Skip to main content
. 2015 Mar 13;77(4):684–696. doi: 10.1002/ana.24370

Table 3.

KM and Cox Regression Models for LoA: Median Age at LoA, Log-Rank p-Values, Cox Hazard Ratios, and p-Values by SPP1-LTBP4 Genotypes and GC Treatment

Cohort All Participants
GC Treated
GC Untreated
No. Median Age at LoA, yr p, KM Log-Rank HR (95% CI), Cox p No. Median Age at LoA, yr p, KM Log-Rank HR (95% CI), Cox p No. Median Age at LoA, yr p, KM Log-Rank
Whole genotyped DNHS cohort, n = 283a
SPP1 rs28357094 0.048b 1.22 (0.89–1.68), 0.22 0.032b 1.61 (1.09-2.37), 0.016b 0.6
TT 195 13.0 150 13.9 45 10.0
TG/GG 84 11.8 63 12.0 21 10.0
LTBP4 rs10880 0.20 0.78 (0.49–1.24), 0.29 0.27 0.74 (0.44-1.26), 0.27
CC/CT 242 12.0 178 13.3 64 10
TT 32 13.9 27 13.9 5 9.1
Caucasian cohort, controlled for population stratification by MDS, n = 118a
SPP1 rs28357094 0.047b 1.54 (0.93–2.54), 0.09 0.07 1.85 (1.01–3.38), 0.047b 0.7
TT 81 13.9 67 13.9 14 10.0
TG/GG 35 12.0 26 12.0 9 9.0
LTBP4 rs10880 0.024b 0.49 (0.23–1.07), 0.07 0.046b 0.47 (0.20–1.09), 0.08
CC/CT 103 12.6 80 13.8 23 10
TT 12 15.0 11 16.0 1 N/A
a

Total number may not correspond exactly to genotype group number because of a few ungenotyped patients (limited DNA availability).

b

Statistically significant effect of genotype on LoA.

CI = confidence interval; DNHS = Duchenne Natural History Study; GC = glucocorticoid corticosteroid; GC treated = GC treatment at least 1 year while patient was ambulatory; GC untreated = no or <1 year of GC treatment while patient was ambulatory; HR = hazard ratio for genotype in Cox regression model with GC treatment as a time-varying covariate; KM = Kaplan–Meier survival analysis with log-rank comparison of median age at LoA; LoA = loss of ambulation; MDS = multidimensional scaling; N/A = not available.