Journal of
J. Trace Elements Med. BioI. Vol. 12, pp. 231-235 (1998)
Trace Elements
in Medicine and Biology
© 1999 by Urban & Fischer
Validation of a Questionnaire Method for Estimating Extent
of Menstrual Blood Loss in Young Adult Women
A.-L. M. HEATH', C. M. SKEAFF and R. S. GIBSON
Department of Human Nutrition, University of Otago, PO Box 56, Dunedin, New Zealand
(JanuarylMay 1998)
Summary
The objective of this study was to validate two indirect methods for estimating the extent of menstrual
blood loss against a reference method to determine which method would be most appropriate for use in a pop-
ul ation of young adult women. Thirty-two women aged 18 to 29 years (mean ± SO; 22.4 ± 2.8) were recruited
by poster in Dunedin (New Zealand ). Data are presented for 29 women . A recall method and a record meth od
for estimating extent of menstrual loss were validated against a weighed reference method. Spearman rank
correlation coefficients between blood loss assessed by Weighed Menstrual Loss and Menstrual Record was
r,=0.47 (p=0.012), and between Weighed Menstrual Loss and Menstrual Recall , was rs=0 .61 (p=O.OOI) . The
Record method correctly classified 66% of participants into the same terti Ie , gross ly misclassifying 14%. The
Recall method correctly classified 59% of participants, grossly misclassifying 7%. Reference method men-
strualloss calculated for surrogate categories demonstrated a significant difference between the second and
third tertiles fo r the Record method, and between the first and third tertiles for the Recall method. The Men-
strual Recall method can differentiate between low and hi gh levels of menstrual blood loss in young adult
women, is quick to complete and analyse, and has a low parti ci pant burden.
Keywords: Menstruation, questionnaires, iron, adult, female, human.
Introduction
Iron def iciency states are the most common nutrition-
al deficiencies in developed countries. Menstrual blo od
loss has been shown to be an important determinant of
iron status amongst women of child-bearing age (1-4).
Therefore, any study investigating th e aetiology of iron
deficiency in this population must estimate the extent of
the participants' menstrual blood lo ss.
There is general th at the most accurate
method for measuring menstrual blood loss is chemical
analysis of the haem iron content of used menstrual prod-
ucts (5). The method of weighing menstrual fluid loss is
'To whom correspondence should be addressed
an appropriate reference method for ranking individuals
because of the high correlation between menstrual blood
and fluid loss (6). However, both these methods require
women to collect all used tampons and pads during a
menstrual cycle so are inappropriate in an epidemiologi-
cal research setting. They are inappropriate not only be-
cause of the high respondent burden, but also because of
the logistic problems involved in ensuring the safe and
complete collection, storage, transport and analysis of the
products. Alternate methods have asked participants to
estimate their losses as high, medium or low. The results
of these studies have shown that women cannot accurate-
ly estimate the extent of their own menstrual blood lo ss in
relation to the losses of women in general. Hallberg et al