PEDIATRICS Vol. 120 No. 4 October 2007, pp. 931-932 (doi:10.1542/peds.2007-2332)
LETTER TO THE EDITOR |
Cardiovascular Effects of Sibutramine: In Reply
Stephen R. Daniels, MD, PhDDepartment of Pediatrics,
Children's Hospital,
Denver, CO 80218
We appreciate the comments by Dr Pisarik, because they illustrate the reason for the more in-depth analysis of blood pressure (BP) from our study published in Pediatrics.1 It is important to recognize that different statistical conclusions can result when different end points (ie, evaluation points) and statistical models are considered. The most important question, however, is how best to characterize the counterbalancing effects of sibutramine and weight loss on BP. In this study, the overall effect of sibutramine resulted in weight loss without meaningful adverse effects on BP.
In our previous Annals of Internal Medicine article,2 a linear mixed-effects model for repeated measures (up to 17 visits per patient) that included baseline value as a covariate was used on change-from-baseline data to test for overall differences between treatment groups and for treatment differences at each visit. Small but statistically significant overall mean differences in systolic BP (SBP) (–0.9 mmHg [for sibutramine] vs –1.9 mmHg [for placebo]; P = .03) and diastolic BP (DBP) (+0.3 [for sibutramine] vs –1.4 mmHg [for placebo]; P < .001) were reported. In addition, statistically significant differences between treatments were seen at most (10 of 17) visits to month 12 for DBP; however, for SBP, such differences were seen at only 3 of the visits.
In our article, an analysis-of-covariance model with a factor for treatment group and baseline value as a covariate was used to compare treatment groups with respect to changes from baseline to the end point (month 12, or last postbaseline value for noncompleters) in BP.1 The difference between treatment groups in this protocol-specified analysis was not statistically significant for either SBP (mean change: –2.1 mmHg for both sibutramine and placebo) or DBP (–0.1 mmHg [for sibutramine] vs –1.1 mmHg [for placebo]).
Perhaps the best overview of BP changes was provided by the observed data from Figs 2 and 3 in our article,1 which showed that mean SBP generally decreased over time in both treatment groups. Early in treatment, mean DBP increased slightly in the sibutramine group but, by month 6.5, had returned to baseline and stabilized, whereas mean DBP in the placebo group declined early in treatment, and although there was substantial variability, it remained below baseline. Although not shown in our figures, the observed changes from baseline to the end point (unadjusted for baseline value) for sibutramine and placebo, respectively, were –2.1 and –2.0 mmHg for SBP and 0.0 and –1.3 mmHg for DBP.
Analyses of 3 BMI-response subgroups in this study were performed in an effort to better understand the effects of weight loss on sibutramine-associated BP changes in obese adolescents. The subgroup (n = 137 [sibutramine] and 104 [placebo]) with the least reduction in BMI (<5% at the end point) showed no change in mean SBP and an increase of 1.8 mmHg in mean DBP in the sibutramine arm. The placebo-subtracted treatment differences in this subgroup (1.8 mmHg [SBP] and 2.3 mmHg [DBP]), which favored placebo, were both statistically significant. In the 
5% BMI-reduction subgroup, which included approximately two thirds of the sibutramine-treated patients but only approximately one fifth of the placebo-treated patients, the mean SBP/DBP changes of –3.2/–1.2 mmHg in the sibutramine group were not statistically different from placebo; however, we agree with Pisarik that there was insufficient power to test for treatment differences, because the placebo group was too small. Further restricting the BMI-response group, approximately one third of those in the sibutramine group achieved 10% BMI reduction, with greater mean changes of –4.5/–1.9 mmHg for SBP/DBP. Although it is possible that a study adequately powered to examine treatment differences at 5% BMI reduction might show statistically significant treatment differences in BP, such an observation does not address the more meaningful clinical point of interest when considering the effects of sibutramine in obese adolescents. The pattern of BP changes in the sibutramine group across the spectrum of BMI reduction shows that as weight loss increased, the effect of sibutramine on BP was overridden by weight loss-induced BP reduction.
Patients who were taking sibutramine and who at month 6 had not lost >10% of their initial BMI were increased from 10 to 15 mg per day; none were withdrawn from sibutramine and treated with placebo. Analogously, placebo-treated patients were up-titrated to a "higher dose" of placebo at month 6 if they had not lost >10% of their initial BMI.
The important issue is the conclusions that may be drawn from the data. It seems that sibutramine may transiently increase mean BP in obese adolescents. On the other hand, the potential impact of sibutramine on BP seems to be counteracted by weight loss, especially weight loss of 5%. It is also important to reemphasize that even with the potential adverse effect of sibutramine on BP, average BP declined in both the sibutramine and placebo groups, making this an overall advantageous effect.
Finally, it should be noted that adolescents with abnormal BP at baseline were not included in this study. Thus, it would be inappropriate to extrapolate the reported results to adolescents with BP elevation.
REFERENCES
- Daniels SR, Long B, Crow S, et al. Cardiovascular effects of sibutramine in the treatment of obese adolescents: results of a randomized, double-blind, placebo-controlled study. Pediatrics. 2007;120(1) . Available at: www.pediatrics.org/cgi/content/full/120/1/e147
- Berkowitz RI, Fujioka K, Daniels SR, et al. Effects of sibutramine treatment in obese adolescents: a randomized trial.
Ann Intern Med. 2006;145
:81
–90
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
