Perceived stress, cytomegalovirus titers, and late-differentiated T and NK cells: Between-, within-person associations in a longitudinal study of older adults

Abstract

Cytomegalovirus (CMV) and psychological stress are implicated as drivers of immunological aging. It is unknown, however, whether associations among CMV titers, stress, and immune aging are more stable or dynamic over time. The present investigation tested the between-person (stable differences) and within-person (dynamic fluctuations) associations of CMV titers and perceived stress on late-differentiated T and natural killer (NK) peripheral blood cells in a longitudinal study of older adults aged 64-92 years (N = 149). Participants reported stress levels and provided blood biannually for 2.5 years (up to 5 waves per person) to assess CMV IgG titers and composites of late-differentiated CD8 T cells (CD28- and CD57 + subsets) and CD56^dim NK cells (CD57+, NKG2C+, and FcεRIγ- subsets). In multilevel models that controlled for demographic variables, higher CMV titers were associated with higher proportions and counts of aged T and NK cells between people and lower counts of aged T cells within people. Perceived stress was associated with higher counts of aged T cells between people, but was not associated with aged NK cells. A significant interaction between stress and CMV titers on T cells between people indicated that older adults with lower stress levels and lower CMV titers had the lowest proportions of late-differentiated T cells, whereas those with higher stress levels had high proportions, regardless of CMV control. Our results provide evidence for longer-term, between-person associations among CMV titers, stress, and immunological aging, rather than dynamic within-person associations. We propose that targeting factors that promote low, stable perceived stress in older adults may retard T cell differentiation and ultimately support healthy aging.

Keywords: Aging; Cytomegalovirus; Immunosenescence; Longitudinal; Psychological stress.

Figure 1.

Illustrative plots of analyzed T cell populations. Lymphocytes were identified on the forward scatter/side scatter plot and calculated as a percentage of live cells. Viable lymphocytes (eBio viability stain) were identified. CD3+ cells were selected for and further identified as CD8+ cells. Proportions of CD8+ T cells expressing CD57+ and CD28− were calculated.

Figure 2.

Illustrative plots of analyzed NK cell populations. Lymphocytes were identified on the forward scatter/side scatter plot and calculated as a percentage of live cells. Viable CD3- cells were selected for and further identified as CD56^dim NK cells (low to moderate CD56 expression and high CD16 expression). Percentages of CD56^dim cells expressing CD57+ were calculated. In addition, the CD56^dim cells were divided into quadrants according to their NKG2C and FcεR1γ expression (quadrants Q1–4). Cells (%) expressing NKG2C+ were calculated by summing quadrants Q1 (FcεR1γ− NKG2C+) and Q2 (FcεR1γ+ NKG2C+). Cells (%) expressing FcεR1γ− were calculated by summing quadrants Q1 (NKG2C+ FcεR1γ−) and Q4 (and NKG2C- FcεR1γ-). Expression of NKG2C and FcεR1γ is displayed for a representative CMV seronegative donor and CMV seropositive donor.

Figure 3.

Simple slopes depicting the between-person differences in cytomegalovirus (CMV) titers and perceived stress on T cell composite (%) in the entire sample. Low CMV titers and low stress are 1 standard deviation (SD) units below the grand means; high CMV titers and high stress are 1 SD units above. Error bars represent standard errors of the estimates. The model was adjusted for lab scientist, CMV extrapolation, time, age at first wave, gender, income, and education.

Figure 4.

Simple slopes depicting the between-person differences in cytomegalovirus (CMV) titers and perceived stress on T cell composite (%) in the subset of CMV-positive older adults. Low CMV titers and low stress are 1 standard deviation (SD) units below the grand means; high CMV titers and high stress are 1 SD units above. Error bars represent standard errors of the estimates. The model was adjusted for lab scientist, CMV extrapolation, time, age at first wave, gender, income, and education.