Joost Peerbooms

109 PRP in plantar fasciitis for handling missing data on the outcome measures by means of maximum likelihood estimation in the mixed model analysis, as this method assumes the missing values to be either missing at random or missing completely at random. Table 1 Baseline Characteristics for Patients in the PRP and Control Groups α PRP group (n=63) Control group (n=52) Baseline characteristics M (SD) / n (%) M (SD) / n (%) Female sex 48 (76.2) 34 (65.4) Age, y 50.73 ± 11.33 47.5 ± 11.19 Length, cm 170.33 ± 10.16 174.65 ± 10.11 Weight, kg 84.27 ± 15.62 91.92 ± 19.39 Body Mass Index 29.12 ± 5.17 30.16 ± 6.29 Duration of Symptoms, weeks b 70 (40-130) 52 (35- 90) Smoking, yes 9 (14.3) 10 (19.2) Previous Foot Surgery: yes 2 (3.2) 3 (5.8) Codisease: yes 15 (23.8) 11 (21.2) Comedication: yes 20 (31.7) 13 (25.0) Baseline scores of outcome measures M (SD) M (SD) FFI Pain 64.7 ± 16.95 56.96 ± 20.13 FFI Disability 51.56 ± 20.6 41.15 ± 21.71 FFI Activity 31.64 ± 16.64 25.2 ± 16.57 AOFAS 49.44 ± 15.29 57.71 ± 16.61 WHOQOL-BREF Overall QOL and General Health 7.17 ± 1.55 7.51 ± 1.29 Physical Health 12.26 ± 2.74 12.99 ± 2.58 Psychological Health 14.98 ± 2.29 15.37 ± 2.44 Social Relationships 15.44 ± 2.79 16.42 ± 1.89 Environment 16.2 ± 4.32 16.1 ± 2 α Data are provided as n (%) or mean ± SD. AOFAS, American Orthopaedic Foot & Ankle Society; FFI, Foot Function Index; PRP, platelet-rich plasma; WHOQOL-BREF, World Health Organization Quality of Life. b Median (interquartile range) reported because this variable is not normally distributed. Table 2 presents the results of the linear mixed modelling analysis for all outcome measures. For each outcome, the treatment × time interaction effects in the second column pertain to a model including only a linear time effect, while the similar tests in the third column are derived from a model including both linear as well as quadratic 7