JevalentThis paper would tend to support your conclusion the...

Jevalent
This paper would tend to support your conclusion the clinical trial was flawed in placebo design and diet and exercise masking the result.
Also as I surmised a trial should run for a longer period of time.
http://onlinelibrary.wiley.com/doi/10.1002/j.1550-8528.1995.tb00209.x/pdf

Note that AOD9604 complies favourably with nearly all the desirable features of a weight loss drug with only dosage uncertainty ie
. Characteristics of an ideal weight loss drug
1. Orally Active
2. Few or no side effects
3. Dose-dependent reduction in body fat
4. Reduces visceral fat
5. Inexpensive
6. Long-acting
7. Non-toxic

As noted in the paper

The effect of placebo treatment in a number of clinical
trials for anti-obesity drugs that lasted fewer than 30 weeks
are shown in Figure 2 and those lasting 31 to 60 Table 1

Figure 3. The magnitude of weight loss in the placebo-treated
patients varied substantially from one trial to the next. Be-
cause of this variation, the effects of drug treatments may be
difficult to see, if the magnitude of placebo effect is great.
With both placebos and active drugs, the initial rate of weight
loss slows. In several long-term trials lasting up to 3 years
using dietary approaches to treat hypertension (Figure 4), it
can be seen that weight loss in all cases tends to return to or
exceed baseline.
Behaviour modification was originally introduced for the
treatment of obesity in 1968 by Stuart (42) and has become an
important part of most treatment programs. When vigorous
behavioural treatment programs are introduced, they may
mask the effect of drug treatment (43). In amulti-center trial
shown in Figure 5, one centre used a behavioural program for
all patients, including the placebo and drug treatment groups
and a group with no pills (13,46). When this was compared
With the centres in which no behavioural treatment was used,
It is clear that the behaviour treatment group in one centre did
as well as the drug treatments for the other four centres
combined. For these reasons, the use of vigorous behavioural
Body Weight (kgs)
treatment programs as a placebo for comparison with an
appetite suppressant may obscure the effect of the drug and
be inappropriate. Use of very-low-calorie diet as a placebo
treatment is also inappropriate. In at least two instances
(3, 30), the use of very-low-calorie diet obscured the effect of
the drug. Thus, when designing trials for anti-obesity drugs,
use of aggressive behavioural modification techniques and/or
Very-low-calorie diets should be avoided.
For an anti-obesity drug, weight loss should be the
primary end-point. Since central fat distribution is as impor-
tant, if not more important in some people than total fatness,
central fat might be a subsidiary primary end-point. A
demonstration that visceral fat was significantly reduced
even if total fat was not reduced might be perfectly appro-
priate basis for approving a drug for clinical use.
A growing body of information suggests that a loss of
5% to 10% of body weight will provide significant benefit
from weight loss, and that a goal of “desirable body weight”
is inappropriate for most overweight individuals. It may thus
be appropriate to examine the per cent of drug-treated and
placebo-treated patients who lose varying amounts of weight.
This categorical approach was used by Scoville (39) in his
analysis of the early submissions of appetite suppressant
drug application to the FDA. He found that 44% of drug-
treated subjects lost more than 1 lb week (0.45 kg week) vs.
26% in placebo-treated subjects.