Table 3—

Effects of physical activity counseling on costs for medical prescriptions, other direct health system costs, direct and indirect social costs, and total costs in 179 type 2 diabetic subjects with different amounts of energy expenditure (METs per hour per week)

Drugs
Other health care
Direct social
Indirect social
Total costs
BasalChangeBasalChangeBasalChangeBasalChangeBasalChange
Group 0 (n = 28) A1,173 ± 44393 (111 ± 675)*1,456 ± 77123 (−25 ± 271)14 ± 1.010 (−6 ± 26)1,444 ± 90302 (51 ± 553)*4,253 ± 289828 (224 ± 1,432)*
Group 1–10 (n = 28) B1,207 ± 52206 (−34 ± 446)1,541 ± 67−7 (−15 ± 1)12 ± 0.850* (9 ± 91)1,356 ± 105−24 (−71 ± 23)4,116 ± 311224 (−29 ± 477)
Group 11–20 (n = 28) C1,087 ± 43−196 (−342 ± −50)*1,384 ± 86−92 (−191 ± 7)15 ± 1.2152 (41 ± 263)*1,281 ± 68−250 (−438 ± −62)*3,767 ± 235−386 (−663 ± −109)*
Group 21–30 (n = 28) D1,085 ± 39−593 (−1,025 ± −161)*1,414 ± 59−296 (−518 ± −74)*11 ± 0.7199 (48 ± 350)*1,491 ± 93−762 (−1,394 ± −130)*4,001 ± 280−1,452 (−2,364 ± −540)*
Group 31–40 (n = 28) E1,177 ± 56−660 (−1,078 ± −242)*1,543 ± 73−588 (−1,011 ± −165)*16 ± 1.2256 (88 ± 424)*1,434 ± 73−909 (−1,551 ± −267)*4,169 ± 228−1,902 (−3,217 ± −587)*
Group >40 (n = 28) F1,087 ± 45−579 (−905 ± −253)*1,645 ± 84−794 (−1,241 ± −347)*14 ± 0.9300 (83 ± 517)*1,468 ± 80−997 (−1,601 ± −393)*4,214 ± 247−2,070 (−3,325 ± −815)*
All 179 subjects1,134 ± 55−259 (−422 ± −96)*1,528 ± 73−298 (−521 ± −75)*14 ± 1.3168 (41 ± −269)*1,412 ± 78−466 (−809 ± −123)*4,087 ± 235−855 (−1,520 ± −190)*
50% simulation1,177 ± 63−101 (−180 ± −6)*1,580 ± 75−207 (−362 ± −52)*14 ± 1133 (30 ± 236)*1,440 ± 82−303 (−539 ± −67)*4,210 ± 266−478 (−817 ± −139)*
33% simulation1,175 ± 7177 (−41 ± 195)1,571 ± 79−91 (−200 ± 18)13 ± 1.187 (9 ± 165)*1,411 ± 82−106 (−191 ± −21)*4,170 ± 246−34 (−80 ± 12)
Between-group comparisons P<0.001<0.001<0.001<0.001<0.001
Post hoc analysisA vs. CDEF, B vs. CDEF, C vs. EABC vs. DEF, D vs. FAB vs. F, A vs. EAB vs. DEF, C vs. EFAB vs. DEF, C vs. EF
  • Data are means ± SE and means (95% CI). Sensitive analysis assumes that instead of 125 of 179 (69%) only 90 of 179 (50%) or 60 of 179 (33%) diabetic subjects achieved the target of the intervention (>10 METs · h−1 · week−1).

  • *

    * P < 0.05 vs. basal, costs in U.S. dollars per capita per year. Between-group comparisons are reported in the last two rows. Significant differences are then followed by post hoc results (e.g., A vs. CDEF means that group A is different from groups C–F; CDEF are not different from each other).