Background: Increased physical activity is associated with better physical andmental well-being. In postmenopausal women, land-based exercise has a modesteffect on bone strength which is predominantly reflected in the spatial distribution(geometry) of bone mineral at the skeletal sites targeted by the exercise. However, the risks and benefits of exercise for women with osteoporotic vertebralfracture (VFs) who are at high risk for future fracture are not known. Women withestablished osteoporosis may prefer exercising in water where compressiveloads are reduced. However, it is not known if the reduced compressive loadingis detrimental to bone health. Also contributing to this evidence gap is the limitedmethods available for measuring vertebral bone geometry which may beexpected to respond to exercises targeting the trunk. Bone geometry can bemeasured using computed tomography scans but this involves exposure to asubstantial dose of radiation. Semi-automated analyses of dual energy X-rayabsorptiometry (DXA) vertebral fracture assessment (VFA) scans providemeasures of vertebral height (VH); however, the measurement properties of thisoutcome have not been established. Measures of mechanical bone strength andvolumetric density also provide insight into bone adaptations to exercise. Noresearch has yet investigated the effect of exercising in water on bone geometryin women with osteoporotic VFs.
Purpose: The overall purpose of this thesis was to investigate methods ofmeasuring bone geometry in women with osteoporotic VFs that could be used ina future clinical trial to determine the effect of water exercises on bone in womenwith VFs. The first objective was to determine the relative and absolute intra-raterreliability of VHs in postmenopausal women with and without VFs. In the secondstudy, protocols for recruitment of postmenopausal women with osteoporotic VFsfor a 6 month water exercise intervention and protocols for assessing proposedoutcome measures were piloted to determine the recruitment, adherence to theintervention, adherence to the assessment protocol, safety of the intervention andassessment and retention.
Methods: To address the first objective, DXA VFA scans were acquired for 32women [mean (SD) age 70(7)] and analyzed on 2 occasions, 4 weeks apart, by asingle rater using a predetermined protocol. Semi-automated software derivedmeasures of anterior, middle, and posterior VH. Intra-rater relative reliability wasestimated using the intraclass correlation coefficient (ICC) with 95% confidenceintervals (95% CI). Absolute reliability was estimated using standard error ofmeasurement (SEM) with 95% CI. To address the second objective, women 60years and older with one or more VF were recruited through two osteoporosisclinics and poster advertisements over two months. Feasibility of recruitment wassummarized using the CONSORT flow diagram. Adherence to the six monthcommunity-based aquatic exercise program (74 sessions) was evaluated bypercentage of sessions attended. Adherence to the assessment protocol wasevaluated based on the number of data points lost. The safety was assessedbased on the occurrence of adverse events that were documented as major andminor. Retention was assessed as number of participants returning for follow-up.
Results: DXA-based VH intra-rater reliability could be estimated from T9 to L4,with reduced visibility from T4 to T8. The ICCs were > 0.80 and the SEM wasless than 1.17 mm for all VH except for the posterior aspect of T9 (ICC = 0.62(0.15, 0.84), SEM = 0.92 mm). For the feasibility study, 10 participants wererecruited in 10 weeks by expanding the inclusion criteria. The average adherenceto the intervention was 68%. No measures of VH were obtained for twoparticipants. Movement during acquisition and unanticipated protocol changesresulted in loss of 46% of pQCT scans and 10% of physical performancemeasurement data. There was one major adverse event (fracture). Retention was100% at 6 months and 89% at 12 months.Conclusion: Intra-rater reliability was acceptable for VH between T10 and L4.Further study is needed to assess other measurement properties of DXA-basedVH measures and to identify methods for assessing more proximal vertebrallevels. Further study is needed to determine feasible protocols for recruitmentand assessment of outcome measures. Screening tests for falls risk and protocolfor implementing suitable safety precautions are recommended.