Table 3.

Subgroup trends in estimated treatment effects for left ventricular end diastolic volume (percent difference in geometric mean changes)

Subgroup FactorDaily TrialNocturnal Triala
SubgroupEstimated Effect with 95% CIbP Value InteractionSubgroupEstimated Effect with 95% CIb
Age (yr)≤50−14.67 (−21.59 to −7.13)0.97≤50−14.29 (−29.82 to 4.68)
>50−7.22 (−14.85 to 1.11)>50−0.82 (−15.60 to 16.56)
SexMen−9.51 (−15.97 to −2.56)0.41Men−5.86 (−19.41 to 9.97)
Women−14.12 (−22.15 to −5.25)Women−3.82 (−22.52 to 19.38)
Diabetic statusNondiabetes−11.44 (−18.05 to −4.3)0.82Nondiabetes−3.83 (−18.36 to 13.29)
Diabetes−10.18 (−18.18 to −1.39)Diabetes−7.82 (−23.51 to 11.10)
RaceWhite−10.78 (−18.46 to −2.38)0.90White−5.43 (−18.91 to 10.30)
Black−11.5 (−19.12 to −3.15)Black−5.77 (−25.23 to 18.77)
Anthropometric volume (L)≤35−16.16 (−24.93 to −6.38)0.55≤353.36 (−19.29 to 32.35)
>35−8.98 (−15.09 to −2.44)>35−6.89 (−19.36 to 7.52)
Anthropometric volume (L)≤40−13.04 (−19.53 to −6.02)0.23≤407.39 (−9.98 to 28.12)
>40−9.52 (−17.53 to −0.73)>40−16.98 (−29.89 to −1.69)
Vintage (yr)<4−8.2 (−15.17 to −0.66)0.35<21.23 (−13.23 to 18.10)
≥4−14.27 (−22.03 to −5.73)≥2−17.02 (−32.98 to 2.75)
Urine volume (ml)≤100−14.2 (−20.15 to −7.81)0.02≤500−11.52 (−25.48 to 5.04)
>100−3.25 (−13.01 to 7.6)>5005.08 (−12.45 to 26.13)
CHFNo CHF−10.67 (−16.43 to −4.52)0.84No CHF−11.34 (−21.79 to 0.51)
CHF−12.14 (−23.64 to 1.1)CHF
Baseline LVM (g)<132−9.28 (−16.42 to −1.53)0.08<132−6.05 (−21.29 to 12.14)
≥132−12.32 (−19.39 to −4.64)≥132−0.16 (−16.01 to 18.68)
β-Blockers useNo−10.15 (−18.27 to −1.23)0.88No−12.41 (−28.58 to 7.41)
Yes−11.46 (−18.07 to −4.32)Yes−2.18 (−16.98 to 15.26)
ACEI or ARB useNo−9.12 (−15.04 to −2.79)0.12No−5.13 (−17.37 to 8.93)
Yes−17.17 (−27.05 to −5.95)Yes−8.65 (−40.64 to 40.59)
  • 95% CI, 95% confidence interval; CHF, congestive heart failure; LVM, left ventricular mass; ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker.

  • a Because of its limited sample size, we considered subgroup analyses in the nocturnal trial in an exploratory fashion without inference testing.

  • b Treatment effects expressed as percent difference based on log-transformed analysis.