Quantitating hair loss in women: a critical approach.
BACKGROUND: Assessing how many hairs are actually shed by patients complaining of hair loss is a difficult task. Many methods have been proposed, but all need standardization. METHODS: We examined 234 women complaining of hair loss. Alopecia areata and scarring alopecias were excluded. Eighty-nine of the patients had an apparently normal density of hair. Seventy-four were classified as Ludwig stage I, 37 as Ludwig II and 15 as Ludwig III. In 19 patients, classification had not been recorded. They were tested with the pull test (PT), daily count (DC) and wash test (WT). Telogen percentage was obtained by a trichogram in 43 patients. RESULTS: Dispersion of the data was very high. The medians of the PT, DC and WT were 0.6, 60.5 and 122 hairs, respectively. The telogen median of percentage was 16. In patients with normally dense hair, the PT, DC and WT means were significantly higher than in those with Ludwig stages I-III. Telogen means did not differ. WT values correlated significantly with those of the DC and PT, and DC values did so with those of the PT. By contrast, telogen percentage did not correlate with WT, DC and PT values. WT values and telogen percentage increased in autumn, the latter preceding the WT by 1 month. CONCLUSIONS: The methods adopted and the problem of 'normality' are critically analyzed. Subjects complaining of hair loss proved to shed a higher number of hairs than those with various degrees of baldness. The PT is a poorly sensitive method, while telogen percentage is not correlated with the severity of hair loss. As the DC is a cumbersome procedure, the WT is probably the best method to adopt. Standardization of methods and assessment of normality in prepubertal children are imperative.