.05). The figure also indicates a very low IL-2 production for the 4 individuals with MCD and a CD4/CD8 ratio less than one compared with other subgroups. In addition, correlation analysis revealed a significant positive association between IL-2 production and cognitive functioning according to the MMSE score (r = 0.55, p <.05) in individuals with a CD4/CD8 ratio less than one. No association between the variables was found among those participants with a CD4/CD8 ratio greater than one or in the subgroup of individuals with a CD4/CD8 ratio greater than one combined with cognitive impairment. Similar results were obtained for stimulation with Con A at 5 μg/ml (no data shown).
A pattern similar to the results for the IL-2 production was obtained for the mitogen response to Con A. For Con A stimulation at 20 μg/ml the means 2862, 5566, 18,270, and 17,988 dpm for the subgroups with CD4/CD8 < 1, MCD; CD4/CD8 < 1, cognitively intact; CD4/CD8 > 1, MCD; and CD4/CD8 > 1, cognitively intact; were obtained, respectively (p <.05). This also occurred in the results for the Con A stimulation at 5 μg/ml (no data shown).
Discussion
We investigated 4-year mortality in very old individuals with an IRP defined as a CD4/CD8 ratio less than one and in those with cognitive impairment, conditions previously identified as major predictors of nonsurvival in late life (6,7,19). The results confirm the previous findings of an elevated mortality risk in individuals with the IRP. The IRP predicted 2-year nonsurvival using immune data both at baseline in 1989 and 2 years later in the previous OCTO immune longitudinal study using exclusion criteria to select a sample with relatively good health (6,7). The present findings demonstrate that the results can be generalized also to the more broadly defined NONA population-based sample, not excluding individuals with compromised health. In addition, an elevated risk of nonsurvival for elderly individuals with inversion of the CD4/CD8 ratio was confirmed in a UK sample of the Healthy Ageing Study emerging from Cambridge and Nottingham and including individuals without physical frailty and cognitive impairment (8). The validity of the CD4/CD8 cutoff of 1.00 used in the present and the UK studies could be questioned. Analysis excluding eight NONA immune cases at the border (0.90–1.10), however, did not change the outcome of the present study in any way.
The present study also confirms previous findings from the OCTO longitudinal study of an elevated mortality risk in individuals with cognitive impairment. In addition, our results show that the two conditions independently predicted a majority of the observed deaths even controlling for age, sex, and the seven most prevalent diseases in the NONA immune sample (26). This result is in agreement with our previous findings that the IRP appears independently of the individuals' state of health (26). Although IRP and cognitive impairment were independently predictive of mortality, a small number of individuals with both conditions were also identified. An annual mortality rate of 8.5% among individuals without IRP or cognitive impairment compared with 15% among individuals with one or the other of the factors and 42% among those with both risk factors corresponds to the relative mortality rates of 1:2:2:5. This finding indicates an increase of the mortality risk in individuals having both risk factors that exceeded that of an additive effect suggesting an allostatic overload (29). McEwen and colleagues proposed allostatic load as a term for cumulative effects of physiologic dysregulation across multiple systems, with particular importance in late life (1,2,29,30). It was hypothesized that allostatic load may have considerable impact on health and survival. When the condition of only mild cognitive impairment was combined with IRP in our study, the mortality rate increased 2.5 times, indicating a substantially elevated risk.
An allostatic load in IRP individuals with cognitive impairment is also supported by changes in the levels of cytokines, suggested to be primary mediators of allostasis (30). Excessive increases in the plasma levels of the proinflammatory cytokine IL-6, associated with neurodegenerative processes and cognitive impairment (21), and pronounced decreases in IL-2 responsiveness, indicating a failure in the T cellular response, represent changes in these individuals associated with an allostatic load (30). Although inflammatory and immune responses are adaptive in the short term, they are likely to be damaging when excessive in duration or magnitude (29).
IRP and Persistent Viral Infection
The present findings in the NONA study of an association between the IRP and the evidence of persistent CMV infection (prevalence more than 90%) confirms our results from the previous OCTO immune study (9). Similar immune changes as in the OCTO were also confirmed in the present study pinpointing a CD8+CD27−CD28−CD45RA+/RO+ phenotype as markedly expanded for IRP individuals. This finding suggests allostatic T-cell changes by generation of effector CD8+ cells against the CMV infection. Several authors have reported that clonal expansions in the CD8+CD28− subsets occur and persist in very old individuals (31–38). Significant expansions of CMV-specific CD8+ cells for the dominant human pp65-derived epitope NLVPMVATV associated with the IRP was also demonstrated in OCTO participants using tetramer technology (39). The relative percentages of CMV-specific expansions were in the range of 1%–27% of total CD8+ cells, similar to findings by Khan and coworkers (38), who also demonstrated that these CMV-specific cells are composed of lately differentiated effector cells with a CD28−CD57+CCR7− phenotype. Such huge expansions within one or a few clones might give rise to loss of clonal diversity and immune protective capability (39–41). Increasing evidence also suggests that increased proportions of lately differentiated CD8+ cells indicate characteristics of replicative senescence with possible cell exhaustion and decline in T-cell competence (42,43). In line with these findings, it has been found that individuals with chronic CMV infection and/or high proportions of associated CD8+CD28− T cells show markedly diminished antibody responses to influenza vaccination (44–46).
The present study also supports the view that, besides persistent CMV infection, persistent EBV infection plays a role as a bystander that is associated with the IRP. IRP individuals were in all cases both CMV and EBV seropositive, suggesting that a chronic viral load in elderly persons might be important for developing an IRP. Significant expansions of EBV-specific CD8+ cells relative the human leukocyte antigen (HLA)-A2-restricted epitope GLCTLVAML were indeed also demonstrated in NONA immune participants, although their frequency was 10-fold lower than was the frequency of the CMV-specific cells (47). The present study demonstrates increased numbers of lately differentiated CD8+CD28−CD27− cells, characteristic of the IRP, particularly in CMV- and EBV-coinfected individuals, to a less but still significant extent in those individuals being infected with CMV only, and to a small extent in those individuals infected with EBV only. Enrichments of cells at different stages of differentiation or depletions of a particular phenotype according to the viral specificity or to a coinfecting agent during chronic infection have been demonstrated by others (16,48).
The fact that only about one fourth of the individuals with persistent CMV infection, with EBV as bystander, resided in the IRP category indicates that, besides being seropositive, there are other factors of importance in the development of the IRP. Various genetic influences (43), the functioning of thymus, and the duration of chronic viral infections may affect the individuals' ability to manage chronic infection. The immune system changes associated with the IRP may also reflect various stress-related factors that are quite different but ultimately result in similar immune changes (41).
Cognitive Functioning, IL-6, and IL-2
Various studies indicate that an increased level of the proinflammatory cytokine IL-6 is highly associated with age and many age-related diseases such as cardiovascular disease, arthritis, osteoporosis, type 2 diabetes, and Alzheimer's disease (49), many of which are commonly represented in our population-based studies. Our data indicate that IL-6 in plasma was particularly associated with poor cognitive functioning and, thereby, to increased mortality. Individuals with cognitive impairment showed significantly higher IL-6 plasma levels and mortality rates than did those individuals who were cognitively intact. It is noteworthy that we found weak associations between IL-6 and other age-related diseases such as cardiovascular disease (p =.053). Individuals with the combination of IRP and cognitive impairment, however, showed significantly higher plasma IL-6 as well as mortality than did individuals with cognitive impairment only, suggesting that the IRP may magnify the risk for a more pronounced and rapid decline in vitality in individuals with cognitive impairment. These results support findings that have linked IL-6 to progressive inflammatory disorders of the brain, including poorer cognitive functioning and increased risk for a cognitive decline and mortality (50).
In the present study, as expected, the majority of dementia cases were those of Alzheimer's type. Amyloid peptides involved in Alzheimer's disease are derived from the amyloid precursor protein (APP) with progressive deposition of the cleaved fragments Abeta 1-42,43 considered to be of importance in development of the disease. APP peptides in humans are shown to be targets for the immune system as evidenced by IL-2 expression and cell proliferation when stimulating lymphocytes from healthy individuals. A lack of responsiveness of lymphocytes to APP peptides is however seen in Alzheimer's disease patients (51). In mice it was demonstrated that intracerebral injection of an APP peptide produces an IL-6 increase both in the brain and in plasma, suggesting a direct IL-6 induction associated with Alzheimer's disease (52). This suggests the possibility of using IL-6 as a marker to characterize the immune response to Alzheimer's disease during disease progression, in line with the finding that peripheral IL-6 and IL-1beta secretion from lipopolysaccharide-stimulated blood cells of Alzheimer's disease patients are significantly decreased with disease severity (53). The finding that caregivers for a spouse with dementia had an average increase in IL-6 about four times that among noncaregivers (54) suggests that also chronic stress, likely to be associated with early stages of Alzheimer's disease, may accelerate the increases in IL-6 during progression of Alzheimer's disease.
IRP and Cognitive Impairment, IL-2, IL-6, and T-Cell Differentiation
On antigen recognition, specific CD8+ T cells need to differentiate from an early naive CD27+CD28+CD45RA+ stage, via intermediate CD27+CD28+CD45RO+ and CD27−CD28−CD45RO+ memory and/or effector stages, and finally into a lately CD27−CD28−CD45RA+ effector cell stage (11,12,14,16). Our data suggest a close association between subsets with the CD27 and CD28 markers, indicating that the CD8+CD27+CD45RA+ subset measured in our study includes naive cells, the CD8+CD27+CD45RA− and CD8+CD27−CD45RA− intermediate stages, and CD8+CD27−CD45RA+ later effector cell stages. Our finding that IRP individuals with cognitive impairment had fewer cells in the late-differentiated stages and higher numbers in the intermediate stages when compared with cognitively intact IRP individuals indicates impaired CD8+ maturation into fully differentiated effector cells in these individuals. The failing mechanism of the immune CD8+ T-cell differentiation into effector cells has been demonstrated to be an important factor in the progression of HIV into clinically manifested AIDS and to be involved in the decreased immunity to latent EBV and CMV infections in immunosuppressed individuals (16).
A failure of CD8+ cells to function and to properly differentiate into CTLs might be a consequence of impaired T-cell help in response to viral antigens such as CMV and EBV (16). IL-2 is important in the CD4+ helper function for expansion of CD8+ cells during infection as well as for the maintenance of CD8+ CTLs (55). A suppression of the cellular response already in the early T-helper cell activation in IRP individuals is supported by our finding of poorer IL-2 production in response to Con A in these individuals compared with non-IRPs. An association between the degree of responsiveness and cognitive functioning in IRP individuals also supports an interaction between nervous- and peripheral immune system dysfunctions with further down-regulation of the cellular response in those individuals with cognitive impairment. Elevated IL-6 levels might have contributed to this suppression by acting as an indirect immunosuppressant via the hypothalamic–pituitary-adrenal axis (56). These results might be compatible with findings demonstrating a lack of IL-2 responsiveness of lymphocytes to APP peptides in Alzheimer patients (51) and those findings showing that peripheral IL-6 and IL-1beta secretion from LPS-stimulated blood cells of Alzheimer's disease patients are significantly decreased in individuals with severe Alzheimer's disease (53).
The present longitudinal study focused on changes in the immune and nervous systems relative to the allostatic effects of the combination of cognitive impairment, persistent CMV and EBV persistent infection, changes in CD8+ T-cell differentiation, plasma IL-6 levels, and IL-2 production in population-based samples of very old persons. The study supports the potential value of the application of measurements from multiple systems to evaluate the effects associated with allostatic load and morbidity and mortality in very late life.
Decision Editor: James R. Smith, PhD
![survival curves for [1] CD4/CD8 < 1, cognitively impaired (n = 8); [2] CD4/CD8 < 1, cognitively intact (n = 14); [3] CD4/CD8 > 1, cognitively impaired (n = 32); and [4] CD4/CD8 > 1, cognitively intact (n = 83) subgroups of NONA immune individuals. Test for equality of survival distribution for the subgroups were: log rank: 44.80, p <.0001](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/biomedgerontology/60/5/10.1093_gerona_60.5.556/1/m_biomedgerontology_60_5_556_f1.gif?Expires=1726687026&Signature=O~gduJT-LHtOxbiTI-M~hdDc7VdGmL3agPjJ2hNNRkTWUIGZETBjK0CGcdUdvqKYj9sW9PSf6jDFcSRvjDUXDkPGFj5L7Wg0k4DOse62gzcMaoXuLFZEiUVuEkqEfImOgTt7aDDn~7AM5UyvvdlvmFXSXszTKVXaI5-80guHPsFjaL~DV0fpaXmZpbfb7VDhJwtCwlu7CBAvYfMnGW5LGrBE-HDaozQn6fI5PLeuZOYazK7ILBpfss8eha5JlkPliPA1sinFDFE~3SzIfAOMglwhVx7fyMq4TjcJvekI2TCLfB7K~-iq1p~G5auE2BOYaUuAgaWs1l3bTVA2VB~sPQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
survival curves for [1] CD4/CD8 < 1, cognitively impaired (n = 8); [2] CD4/CD8 < 1, cognitively intact (n = 14); [3] CD4/CD8 > 1, cognitively impaired (n = 32); and [4] CD4/CD8 > 1, cognitively intact (n = 83) subgroups of NONA immune individuals. Test for equality of survival distribution for the subgroups were: log rank: 44.80, p <.0001
Mean of interleukin 2 (IL-2) production in response to concanavalin A (Con A; 20 μg/ml) in subgroups of OCTO immune individuals. 1, CD4/CD8 < 1 and cognitively impaired (n = 4); 2, CD4/CD8 < 1 and cognitively intact (n = 12); 3, CD4/CD8 > 1 and cognitively impaired (n = 26); and 4, CD4/CD8 > 1 and cognitively intact (n = 53). One-way analysis of variance including a post hoc Scheffe test indicated that the groups with CD4/CD8 ratio less than one had significantly lower IL-2 production compared to the groups with a ratio greater than one (p <.05)
Chi-Square Analysis of Relationships of Survivors and Nonsurvivors in Subgroups of Very Old NONA Immune Individuals by CD4/CD8 Ratio and Cognitive Status.
| Subgroup | Number of | p < | ||||
|---|---|---|---|---|---|---|
| Survivors | Nonsurvivors | |||||
| CD4/CD8 | ||||||
| <1.0 | 5 | 17 (77%)* | .005 | |||
| >1.0 | 65 | 50 (43%) | ||||
| Cognitive status | ||||||
| Impairment | 10 | 30 (75%) | .001 | |||
| Intact | 60 | 38 (39%) | ||||
| Subgroup | Number of | p < | ||||
|---|---|---|---|---|---|---|
| Survivors | Nonsurvivors | |||||
| CD4/CD8 | ||||||
| <1.0 | 5 | 17 (77%)* | .005 | |||
| >1.0 | 65 | 50 (43%) | ||||
| Cognitive status | ||||||
| Impairment | 10 | 30 (75%) | .001 | |||
| Intact | 60 | 38 (39%) | ||||
Note: *Relative 4-year mortality rate.
Chi-Square Analysis of Relationships of Survivors and Nonsurvivors in Subgroups of Very Old NONA Immune Individuals by CD4/CD8 Ratio and Cognitive Status.
| Subgroup | Number of | p < | ||||
|---|---|---|---|---|---|---|
| Survivors | Nonsurvivors | |||||
| CD4/CD8 | ||||||
| <1.0 | 5 | 17 (77%)* | .005 | |||
| >1.0 | 65 | 50 (43%) | ||||
| Cognitive status | ||||||
| Impairment | 10 | 30 (75%) | .001 | |||
| Intact | 60 | 38 (39%) | ||||
| Subgroup | Number of | p < | ||||
|---|---|---|---|---|---|---|
| Survivors | Nonsurvivors | |||||
| CD4/CD8 | ||||||
| <1.0 | 5 | 17 (77%)* | .005 | |||
| >1.0 | 65 | 50 (43%) | ||||
| Cognitive status | ||||||
| Impairment | 10 | 30 (75%) | .001 | |||
| Intact | 60 | 38 (39%) | ||||
Note: *Relative 4-year mortality rate.
Chi-Square Analysis of Relationships of Survivors and Nonsurvivors in Subgroups of Very Old NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| CD4/CD8 Cognitive Status | Number of | ||
|---|---|---|---|
| Survivors | Nonsurvivors | ||
| Less than one, impairment | 0 | 8 (100%)* | |
| Less than one, intact | 5 | 9 (64%) | |
| Greater than one, impairment | 10 | 22 (69%) | |
| Greater than one, intact | 55 | 28 (34%) | |
| CD4/CD8 Cognitive Status | Number of | ||
|---|---|---|---|
| Survivors | Nonsurvivors | ||
| Less than one, impairment | 0 | 8 (100%)* | |
| Less than one, intact | 5 | 9 (64%) | |
| Greater than one, impairment | 10 | 22 (69%) | |
| Greater than one, intact | 55 | 28 (34%) | |
Notes: Chi square = 22.4; p <.001.
*Relative 4-year mortality.
Chi-Square Analysis of Relationships of Survivors and Nonsurvivors in Subgroups of Very Old NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| CD4/CD8 Cognitive Status | Number of | ||
|---|---|---|---|
| Survivors | Nonsurvivors | ||
| Less than one, impairment | 0 | 8 (100%)* | |
| Less than one, intact | 5 | 9 (64%) | |
| Greater than one, impairment | 10 | 22 (69%) | |
| Greater than one, intact | 55 | 28 (34%) | |
| CD4/CD8 Cognitive Status | Number of | ||
|---|---|---|---|
| Survivors | Nonsurvivors | ||
| Less than one, impairment | 0 | 8 (100%)* | |
| Less than one, intact | 5 | 9 (64%) | |
| Greater than one, impairment | 10 | 22 (69%) | |
| Greater than one, intact | 55 | 28 (34%) | |
Notes: Chi square = 22.4; p <.001.
*Relative 4-year mortality.
Student's t Test for the Number of Cells (cells/mm3) in CD3CD8CD28 and CD8CD27CD45RA Subsets in Very Old NONA Immune Individuals With a CD4/CD8 Ratio Less Than or Greater Than One.
| Subset | Group | p < | ||
|---|---|---|---|---|
| CD4/CD8 < 1 | CD4/CD8 > 1 | |||
| CD3+CD8+CD28+ | 250 ± 29 (22)* | 168 ± 10 (113) | .01 | |
| CD3+CD8+CD28− | 611 ± 63 (22) | 200 ± 15 (113) | .001 | |
| CD8+CD27+CD45RA+ | 123 ± 24 (22) | 85 ± 6 (115) | NS | |
| CD8+CD27+CD45RA− | 154 ± 21 (22) | 116 ± 8 (115) | NS | |
| CD8+CD27−CD45RA− | 187 ± 42 (22) | 72 ± 8 (115) | .05 | |
| CD8+CD27−CD45RA+ | 476 ± 70 (22) | 199 ± 12 (115) | .001 | |
| Subset | Group | p < | ||
|---|---|---|---|---|
| CD4/CD8 < 1 | CD4/CD8 > 1 | |||
| CD3+CD8+CD28+ | 250 ± 29 (22)* | 168 ± 10 (113) | .01 | |
| CD3+CD8+CD28− | 611 ± 63 (22) | 200 ± 15 (113) | .001 | |
| CD8+CD27+CD45RA+ | 123 ± 24 (22) | 85 ± 6 (115) | NS | |
| CD8+CD27+CD45RA− | 154 ± 21 (22) | 116 ± 8 (115) | NS | |
| CD8+CD27−CD45RA− | 187 ± 42 (22) | 72 ± 8 (115) | .05 | |
| CD8+CD27−CD45RA+ | 476 ± 70 (22) | 199 ± 12 (115) | .001 | |
Notes: *Mean ± SE (n).
NS = not significant.
Student's t Test for the Number of Cells (cells/mm3) in CD3CD8CD28 and CD8CD27CD45RA Subsets in Very Old NONA Immune Individuals With a CD4/CD8 Ratio Less Than or Greater Than One.
| Subset | Group | p < | ||
|---|---|---|---|---|
| CD4/CD8 < 1 | CD4/CD8 > 1 | |||
| CD3+CD8+CD28+ | 250 ± 29 (22)* | 168 ± 10 (113) | .01 | |
| CD3+CD8+CD28− | 611 ± 63 (22) | 200 ± 15 (113) | .001 | |
| CD8+CD27+CD45RA+ | 123 ± 24 (22) | 85 ± 6 (115) | NS | |
| CD8+CD27+CD45RA− | 154 ± 21 (22) | 116 ± 8 (115) | NS | |
| CD8+CD27−CD45RA− | 187 ± 42 (22) | 72 ± 8 (115) | .05 | |
| CD8+CD27−CD45RA+ | 476 ± 70 (22) | 199 ± 12 (115) | .001 | |
| Subset | Group | p < | ||
|---|---|---|---|---|
| CD4/CD8 < 1 | CD4/CD8 > 1 | |||
| CD3+CD8+CD28+ | 250 ± 29 (22)* | 168 ± 10 (113) | .01 | |
| CD3+CD8+CD28− | 611 ± 63 (22) | 200 ± 15 (113) | .001 | |
| CD8+CD27+CD45RA+ | 123 ± 24 (22) | 85 ± 6 (115) | NS | |
| CD8+CD27+CD45RA− | 154 ± 21 (22) | 116 ± 8 (115) | NS | |
| CD8+CD27−CD45RA− | 187 ± 42 (22) | 72 ± 8 (115) | .05 | |
| CD8+CD27−CD45RA+ | 476 ± 70 (22) | 199 ± 12 (115) | .001 | |
Notes: *Mean ± SE (n).
NS = not significant.
One-Way Analysis of Variance of Number of Cells (cells/mm3) in CD3CD8CD28 and CD8CD27CD45RA Subsets in NONA Immune Individuals Relative to Cytomegalovirus (CMV) Immunoglobulin G (IgG) and Epstein–Barr Virus (EBV) IgG Antibody Status.
| Subset | Subgroup | p < | ||||
|---|---|---|---|---|---|---|
| CMV+EBV+N = 115 | CMV+EBV−N = 5 | CMV−EBV+N = 18 | ||||
| CD3+CD8+CD28+ | 183 ± 11 | 217 ± 18 | 160 ± 32 | NS | ||
| CD3+CD8+CD28− | 308 ± 24* | 204 ± 58 | 64 ± 16 | .001 | ||
| CD8+CD27+CD45RA+ | 94 ± 7 | 110 ± 13 | 72 ± 12 | NS | ||
| CD8+CD27+CD45RA− | 121 ± 8 | 143 ± 32 | 127 ± 24 | NS | ||
| CD8+CD27−CD45RA− | 110 ± 13† | 64 ± 32 | 19 ± 4 | .05 | ||
| CD8+CD27−CD45RA+ | 265 ± 20† | 201 ± 51 | 112 ± 14 | .05 | ||
| Subset | Subgroup | p < | ||||
|---|---|---|---|---|---|---|
| CMV+EBV+N = 115 | CMV+EBV−N = 5 | CMV−EBV+N = 18 | ||||
| CD3+CD8+CD28+ | 183 ± 11 | 217 ± 18 | 160 ± 32 | NS | ||
| CD3+CD8+CD28− | 308 ± 24* | 204 ± 58 | 64 ± 16 | .001 | ||
| CD8+CD27+CD45RA+ | 94 ± 7 | 110 ± 13 | 72 ± 12 | NS | ||
| CD8+CD27+CD45RA− | 121 ± 8 | 143 ± 32 | 127 ± 24 | NS | ||
| CD8+CD27−CD45RA− | 110 ± 13† | 64 ± 32 | 19 ± 4 | .05 | ||
| CD8+CD27−CD45RA+ | 265 ± 20† | 201 ± 51 | 112 ± 14 | .05 | ||
Notes: Mean ± SE.
*p <.001 compared with CMV−EBV+ (Scheffe test).
† p <.05 compared with CMV−EBV+ (Scheffe test).
NS = not significant.
One-Way Analysis of Variance of Number of Cells (cells/mm3) in CD3CD8CD28 and CD8CD27CD45RA Subsets in NONA Immune Individuals Relative to Cytomegalovirus (CMV) Immunoglobulin G (IgG) and Epstein–Barr Virus (EBV) IgG Antibody Status.
| Subset | Subgroup | p < | ||||
|---|---|---|---|---|---|---|
| CMV+EBV+N = 115 | CMV+EBV−N = 5 | CMV−EBV+N = 18 | ||||
| CD3+CD8+CD28+ | 183 ± 11 | 217 ± 18 | 160 ± 32 | NS | ||
| CD3+CD8+CD28− | 308 ± 24* | 204 ± 58 | 64 ± 16 | .001 | ||
| CD8+CD27+CD45RA+ | 94 ± 7 | 110 ± 13 | 72 ± 12 | NS | ||
| CD8+CD27+CD45RA− | 121 ± 8 | 143 ± 32 | 127 ± 24 | NS | ||
| CD8+CD27−CD45RA− | 110 ± 13† | 64 ± 32 | 19 ± 4 | .05 | ||
| CD8+CD27−CD45RA+ | 265 ± 20† | 201 ± 51 | 112 ± 14 | .05 | ||
| Subset | Subgroup | p < | ||||
|---|---|---|---|---|---|---|
| CMV+EBV+N = 115 | CMV+EBV−N = 5 | CMV−EBV+N = 18 | ||||
| CD3+CD8+CD28+ | 183 ± 11 | 217 ± 18 | 160 ± 32 | NS | ||
| CD3+CD8+CD28− | 308 ± 24* | 204 ± 58 | 64 ± 16 | .001 | ||
| CD8+CD27+CD45RA+ | 94 ± 7 | 110 ± 13 | 72 ± 12 | NS | ||
| CD8+CD27+CD45RA− | 121 ± 8 | 143 ± 32 | 127 ± 24 | NS | ||
| CD8+CD27−CD45RA− | 110 ± 13† | 64 ± 32 | 19 ± 4 | .05 | ||
| CD8+CD27−CD45RA+ | 265 ± 20† | 201 ± 51 | 112 ± 14 | .05 | ||
Notes: Mean ± SE.
*p <.001 compared with CMV−EBV+ (Scheffe test).
† p <.05 compared with CMV−EBV+ (Scheffe test).
NS = not significant.
One-Way Analysis of Variance of Number of Cells (cell/mm3) in CD8CD27CD45RA Subsets in Subgroups of NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| Subset | CD4/CD8 < 1 | CD4/CD8 > 1 | p < | ||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| CD8+CD27+CD45RA+ | 89 ± 16 | 143 ± 37 | 85 ± 10 | 85 ± 7 | NS | ||
| CD8+CD27+CD45RA− | 178 ± 26 | 140 ± 26 | 128 ± 17 | 111 ± 8 | NS | ||
| CD8+CD27−CD45RA− | 262 ± 98* | 144 ± 34 | 63 ± 11 | 75 ± 10 | .001 | ||
| CD8+CD27−CD45RA+ | 385 ± 111 | 528 ± 90† | 191 ± 23 | 203 ± 15 | .001 | ||
| Subset | CD4/CD8 < 1 | CD4/CD8 > 1 | p < | ||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| CD8+CD27+CD45RA+ | 89 ± 16 | 143 ± 37 | 85 ± 10 | 85 ± 7 | NS | ||
| CD8+CD27+CD45RA− | 178 ± 26 | 140 ± 26 | 128 ± 17 | 111 ± 8 | NS | ||
| CD8+CD27−CD45RA− | 262 ± 98* | 144 ± 34 | 63 ± 11 | 75 ± 10 | .001 | ||
| CD8+CD27−CD45RA+ | 385 ± 111 | 528 ± 90† | 191 ± 23 | 203 ± 15 | .001 | ||
Notes: Mean ± SE.
*p <.001 compared with CD4/CD8 > 1 cognitively impaired and intact (Scheffe test).
† p <.001 compared with CD4/CD8 > 1 cognitively impaired and intact (Scheffe test).
NS = not significant.
One-Way Analysis of Variance of Number of Cells (cell/mm3) in CD8CD27CD45RA Subsets in Subgroups of NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| Subset | CD4/CD8 < 1 | CD4/CD8 > 1 | p < | ||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| CD8+CD27+CD45RA+ | 89 ± 16 | 143 ± 37 | 85 ± 10 | 85 ± 7 | NS | ||
| CD8+CD27+CD45RA− | 178 ± 26 | 140 ± 26 | 128 ± 17 | 111 ± 8 | NS | ||
| CD8+CD27−CD45RA− | 262 ± 98* | 144 ± 34 | 63 ± 11 | 75 ± 10 | .001 | ||
| CD8+CD27−CD45RA+ | 385 ± 111 | 528 ± 90† | 191 ± 23 | 203 ± 15 | .001 | ||
| Subset | CD4/CD8 < 1 | CD4/CD8 > 1 | p < | ||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| CD8+CD27+CD45RA+ | 89 ± 16 | 143 ± 37 | 85 ± 10 | 85 ± 7 | NS | ||
| CD8+CD27+CD45RA− | 178 ± 26 | 140 ± 26 | 128 ± 17 | 111 ± 8 | NS | ||
| CD8+CD27−CD45RA− | 262 ± 98* | 144 ± 34 | 63 ± 11 | 75 ± 10 | .001 | ||
| CD8+CD27−CD45RA+ | 385 ± 111 | 528 ± 90† | 191 ± 23 | 203 ± 15 | .001 | ||
Notes: Mean ± SE.
*p <.001 compared with CD4/CD8 > 1 cognitively impaired and intact (Scheffe test).
† p <.001 compared with CD4/CD8 > 1 cognitively impaired and intact (Scheffe test).
NS = not significant.
Test of Plasma Interleukin 6 (IL-6; pg/ml) in Subgroups of NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| CD4/CD8 < 1 | CD4/CD8 > 1 | p < | |||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| IL-6 | 108.5 (8.2)* | 58.5 (2.4) | 84.1 (5.5) | 61.1 (2.5) | .001 | ||
| CD4/CD8 < 1 | CD4/CD8 > 1 | p < | |||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| IL-6 | 108.5 (8.2)* | 58.5 (2.4) | 84.1 (5.5) | 61.1 (2.5) | .001 | ||
Note: *Mean rank (median).
Test of Plasma Interleukin 6 (IL-6; pg/ml) in Subgroups of NONA Immune Individuals Created by CD4/CD8 Ratio Combined With Cognitive Status.
| CD4/CD8 < 1 | CD4/CD8 > 1 | p < | |||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| IL-6 | 108.5 (8.2)* | 58.5 (2.4) | 84.1 (5.5) | 61.1 (2.5) | .001 | ||
| CD4/CD8 < 1 | CD4/CD8 > 1 | p < | |||||
|---|---|---|---|---|---|---|---|
| Impairment (N = 8) | Intact (N = 14) | Impairment (N = 32) | Intact (N = 83) | ||||
| IL-6 | 108.5 (8.2)* | 58.5 (2.4) | 84.1 (5.5) | 61.1 (2.5) | .001 | ||
Note: *Mean rank (median).
We acknowledge the considerable support of the European Union project “T Cell Immunity and Ageing” (T-CIA), contract no. QLK6-CT-2001-02283, the Research Board in the County Council of Jönköping, and the Research Council in the Southeast of Sweden (FORSS) for funding of this project. We also acknowledge Länssjukhuset Ryhov for providing laboratory resources for completion of the studies, and are also indebted to our coworker Per-Erik Evrin. We also would particularly like to thank the nursing staff including Lena Svensson, Lena Blom, Monica Janeblad, and Gun Karlsson for their efforts in obtaining blood samples and Florence Confer, Lisa Stark, and Andrea Tompa for their technical assistance.
References
McEwen BS, Stellar E. Stress and the individual: mechanism leading to disease.
McEwen BS. Interacting mediators of allostasis and allostatic load: towards an understanding of resilience in aging.
Murasko DM, Weiner P, Kaye D. Decline in mitogen induced proliferation of lymphocytes with increasing age.
Roberts-Thomson IC, Wittingham S, Youngchaiyud U, Mackay IR. Ageing, immune response and mortality.
Wayne SJ, Rhyne RL, Garry PJ, Goodwin JS. Cell mediated immunity as a predictor of morbidity and mortality in subjects over 60.
Ferguson FG, Wikby A, Maxson P, Olsson J, Johansson B. Immune parameters in a longitudinal study of a very old population of Swedish people: a comparison between survivors and nonsurvivors.
Wikby A, Maxson P, Olsson J, Johansson B, Ferguson FG. Changes in CD8 and CD4 lymphocyte subsets, T cell proliferation responses and non-survival in the very old: the Swedish longitudinal OCTO-immune study.
Huppert FA, Pinto EM, Morgan K, Brayne C. Survival in a population sample is predicted by proportions of lymphocyte subsets.
Olsson J, Wikby A, Johansson B, Löfgren S, Nilsson BO, Ferguson FG. Age-related change in peripheral blood T-lymphocyte subpopulations and cytomegalovirus infection in the very old: the Swedish longitudinal OCTO immune study.
Wikby A, Johansson B, Olsson J, Löfgren S, Nilsson BO, Ferguson F. Expansions of peripheral blood CD8 T-lymphocyte subpopulations and an association with cytomegalovirus seropositivity in the elderly: the Swedish NONA immune study.
Hamann D, Roos MT, van Lier RA. Faces and phases of human CD8 T-cell development.
Tomiyama H, Matsuda T, Takiguchi M. Differentiation of human CD8+ cells from a memory to memory/effector phenotype.
Wills MR, Okecha G, Weekes MP, Gandhi MK, Sissons PJ, Carmichael AJ. Identification of naive or antigen-experienced human CD8+ T cells by expression of costimulation and chemokine receptors: analysis of the human cytomegalovirus-specific CD8+ T-cell response.
van Baarle D, Kostense S, van Oers MHJ, Hamann D, Miedema F. Failing immune control as a result of impaired CD8+ T-cell maturation: CD27 might provide a clue.
Altman JD, Moss PA, Goulder PJ, et al. Phenotypic analysis of antigen-specific T lymphocytes.
Appay V, Dunbar PR, Callin M, et al. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections.
Fratiglioni L, De Ronchi D, Aguero-Torres H. Worldwide prevalence and incidence of dementia.
Wilson RS, Beckett LA, Bienias JL, Evans DA, Bennet DA. Terminal decline in cognitive function.
Johansson B, Zarit SH. Early cognitive markers of the incidence of dementia and mortality: a longitudinal population-based study of the oldest old.
Nybo H, Petersen HC, Gaist D, et al. Predictors of mortality in 2249 nonagenarians: the Danish 1905 Cohort Survey.
Wilson CJ, Finch CE, Cohen HJ. Cytokines and cognition: the case for a head-to-toe inflammatory paradigm.
Akiyama H, Barger S, Barnum S, et al. Inflammation and Alzheimer's disease.
Johansson B, Zarit SH, Berg S. Changes in cognitive functioning of the oldest old.
Wikby A, Johansson B, Ferguson F, Olsson J. Age-related changes in immune parameters in a very old population of Swedish people: a longitudinal study.
Nilsson BO, Ernerudh J, Johansson B, et al. Morbidity does not influence the T-cell immune risk phenotype in the elderly: findings in the Swedish NONA Immune Study using sample selection protocols.
Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical method for grading the cognitive state of patients for the clinician.
Karlamangla AS, Singer BH, McEven BS, Rowe JW, Seeman TE. Allostatic load as a predictor of functional decline: MacArthur studies of successful aging.
McEwen BS, Wingfield JC. The concept of allostasis in biology and biomedicine.
Wang ECY, Moss PAH, Frodsham P, Lehner PJ, Bell JI, Borysiewicz LK. CD8highCD57+ T lymphocytes in normal, healthy individuals are oligoclonal and respond to human cytomegalovirus.
Schwab R, Szabo P, Manavalan JS, et al. Expanded CD4+ and CD8+ T cell clones in elderly humans.
Mugniani EN, Egeland T, Spurkland A, Brinchmann JE. The T-cell receptor repertoire of CD8+CD28- T-lymphocytes is dominated by expanded clones that persist over time.
Posnett DN, Edinger JW, Manavalan JS, Irwin C, Marodon G. Differentiation of human CD8 T cells: implications for in vivo persistence of CD8+CD28− cytotoxic effector clones.
Weekes MP, Wills MR, Mynard K, Carmichael AJ, Sissons JGP. The memory cytotoxic T lymphocyte (CTL) response to human cytomegalovirus infection contains individual peptide-specific CTL clones that have undergone extensive expansion in vivo.
Weekes MP, Carmichael AJ, Wills MR, Mynard K, Sissons JGP. Human CD28-CD8+ T-cells contain greatly expanded functional virus-specific memory CTL clones.
Ricalton NS, Roberton C, Norris JM, Rewers M, Hamman RF, Kotzin BL. Prevalence of CD8+ T cell expansion in relation to age in healthy individuals.
Khan N, Shariff N, Cobbold M, et al. Cytomegalovirus seropositivity drives the CD8 T cell repertoire toward greater clonality in healthy elderly individuals.
Ouyang Q, Wagner WM, Wikby A, et al. Large numbers of dysfunctional CD8+ T lymphocytes bearing receptors for a single dominant CMV epitope in the very old.
LeMaoult J, Messaoudi I, Manavalan JS, et al. Age-related dysregulation in CD8 T cell homeostasis: kinetics of a diversity loss.
Wikby A, Johansson B, Ferguson F. The OCTO and NONA immune longitudinal studies: a review of 11 years studies of Swedish very old humans.
Papagno L, Spina CA, Marchant A, et al. Immune activation and CD8(+) T-cell differentiation towards senescence in HIV-1 infection.
Pawelec G, Akbar A, Caruso C, Effros R, Glubeck-Loebenstein B, Wikby A. Is immunosenescence infectious?
Trzonkowski P, Mysliwska J, Szmit E, et al. Associations between cytomegalovirus infection, enhanced proinflammatory response and low levels of anti-hemagglutinins during the anti-influenza vaccination–an impact of immunosenescence.
Saurwein-Teissl M, Lung TL, Marx F, et al. Lack of antibody production following immunization in old age: association with CD8(+)CD28(−) T cell clonal expansions and an imbalance in the production of Th1 and Th2 cytokines.
Goronzy JJ, Fulbright JW, Crowson CS, et al. Value of immunological markers in predicting responsiveness to influenza vaccination in elderly individuals.
Ouyang Q, Wagner WM, Walter S, et al. An age-related increase in the number of CD8(+) T cells carrying receptors for an immunodominant Epstein-Barr virus (EBV) epitope is counteracted by a decreased frequency of their antigen-specific responsiveness.
Elrefaei M, El-Sheikh N, Kamal K, Cao H. HCV-specific CD27-CD28- memory T cells are depleted in hepatitis C virus and Schistosoma mansoni co-infection.
Forsey RJ, Thompson JM, Ernerudh J, et al. Plasma cytokine profiles in elderly humans.
Weaver JD, Huang MH, Albert M, Harris T, Rowe JW, Seeman TE. Interleukin-6 and risk of cognitive decline: MacArthur Studies of Successful Aging.
Trieb K, Ransmayr G, Sgonc R, Lassmann H, Grubeck-Loebenstein B. APP peptides stimulate lymphocyte proliferation in normals, but not in patients with Alzheimer's disease.
Song DK, Im YB, Jung JS, Cho J, Suh HW, Kim YH. Central beta-amyloid peptide-induced peripheral interleulin-6 responses in mice.
Sala G, Galimberti G, Canvari C, et al. Peripheral cytokine release in Alzheimer patients: correlation with disease severity.
Kiecolt-Glaser JK, Preacher KJ, MacCallum RC, Atkinson C, Malarkey WB, Glaser R. Chronic stress and age-related increases in the proinflammatory cytokine Il-6.
Francavilla V, Accapezzato D, De Salvo M, et al. Subversion of effector CD8+ T cell differentiation in acute hepatitis C virus infection: exploring the immunological mechanisms.
