45, P < .048], CD3+CD69+ [REL: −0.53,

P < .016]) and only one negative correlation (with the cell activation marker CD4+CD69+) was significant for the group as a whole. In terms of muscle strength, significant positive correlations were found for several T cell activation markers and memory cell counts: CD3+HLA-DR+, CD3+CD25+HLA-DR+, CD4+CD25+HLA-DR+ and CD8+CD45RA+CD45RO+, although significant relationships were limited to the stronger half of our sample. Data for the group as a whole showed similar (but weaker) positive relationships and also a negative correlation with the CD3+CD4+CD8+ count ( Fig. 1). Neither natural killer cell cytotoxic activity nor lymphocyte proliferation data were significantly correlated with either aerobic power or muscle strength (data not shown). For the purpose of multiple regression analyses, a FITscore was calculated as a half of the sum of [Z aerobic power + Z selleck screening library muscle strength]. Other variables introduced into the equations were the depression, fatigue and quality of life indices and the carbohydrate intake. After appropriate Bonferroni adjustment of probability levels, many of the apparent relationships with the fitness score became non-significant, the only significant items being the numbers of regulatory cells CD3+HLA-DR+ and CD3+CD25+HLA-DR+ (Table 5). The depression score showed a positive association with the relative number of CD3+CD8+

(suppressor) cells, and a negative association with absolute numbers of CD3+CD25+HLA-DR+ EPZ015666 mouse regulatory cells. Fatigue scores showed a strong positive association with the numbers of mature CD56dim cells and with the relative numbers of CD4+CD45RO+ memory cells, and a strong negative relationship with PHA proliferation. A good QOL score also showed Montelukast Sodium positive relationships with the relative number of CD3+CD8+ cells and the relative numbers of CD4+CD45RO+ memory

cells (i.e. the opposite correlations found for depression), and negative associations with activation markers and PHA proliferative response (i.e. the opposite of the correlation that was found for fatigue). Carbohydrate intake showed only one weak positive association with an activation marker. Further regression analyses were calculated, testing a series of immune functions against depression, fatigue, QOL, carbohydrate intake and either aerobic power (Table 6a), muscle strength (Table 6b) or FITscore (Table 6c). The only positive correlations with the fitness variables were for CD3+HLA-DR+ (muscle strength and FITscore) and PHA proliferation (FITscore), although several positive relationships were found for depression, fatigue and QOL. Conclusions were essentially similar on progressively eliminating non-significant beta coefficients from these equations. Our data offer a substantial selection of normative values for lymphocyte subsets in sedentary but otherwise healthy older individuals.