Table 5 Non vertebral fractures occurring during the study

In our prospective controlled 3-year study we found that fluoride given as monofluorophosphate in combination with calcium increases BMD at different skeletal sites and thereby reduces the incidence of new vertebral and Non vertebral fractures in comparison with calcium alone. Comparing the results of the two fluoride groups different responses in terms of BMD changes and fracture rates can be recognized. While in group A, who received low-dose intermittent fluoride, average BMD increased moderately at all sites (e.g., mean annual increase at the spine 4.2%), in group B, who received continuous higher-dose fluoride, there was a higher increase at the more spongy bone sites but a slight average loss at the more cortical areas (Table 3). Although statistically not significant there is a clear tendency toward higher vertebral and non vertebral fracture rates in group B as compared with group A. The following possible explanations for these Obviously dose-related differences between the two fluoride groups are suggested:

1. Lower fluoride doses lead to moderate linear increases at mainly spongy bone sites leaving enough time for adequate mineralization and ameli-oration of mechanical stability. With this dosage same there are no negative effects on more cortical bone sites, some patients even showing slight increases (see Table 3, group A).

2. With higher fluoride levels the average rates of increase at the cancellous bone sites are higher, however, in some patients optimal mechanical stability is not necessarily achieved. These higher rates of increase may occur in part at the expense of the more cortical skeletal regions (see Table 3, group B). This view is supported by comparing the individual changes in BMD at the lumbar spine and the femoral neck given in Fig,. 3. The suggestion that the higher fluoride dosage might have adverse effects on cortical bone in some patients helps to explain the tendency toward a higher rate of Non vertebral fractures in group B but also a higher rate of vertebral fractures. Concerning the latter the important contribution of cortical bone to vertebral strength is often underestimated [9].

    

Positive effects of MFP/calcium therapy on cortical bone were also demonstrated recently in a study from Austria [7], and an observational long-term trial from Spain showed a significant reduction in non vertebral fractures [6]. Typical examples for high fluoride doses are the two placebo-controlled studies from the Mayo Clinic and from the Henry Ford Hospital published in 1990 and 1991 [1 0.11]. In the Mayo Clinic trial the Mean dosage of sodium fluoride (NaF) was 75 mg/day, corresponding to 34 mg of fluoride ions. The BMD of the lumbar spine increased by 36% after 4 years, i.e., 9% per annum. This rapid and rather large increase was not followed by a decrease in vertebral fracture rate. Bone density at the pure cortical measuring site of the radius decreased. After extension of this study and further analysis the same authors reported 4 years later an inverse correlation between fluoride dosage and the rate of avoided vertebral fractures [4]. This strongly supports the above mentioned interpretation of our study results. In the meantime further studies have proved the ability of MFP or sodium fluoride to reduce the number of new vertebral fractures when given in daily doses corresponding to amounts of fluoride ions between 10 and 20 mg [5,6,12,13]. On the other hand a French-Belgian multi-center study [14] failed to show a fracture-reducing effect in  spite off using sodium fluoride or MFP at doses equivalent to 20 mg fluoride ions per day. As possible  explanations the only 2 year duration of that particular trial and the obviously very effective therapy of  the control group with calcium and vitamin D leading to rather low fracture rates in both groups were discussed. In the light of our study (see Tables 3 and Table 4) a lower daily dose of fluoride might have been advantageous. No definite recommendation can currently be pro- vided as to whether fluoride therapy should be given intermittently or continuously. The intervals in the intermittent schedules used to date were quite different [5,8,15]. A recent study from South Africa evaluating, histologic parameters found fewer histomorphometric abnormalities with intermittent than with continuous slow-release sodium fluoride therapy [16]. Our study shows a smaller number of LEPS events in the group with intermittent therapy (Table 2). In conclusion, our findings indicate that compared with calcium therapy 3 years of MFP/calcium therapy giving 15 mg fluoride intermittently or 20 mg continuously significantly increases BMD and reduces vertebral and non vertebral fracture rates in women with established postmenopausal osteoporosis. Despite a lower gain of bone mass at the spine with the lower- dose intermittent MFP schedule there was a clear tendency toward fewer vertebral fracture events with this regimen.   When planning, the therapeutic strategy for individual patients it should also be taken into account that fluoride is cheap in comparison with other modern therapies for postmenopausal osteoporosis and that in the low-dose range of fluoride therapy described in this study only very rarely do mild side effects occur. Furthermore there is increasing evidence that combinations of MFP/calcium with an antiresorptive agent are a very promising therapy. Recent data from a placebo- controlled Danish study show a clear additive effect of MFP and HRT [15].