Summary

The following analysis compares the estimated sample sizes necessary to detect change on various outcomes.

The first tab (dataset) provides descriptions of the data.

Dataset

Full Dataset description

Overall
n 65
days_between_visit (mean (SD)) 286.34 (82.04)
age_at_visit (mean (SD)) 68.20 (7.42)
education (mean (SD)) 15.69 (2.56)
sex_at_birth = M (%) 37 (56.9)
ftldcdr_global (%)
0 1 ( 1.5)
0.5 21 (32.3)
1 28 (43.1)
2 13 (20.0)
3 2 ( 3.1)
genetic_status (%)
MAPT 2 ( 3.1)
NONE-Research genetic testing completed and no mutation identified 48 (73.8)
OTHER-Variant of unknown significance 3 ( 4.6)
Pending results 12 (18.5)
moca_total (mean (SD)) 21.32 (4.67)
psprs_totalcln (mean (SD)) 29.64 (12.01)
ftldcdr_sob (mean (SD)) 5.25 (3.88)
uds3_ef (mean (SD)) -2.03 (0.80)
flu_animal (mean (SD)) 11.59 (5.06)
flu_veg (mean (SD)) 7.70 (3.49)
flu_fwords (mean (SD)) 7.17 (4.24)
flu_lwords (mean (SD)) 6.71 (3.71)
traila_ratio (mean (SD)) 3.32 (2.41)
trailb_ratio (mean (SD)) 13.60 (14.74)
dig_back_sum (mean (SD)) 4.36 (1.90)

Datset by FTLD CDR Global Score

0 0.5 1 2 3 p test
n 1 21 28 13 2
days_between_visit (mean (SD)) NaN (NA) 310.45 (99.82) 270.92 (69.39) 261.46 (56.47) 407.50 (26.16) 0.038
age_at_visit (mean (SD)) 80.00 (NA) 68.38 (6.89) 68.04 (6.54) 70.15 (6.56) 50.00 (7.07) NA
education (mean (SD)) 18.00 (NA) 15.81 (2.46) 16.07 (2.43) 14.46 (2.90) 16.00 (2.83) NA
sex_at_birth = M (%) 1 (100.0) 11 (52.4) 18 (64.3) 6 (46.2) 1 (50.0) 0.698
genetic_status (%) 0.002
MAPT 0 ( 0.0) 0 ( 0.0) 1 ( 3.6) 0 ( 0.0) 1 (50.0)
NONE-Research genetic testing completed and no mutation identified 1 (100.0) 18 (85.7) 18 (64.3) 11 (84.6) 0 ( 0.0)
OTHER-Variant of unknown significance 0 ( 0.0) 0 ( 0.0) 2 ( 7.1) 0 ( 0.0) 1 (50.0)
Pending results 0 ( 0.0) 3 (14.3) 7 (25.0) 2 (15.4) 0 ( 0.0)
moca_total (mean (SD)) 23.00 (NA) 23.24 (3.46) 21.70 (3.89) 18.20 (5.49) 8.00 (NA) NA
psprs_totalcln (mean (SD)) 22.00 (NA) 23.62 (10.01) 30.21 (10.23) 39.50 (8.98) 50.50 (21.92) NA
ftldcdr_sob (mean (SD)) 0.00 (NA) 2.14 (0.95) 4.86 (1.52) 9.50 (2.28) 18.50 (0.71) NA
uds3_ef (mean (SD)) -1.07 (NA) -1.86 (0.98) -2.03 (0.72) -2.32 (0.61) -2.87 (NA) NA
flu_animal (mean (SD)) 13.00 (NA) 14.33 (5.55) 11.22 (4.17) 8.00 (3.50) 3.00 (NA) NA
flu_veg (mean (SD)) 10.00 (NA) 9.42 (3.55) 7.22 (3.03) 6.11 (2.93) 0.00 (NA) NA
flu_fwords (mean (SD)) 11.00 (NA) 8.63 (3.74) 7.33 (4.33) 4.45 (3.59) 1.00 (NA) NA
flu_lwords (mean (SD)) 9.00 (NA) 7.61 (4.19) 6.67 (3.73) 5.27 (2.80) 5.00 (NA) NA
traila_ratio (mean (SD)) 0.96 (NA) 2.85 (1.42) 3.30 (2.61) 4.63 (3.19) NaN (NA) NA
trailb_ratio (mean (SD)) 3.29 (NA) 12.59 (12.88) 12.94 (13.09) 18.29 (22.03) NaN (NA) NA
dig_back_sum (mean (SD)) 5.00 (NA) 4.42 (2.24) 4.28 (1.67) 4.67 (1.94) 2.00 (NA) NA

Sample Size Estimates

Each analysis below is based on the change over time in the data set described above. A difference score was calculated (Visit 2 score - Visit 1 score). This difference was then annualized: [difference score / days between visits ] * 365. I also created a version of this where the raw scores were converted to Z scores based on baseline distributions. Then that z score was annualized. Therefore we have 3 variables of interest (and SD) based on units:

Raw score units: Difference (Diff SD)

Annualized units: Annualized decline (Ann. SD)

Annualized Z score: Annualized Z-Score Decline (Ann. Z. SD)

I calculated necessary sample size for each measure to detect a hypothetical drug effect that reduced decline by 30%. Power was set at .9 with an alpha of .05. Let me know if you want any other values here.

I completed analyses first in the full sample, then in sub samples that were closer to what would be in the actual trial. The tables below present the estimated sample sizes for:

Raw score units: SampleSize_RawUnits

Annualized units: SampleSize_ann

Annualized Z score: SampleSize_AnnZ

Analyses based on full sample

Variable Sample Size Difference Diff. SD Annualized Decline Ann. SD Annualized Z-score Decline Ann. Z. SD SampleSize_rawunits SampleSize_ann SampleSize_annZ
psprs_totalcln 43 9.302 12.118 12.996 18.728 1.034 1.490 141 172 172
perscare 61 0.525 0.744 0.729 1.070 0.925 1.359 167 179 179
ftldcdr_sob 61 1.869 3.063 2.482 4.396 0.634 1.124 223 260 260
homehobb 61 0.393 0.678 0.519 0.950 0.715 1.309 246 278 278
memory 61 0.213 0.442 0.284 0.640 0.574 1.296 357 422 422
moca_total 49 -1.857 4.021 -2.299 5.402 -0.504 1.183 388 457 457
traila_ratio 42 1.477 3.679 1.969 4.875 1.091 2.700 514 508 508
cdrlang 61 0.254 0.681 0.328 0.973 0.596 1.767 595 727 727
commun 61 0.221 0.595 0.274 0.804 0.397 1.165 599 712 712
flu_animal 47 -1.383 4.163 -1.613 5.602 -0.337 1.172 750 999 999
judgment 61 0.123 0.435 0.173 0.593 0.266 0.913 1034 978 978
flu_veg 49 -0.714 2.630 -0.918 3.764 -0.267 1.093 1123 1393 1393
trailb_ratio 31 2.191 8.352 2.662 11.102 0.193 0.804 1204 1441 1441
flu_fwords 50 -0.880 3.420 -1.121 4.730 -0.264 1.112 1251 1473 1473
orient 61 0.123 0.489 0.170 0.693 0.377 1.539 1308 1377 1377
uds3_ef 51 -0.132 0.551 -0.161 0.754 -0.204 0.954 1444 1807 1807
dig_back_sum 47 -0.404 1.861 -0.601 2.534 -0.302 1.275 1755 1474 1474
flu_lwords 49 -0.612 2.971 -0.550 4.649 -0.145 1.222 1950 5917 5917
comport 61 0.016 0.619 0.006 0.993 0.008 1.422 118019 2345235 2345235

Analyses based on FTLD CDR Global <= 1 sample

Variable Sample Size Difference Diff. SD Annualized Decline Ann. SD Annualized Z-score Decline Ann. Z. SD SampleSize_rawunits SampleSize_ann SampleSize_annZ
psprs_totalcln 36 9.889 12.547 13.922 19.248 1.275 1.763 133 158 158
ftldcdr_sob 46 2.120 3.072 2.809 4.361 1.489 2.311 174 200 200
perscare 46 0.543 0.808 0.742 1.153 1.766 2.746 183 200 200
homehobb 46 0.446 0.677 0.574 0.965 1.551 2.610 191 235 235
memory 46 0.217 0.360 0.291 0.550 0.849 1.601 227 295 295
commun 46 0.293 0.602 0.380 0.807 1.030 2.188 348 374 374
trailb_ratio 27 3.273 7.056 4.315 8.768 0.564 1.147 385 342 342
moca_total 39 -1.974 4.264 -2.417 5.732 -0.738 1.750 386 466 466
cdrlang 46 0.293 0.663 0.372 0.879 0.876 2.071 423 463 463
judgment 46 0.174 0.411 0.242 0.563 0.546 1.271 463 449 449
traila_ratio 37 1.413 3.577 1.719 4.121 0.942 2.258 531 476 476
flu_animal 39 -1.410 4.399 -1.508 5.828 -0.324 1.250 806 1236 1236
flu_fwords 41 -1.098 3.639 -1.425 5.029 -0.348 1.228 910 1032 1032
flu_veg 41 -0.707 2.695 -0.878 3.688 -0.258 1.082 1202 1461 1461
uds3_ef 41 -0.129 0.582 -0.153 0.806 -0.187 0.981 1677 2284 2284
orient 46 0.098 0.442 0.151 0.670 0.603 2.672 1694 1627 1627
flu_lwords 40 -0.675 3.214 -0.482 5.004 -0.123 1.278 1877 8930 8930
dig_back_sum 39 -0.231 1.898 -0.350 2.528 -0.176 1.267 5602 4312 4312
comport 46 0.054 0.529 0.058 0.852 0.125 1.830 7858 17633 17633

Analysis based on FTLD CDR Global <=1 sampple; Baseline PSPRS < 25

Variable Sample Size Difference Diff. SD Annualized Decline Ann. SD Annualized Z-score Decline Ann. Z. SD SampleSize_rawunits SampleSize_ann SampleSize_annZ
psprs_totalcln 14 11.357 10.846 15.271 16.048 3.531 3.710 14 17 17
ftldcdr_sob 18 1.472 1.898 1.696 2.058 1.044 1.266 25 23 23
memory 18 0.167 0.243 0.207 0.318 0.697 1.071 32 36 36
homehobb 18 0.306 0.518 0.341 0.740 0.867 1.883 44 72 72
perscare 18 0.333 0.594 0.413 0.740 1.279 2.290 48 49 49
judgment 18 0.194 0.389 0.230 0.455 0.601 1.186 61 59 59
flu_animal 15 -2.067 4.728 -1.685 6.066 -0.330 1.187 80 197 197
commun 18 0.222 0.521 0.276 0.623 0.754 1.702 84 78 78
cdrlang 18 0.167 0.454 0.127 0.486 0.334 1.281 113 224 224
traila_ratio 15 0.525 1.497 0.526 1.614 0.380 1.167 124 143 143
trailb_ratio 11 1.683 7.251 2.078 7.712 0.654 2.427 282 210 210
flu_lwords 16 0.688 3.497 1.574 5.479 0.344 1.198 394 184 184
comport 18 0.056 0.379 0.098 0.527 0.251 1.355 709 443 443
orient 18 0.028 0.208 0.004 0.273 0.018 1.183 854 65393 65393
flu_veg 16 -0.312 2.575 -0.237 3.477 -0.072 1.059 1032 3262 3262
dig_back_sum 16 0.188 1.870 0.359 2.197 0.155 0.952 1512 571 571
moca_total 14 -0.214 3.926 -0.553 5.275 -0.136 1.295 5105 1386 1386
flu_fwords 16 -0.188 4.262 -0.086 5.459 -0.019 1.217 7857 61810 61810
uds3_ef 16 -0.023 0.752 0.064 0.957 0.062 0.935 16168 3454 3454