Table 4 Assumptions of the modelling

1

Age- and gender-specific rates of institutionalization are derived from 2009 and remain unchanged between 2009 and 2025. The rational for this assumption is the uncertainty of further development of age- and gender-specific institutionalisation rates and the limited impact of such changes on the absolute number of community-dwelling persons.

2

Age- and gender-specific fracture rates derived from routine data by AOK Bavaria applies to all other community-dwelling older persons living in Bavaria since more than 50 % of persons aged 65 and older are covered by this insurance.

3

Participation in FPE and bisphosphonate treatment between 2004 and 2009 were not considered, i.e. assumed to be 0 % since there is no valid data on true participation and treatment rates available.

4

Age- and gender-specific fracture rates derived from 01.01.2004 to 30.06.2009 remain unchanged until 2025.

5

Effect size of Fall Prevention Exercise (FPE) is based on a meta-analysis with ‘reduction of any fracture rate’ as endpoint. Effect size of the reduction of femoral fractures is identical since there is no other data available.

6

FPE targets only persons aged 70 to 89 years.

7

Effect size is constant over time for both interventions.

8

Age- and gender-specific distribution of osteoporosis remains unchanged until 2025.

9

Bisphosphonate treatment is limited to people with osteoporosis based on BMD values (T-score -2.5). Concomitant risk factors modifying the threshold for treatment are not considered.

10

The effect size of bisphosphonates in men is identical to that in women since data on effect size in men is rare.