Two hundred and twenty-four women (74 pre-, 90 peri-, 60 post-menopausal).
aged 46-59 years, from a population-based cohort participated in a longitud
inal study of bone mineral density (BMD). BMD was measured by dual-energy X
-ray absorptiometry (DXA) at the lumbar spine and femoral neck and the time
between bone scans was on average 25 (range 14-41) months. The aim of the
study was to assess changes in BMD in relation to changes in normal menopau
sal status. During the study period women who were between 3 and i' months
past their last menstrual period (n = 22, late perimenopausal) at the time
of the second bone scan had a mean (SE) annual change in BMD of -0.9% (0.4%
) at the lumbar spine and -0.7% (0.6%) at the femoral neck (both p < 0.05 c
ompared with women who remained premenopausal). In the women who became pos
tmenopausal (n = 42) the mean annual changes in BMD were -2.5% (0.2%) at th
e lumbar spine and -1.7% (0.2%) at the femoral neck (both p < 0.0005), and
in the women who remained postmenopausal (n = 60) they were -0.7% (0.2%) pe
r year and -0.5% (0.3%) per year respectively (both p < 0.05), compared wit
h women who remained premenopausal. In the 1-3 years after the final menstr
ual period (FMP) there was greater bone loss from the lumbar spine than the
femoral neck (p < 0.05). In women who were menstruating at the time of the
second bone scan and whose FMP could be dated prospectively (n = 35), high
er baseline oestradiol levels were associated with less lumbar spine bone l
oss (p < 0.005). In the women who remained postmenopausal there was an asso
ciation between baseline body mass index (BMI) and percentage change per ye
ar in femoral neck BMD (p<0.05), such that women with higher BMI had less b
one loss. In conclusion, during the time of transition from peri- to post-m
enopause, women had accelerated BMD loss at both the hip and spine.