Premium
Clusters of variables influencing risk of cervical cancer
Author(s) -
Rotkin I. D.,
Cameron J. R.
Publication year - 1968
Publication title -
cancer
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.052
H-Index - 304
eISSN - 1097-0142
pISSN - 0008-543X
DOI - 10.1002/1097-0142(196804)21:4<663::aid-cncr2820210417>3.0.co;2-c
Subject(s) - cervical cancer , medicine , merge (version control) , variable (mathematics) , cluster (spacecraft) , variance (accounting) , principal component analysis , cancer , variables , demography , statistics , mathematics , computer science , mathematical analysis , accounting , sociology , business , information retrieval , programming language
Seven clusters of related variables are generated by a correlation matrix of 120 variables associated with cervical cancer. Ranked in order of relationship to the patient‐control cancer variable, these domains reveal two principal currents of influence—one originating in early sexual experience, the other emerging from sociologic background. Both currents merge in the dimension of early sexuality. The first‐order cluster contributes half of total discovered variance, with age at first coitus as the pivotal variable. The authors propose that adolescence may be a construct of superior meaningfulness over any individual factor or group of variables since carcinogenesis seems most likely to be initiated at this time. Profiles are derived of the female at greatest risk. Multidimensional analysis is discussed as an alternative to conventional single‐variable treatments, the strategy of cluster analysis being used in resolution of multiple interacting influences upon cancer risk.