Wasserman S S, Cohen M M, Schwartz S
Department of Medicine, University of Maryland School of Medicine, Baltimore.
Environ Mol Mutagen. 1990;16(4):255-9. doi: 10.1002/em.2850160407.
The latent "factors" influencing spontaneous and clastogen-induced genetic damage, measured by rates of sister chromatid exchange (SCE) and chromosome breakage (CB), were investigated in a small sample of 20 unrelated, healthy individuals. The covariation of spontaneous and clastogen-induced (bleomycin [BLM], streptonigrin [SN], mitomycin-C [MMC], 4-nitroquinoline-1-oxide [4NQO]) SCEs and CBs was analyzed by maximum-likelihood factor analysis. A single-factor model resulted in large standardized regression coefficients of measured variables on the factor for spontaneous and BLM- and SN-induced SCE frequencies, and a modest regression coefficient for MMC-induced SCEs. A two-factor model, after varimax rotation, yielded one factor strongly associated with spontaneous and BLM- and SN-induced SCE frequencies, and a second factor associated with spontaneous and BLM- and SN-induced CBs. A bootstrap analysis of this data set indicated the statistical significance of one regression coefficient (i.e., P less than or equal to 0.05) and borderline significance (0.07 less than or equal to P less than or equal to 0.11) of three other regression coefficients on the first factor, to be interpreted as an effector of SCE frequencies. However, for the second factor, none of the bootstrapped regression coefficients was significant (P greater than 0.22). Due to the modest sample utilized in this study, the validity of this model should be further explored using additional, larger data sets.