Abstract
This investigation identified three autonomic factors in White and African-American men at rest using confirmatory factor analysis: A cardiac parasympathetic/heart rate variability factor (CPNS), a cardiac sympathetic/inotropic factor (CSNS) and a sympathetic vascular factor (VSNS). Respiratory sinus arrhythmia, heart rate variability, and heart rate maximum minus minimum loaded on CPNS. Heather Index, acceleration index and the ratio QT/QXon loaded on CSNS. Total peripheral resistance and diastolic blood pressure loaded positively on VSNS, whereas stroke volume loaded negatively. Factor scores were calculated and standardized to facilitate within- and between-subject comparisons. The CPNS and CSNS factors correlated 0.22, CPNS and VSNS correlated $-$0.31 and CSNS and VSNS correlated $-$0.87. CPNS scores were positively skewed, whereas CSNS and VSNS scores were normally distributed in the total sample. Age and body mass index (BMI) were significantly correlated with factor scores, with the exception of CPNS and BMI. Regression equations were calculated for each factor and age and BMI. The largest r$\sp2$ was 0.46 between age and VSNS. The CPNS factor and age correlated $-$0.43. After controlling for age and BMI, a trend was observed; seated African-American men had lower CPNS scores compared to their White counterparts. Seated African-American men had higher VSNS scores compared to the seated White men. The correlations of individual variables with the factors were examined; with the exception of SBP and DBP, these correlations were similar for seated and semi-reclined subjects. To graphically illustrate these associations, cardiac autonomic profiles were constructed using the upper and lower deciles of each factor. The results are discussed in the context of modes of autonomic control. It is concluded that the CPNS, CSNS and VSNS factors permit the exploration of autonomic regulatory patterns not possible with a unidimensional factor. Factor scores provide information not contained within an individual variable, and these factor scores may uncover group differences (eg. racial) which otherwise would not be apparent. In this manner, factor scores may be conceptualized as individual difference variables.