RESUMEN
We have examined the value of the Premenstrual Assessment Form (PAF) in 109 women with prospectively validated symptomatic premenstrual change. Results were compared with those of non-complaining women previously published (J Psychosom Res 1990; 34: 439-446.), there was considerable overlap in categorical and numerical scales so that selection of homogeneous populations of complainers could not be achieved. However, fewer than 7% of complainers had a total PAF numerical score in the normal range (defined as 2 SD above the mean total score of controls). Factor analysis of numerical scores in women with symptomatic premenstrual change suggested three factors similar to but not identical with those of non-complaining women. These were physiological depression, anxiety-volatility and increased well-being. While the negative syndromes are compatible with clinical experience, prospective investigation will be required to assess their homogeneity.