Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Mark J. Newman


The relative proportions of lymphocyte subsets in the peripheral blood constitute a measure of immune status. Results of research to establish whether or not individuals with lung cancer differ in their lymphocyte composition have conflicted due to failure to control for factors other than cancer that affect lymphocyte balance. This study was designed to quantitatively estimate the effects of variables such as disease (lung cancer), smoking, age, circadian and circannual rhythms, and dietary factors on lymphocyte subsets. Questionnaires were given to 64 white males with primary lung cancer and 219 cancer-free white males and a blood sample was obtained. Lymphocytes were labelled with subset-specific, fluorescence-tagged antibodies and analyzed on a fluorescence-activated cell sorter (FACS). A computer algorithm for processing the FACS data was invented and utilized to determine proportions positive. The CD3-CD22-CD56 lymphocyte profile was modelled using bivariate loglinear multivariable regression. Disease, age, smoking, hour-of-day, month-of-year, weekly intake of vegetables, and percentages of daily calories from fat and from alcoholic drinks were significant main effects and there were significant age-by-smoking, disease-by-month, and disease-by-vegetable-intake interactions. Smokers' profiles did not mature with increasing age while non-smokers showed an increase in natural killer cells of 10 percentage points from age quartile $$68 years. Subjects with $>$28% of their calories in fat averaged 5% more CD3$\sp+$ and 4% fewer CD22$\sp+$ cells than those with $<$28% fat in their diets. Lung cancer patients had 1% fewer CD22$\sp+$ and 1% more CD3$\sp+$ lymphocytes than controls, had 11.7% HLA-DR$\sp+$ lymphocytes that were not CD22$\sp+$ compared to 6.6% for controls, and averaged 24% lymphocytes compared to 37% for controls. Their lymphocyte profiles were different from controls' in pattern of circannual variation and patients showed more extreme effect from a low-vegetable diet than controls. Patients reported a higher fat intake, and more occupational exposures to carcinogenic substances, but did not differ from controls in $\beta$-carotene consumption.