Maintained intentional weight loss reduces cardiovascular outcomes: results from the Sibutramine Cardiovascular OUTcomes (SCOUT) trial.
ABSTRACT The Sibutramine Cardiovascular OUTcomes trial showed that sibutramine produced greater mean weight loss than placebo but increased cardiovascular morbidity but not mortality. The relationship between 12-month weight loss and subsequent cardiovascular outcomes is explored.
Overweight/obese subjects (N = 10 744), ≥55 years with cardiovascular disease and/or type 2 diabetes mellitus, received sibutramine plus weight management during a 6-week Lead-in Period before randomization to continue sibutramine (N = 4906) or to receive placebo (N = 4898). The primary endpoint was the time from randomization to first occurrence of a primary outcome event (non-fatal myocardial infarction, non-fatal stroke, resuscitated cardiac arrest or cardiovascular death).
For the total population, mean weight change during Lead-in Period (sibutramine) was -2.54 kg. Post-randomization, mean total weight change to Month 12 was -4.18 kg (sibutramine) or -1.87 kg (placebo). Degree of weight loss during Lead-in Period or through Month 12 was associated with a progressive reduction in risk for the total population in primary outcome events and cardiovascular mortality over the 5-year assessment. Although more events occurred in the randomized sibutramine group, on an average, a modest weight loss of approximately 3 kg achieved in the Lead-in Period appeared to offset this increased event rate. Moderate weight loss (3-10 kg) reduced cardiovascular deaths in those with severe, moderate or mild cardiovascular disease.
Modest weight loss over short-term (6 weeks) and longer-term (6-12 months) periods is associated with reduction in subsequent cardiovascular mortality for the following 4-5 years even in those with pre-existing cardiovascular disease. While the sibutramine group experienced more primary outcome events than the placebo group, greater weight loss reduced overall risk of these occurring in both groups.
- [show abstract] [hide abstract]
ABSTRACT: Obesity increases the risk of cardiovascular disease and premature death. Adipose tissue releases a large number of bioactive mediators that influence not only body weight homeostasis but also insulin resistance - the core feature of type 2 diabetes - as well as alterations in lipids, blood pressure, coagulation, fibrinolysis and inflammation, leading to endothelial dysfunction and atherosclerosis. We are now beginning to understand the underlying mechanisms as well as the ways in which smoking and dyslipidaemia increase, and physical activity attenuates, the adverse effects of obesity on cardiovascular health.Nature 01/2007; 444(7121):875-80. · 38.60 Impact Factor
Maintained Intentional Weight Loss Reduces Cardiovascular Outcomes:
Results from the Sibutramine Cardiovascular Outcomes (SCOUT) Trial.
Ian D Catersona, Nick Finerb, Walmir Coutinhoc, Luc F Van Gaald, Aldo P Maggionie,
Christian Torp-Pedersenf, Arya M Sharmag, Udo F Leglerh, Gillian M Shepherdh,
Richard A Rodeh, Renee J Perdoki, Cheryl L Renzh, W Philip T Jamesj on the behalf of
the SCOUT Investigators
aBoden Institute of Obesity Nutrition, Exercise & Eating Disorders, University of Sydney,
NSW, Australia; bUniversity College London Vascular Physiology Unit, London, UK;
cCatholic University of Rio de Janeiro, Brazil; dAntwerp University Hospital, Belgium;
eANMCO Research Center, Florence, Italy; fDepartment of Cardiology, Gentofte University
Hospital, Hellerup, Denmark; g University of Alberta, Royal Alexandra Hospital, Edmonton,
Alberta, Canada; hAbbott Laboratories, Abbott Park, Il, USA; iPreviously Abbott
Laboratories, Abbott Park, Il, USA, currently: Advanced Clinical, Deerfield, Il, USA;
jLondon School of Hygiene and Tropical Medicine, London, UK
Ian D Caterson, Boden Institute of Obesity Nutrition, Exercise & Eating Disorders,
University of Sydney, NSW, Australia
Tel: +61 2 9036 3105
Fax: +61 2 9036 3176
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which
may lead to differences between this version and the Version of Record. Please cite this
article as an ‘Accepted Article’, doi: 10.1111/j.1463-1326.2011.01554.x
Diabetes, Obesity and Metabolism
© 2011 Blackwell Publishing Ltd
The Sibutramine Cardiovascular OUTcomes trial demonstrated that sibutramine produced
greater mean weight loss than placebo but increased cardiovascular morbidity but not
mortality. The relationship between 12-month weight loss and subsequent cardiovascular
outcomes is explored.
Materials and Methods
Overweight/obese subjects (N=10,744), ≥55years with cardiovascular disease and/or type 2
diabetes mellitus, received sibutramine plus weight management during a 6-week Lead-in
Period before randomisation to continue sibutramine (N=4906) or to receive placebo
(N=4898). The primary endpoint was the time from randomisation to first occurrence of a
primary outcome event (nonfatal myocardial infarction, nonfatal stroke, resuscitated cardiac
arrest or cardiovascular death).
For the total population, mean weight change during Lead-in Period (sibutramine) was -
2.54kg. Post-randomisation, mean total weight change to Month 12 was -4.18kg
(sibutramine) or -1.87kg (placebo). Degree of weight loss during Lead-in Period or through
Month 12 was associated with a progressive reduction in risk for the total population in
primary outcome events and cardiovascular mortality over the 5-year assessment. Although
more events occurred in the randomised sibutramine group, on average, a modest weight loss
of approximately 3kg achieved in the Lead-in Period appeared to offset this increased event
rate. Moderate weight loss (3-10kg) reduced cardiovascular deaths in those with severe,
moderate or mild cardiovascular disease.
Modest weight loss over short-term (6 weeks) and longer-term (6-12 months) periods is
associated with reduction in subsequent cardiovascular mortality for the following 4-5years
even in those with pre-existing cardiovascular disease. Whilst the sibutramine group
experienced more primary outcome events than the placebo group, greater weight loss
reduced overall risk of these occurring in both groups.
Keywords: Cardiovascular outcomes, weight loss, sibutramine
Obesity is a known risk factor for cardiovascular disease, the leading cause of reduced life
expectancy and death in the developed world1. No longer confined to affluent countries, its
prevalence is increasing in low and middle-income countries resulting in major public health
problems2. Obesity adversely affects the principal risk factors for cardiovascular disease such
as hypertension and dyslipidaemia3 and also risk factors such as inflammation and insulin
resistance4. Long-term prospective population-based studies have confirmed that the
overweight/obese have excess mortality5-8. While there is some evidence for reduced
mortality after bariatric surgery9,10, non-surgical intentional weight loss has not been shown
to decrease cardiovascular events or mortality11.
The Sibutramine Cardiovascular OUTcomes (SCOUT) trial was the first prospective trial that
attempted to assess the effects of intentional weight loss by lifestyle intervention and
pharmacotherapy on cardiovascular morbidity and mortality in overweight and obese subjects
with pre-existing cardiovascular disease12. In this trial, sibutramine was used in addition to a
weight management programme that included dietary modifications and increased physical
activity. All subjects received sibutramine 10 mg daily during an initial 6-week Lead-in
Period. SCOUT demonstrated that sibutramine improved weight loss long-term but was
associated with increased risk for cardiovascular morbidity (but not mortality). The primary
analyses did not consider whether weight change (loss or gain) affected these variables, and it
must be emphasised that in SCOUT, once randomised, subjects were not required to
discontinue sibutramine (or placebo) treatment if they failed to either lose weight or if they
regained previously lost weight.
This paper explores the relationship between weight change during the first year of treatment
and its effect on cardiovascular outcomes in the overall population and in the two randomised
treatment groups; i.e., those who received sibutramine during the initial 6-week Lead-in
Period and were then randomised to placebo (placebo group) and those who were randomised
to continue on sibutramine (sibutramine group).
Materials and Methods
SCOUT was a randomised, double-blind, placebo-controlled trial conducted in 13 European
countries, Brazil, Mexico and Australia. The first subject entered the 6-week Lead-in Period
in January 2003; recruitment continued for 3 years with the final visit in March 2009. The
protocol has been described elsewhere12. Essentially, eligible subjects included men and
women aged ≥ 55 years, with a body mass index (BMI) of ≥ 27 kg/m2 and ≤ 45 kg/m2, or ≥
25 kg/m2 and < 27 kg/m2 with a large waist circumference (≥ 102/≥88 cm; men/women
respectively). Enrolled subjects had history of cardiovascular disease (defined as coronary
artery disease, stroke, or peripheral arterial occlusive disease) and/or type 2 diabetes mellitus
with at least one additional cardiovascular risk factor [i.e., hypertension, dyslipidaemia,
current smoking, or diabetic nephropathy]. Exclusion criteria included heart failure symptoms
greater than New York Heart Association Functional Class II, blood pressure >160/100
mmHg, pulse >100 bpm, scheduled cardiac surgery or coronary angioplasty, or recent weight
As a safety measure (since similar subjects had not been exposed to sibutramine previously),
all participants were enrolled in a 6-week, single-blind, Lead-in Period and received
sibutramine 10 mg together with a cardioprotective lifestyle programme. This approach was
intended to identify and withdraw, prior to randomisation, subjects with early and persistent
increases in blood pressure and/or pulse.
Following the sibutramine Lead-in Period, eligible subjects continued their individualised
lifestyle programme and were randomised to sibutramine 10 mg (titration to 15 mg was
permitted) or placebo. Investigators were required to ensure that subjects received optimum
monitoring and management of all medical conditions (e.g., hypertension, dyslipidaemia,
diabetes) according to individual national guidelines. Anthropomorphic and vital sign
measurements were performed monthly for the first three months of the Randomisation
Period and then every three months thereafter for subjects on study medication, or annually
after subjects had discontinued study medication. All subjects had quarterly assessments for
morbidity and mortality status with annual laboratory evaluations and electrocardiograms.
The primary endpoint was the time from randomisation to the first occurrence of a primary
outcome event (i.e., nonfatal myocardial infarction, nonfatal stroke, resuscitated cardiac
arrest or cardiovascular death). All outcome events, including causes of death, were evaluated
by an independent adjudication committee which used pre-specified criteria.
The study (ClinicalTrial.gov number: NCT00234832) was conducted under International
Conference on Harmonisation/Good Clinical Practice guidelines with approval from relevant
ethical committees. All subjects gave written informed consent.
Power and sample size considerations for the SCOUT trial have been published12.
Demographic and baseline characteristics at Lead-in Period baseline were summarised for the
intention-to-treat (ITT) population and each treatment group and have been reported
For this analysis cardiovascular disease severity for individual subjects was defined post-hoc
using the following criteria:
• Severe: history of myocardial infarction, stroke, congestive heart failure, peripheral
vascular disease, or revascularisation with either percutaneous or surgical intervention
(defined as: percutaneous transluminal coronary angioplasty; coronary artery bypass
grafting; coronary artery stent placement; cardiac transplant; peripheral vascular
bypass; peripheral angioplasty; and carotid endarterectomy);
• Moderate: history of angina, coronary artery disease, arrhythmia, transient ischaemic
attack, or valvular heart disease with no history of severe cardiovascular disease; and
• Mild: none of the above.
Individual subject weight change measurements were computed from the Lead-in Period
baseline (Week -6) to the Randomisation Period baseline, the Month 6 and the Month 12
Time-to-event analyses of the rates of primary outcome events and the rate of cardiovascular
deaths in the ITT population were performed using Cox models, with factors for randomised
treatment (two levels: sibutramine or placebo), cardiovascular disease severity (three levels:
mild, moderate, or severe), and weight loss as potential covariates. Estimates of the 5-year
risk of primary outcome events (or separately for cardiovascular death), hazard ratios, 95%
confidence intervals (95% CI), and log-rank P values were calculated within the Cox model
framework. The additional weight loss necessary to "offset" risk associated with sibutramine
was determined by visually comparing the placebo point estimate for 5-year survivorship at a
given weight loss to an equivalent 5-year survivorship for sibutramine at its corresponding
Primary analyses assessing the effect of weight loss achieved from the Lead-in Period
baseline to the Randomisation Period baseline (6 weeks) and to Month 12 included all
subjects who were still participating in the trial and who had not experienced a primary
outcome event (or cardiovascular death) prior to the time point of interest. Sensitivity
analyses assessing the effect of weight loss through Month 6 and through Month 12 included
all subjects who were randomised and had weight change measurements at the time point of
interest regardless of the time to onset of their primary outcome event (or cardiovascular
death). In addition, secondary analyses of interactions between treatment and weight loss, as
well as the (potential) quadratic effect related to weight loss, were evaluated using Cox
models. No statistically significant treatment-by-weight loss interaction was detected
(P≥0.100), while the quadratic effect of weight loss was associated with cardiovascular death
for Month 6 (P=0.008) and Month 12 (P=0.002) only. All statistical analyses were performed
using SAS®, version 9.2 (SAS Institute, Cary, NC USA). All statistical tests were conducted
at the (two-tail) 0.05 level of significance unless otherwise specified. Body weight was
imputed using a last observation carried forward (LOCF) approach if measurements were not
recorded at Randomisation Period baseline, Month 6 or Month 12.
Subject characteristics including medical history are summarized in Table 1. Overall, 97%
were Caucasian, 58% were male; mean age was 63.2 years and mean BMI was 34.5 kg/m2.
Eighty-four percent had type 2 diabetes mellitus, with 69% and 13% of subjects having
history of severe or moderate cardiovascular disease, respectively. Treatment groups were
well-balanced with respect to the characteristics summarised in Table 1.
Weight change measurements from Lead-in Period baseline through Randomisation Period
baseline, Month 6 and Month 12 are shown in Table 2 for the overall population and each
treatment group. In the sibutramine-treated, 6-week Lead-in Period, overall mean weight
change was -2.54 kg. At Month 6 post-randomisation, subjects who continued on
sibutramine had an additional mean weight change of -2.04 kg which appeared to be
maintained at Month 12 when total mean weight change was -4.18 kg. In contrast, subjects
subsequently randomised to placebo had a mean total weight change of -1.87 kg, thus
indicating a mean net weight gain over the 12-month post-randomisation period. Although
weight loss distributions for the two randomized treatment groups overlapped at Month 12,
more sibutramine subjects were observed to have lost substantial weight when compared to
placebo subjects; for example, 857 subjects randomized to sibutramine lost 4 to 6 kg
compared to 551 subjects randomized to placebo (Supporting Information Figure 1).
Table 2 also summarises the number of primary outcome events and cardiovascular deaths
following randomisation to Month 12, and from Month 12 to the end of the treatment period
for the overall population and by treatment group. For primary outcome events, the rates were
also similar to Month 12 - placebo 2.6% vs sibutramine 2.8%. After Month 12 in the
randomised sibutramine arm only, there was an increase in primary outcome events
compared to placebo (7.7% vs. 9.0%). For cardiovascular mortality, there was no apparent
difference between the groups at any time point: up to Month 12 - placebo: 0.8% vs.
sibutramine: 0.8%; after Month 12 – placebo 3.9% vs. sibutramine 3.8%.
The effects of early weight change on primary outcome events and cardiovascular death up to
the end of treatment (year 5) are shown in Figure 1. In the overall SCOUT population,
greater weight loss achieved in the initial 6-week sibutramine treatment period was associated
with a greater fall in risk for both primary outcome events and cardiovascular death (Figures
1a and 1b); confidence limits were narrow over the 5-year period. When the analysis was
extended to consider the severity of subjects’ cardiovascular disease at study enrolment and
the subsequent impact of sibutramine treatment (Figures 1c and 1d), the modest weight loss
achieved over the short 6-week period appeared to be beneficial in both groups (i.e., subjects
randomised to continue sibutramine treatment and those randomised to placebo) and
applicable to all three severity grades of pre-existing cardiovascular disease (mild, moderate
and severe: Supporting Information Table 1).
When the effects of weight change to Month 12 were considered (Figures 2a and 2b), the
relationship between the amount of weight loss and subsequent primary outcome events again
appeared to be consistent for both treatment arms. Irrespective of randomised treatment
group, subjects who achieved the most weight loss had the greatest reductions in the absolute
risk of primary outcome event, with the greatest benefit in subjects with “severe”
cardiovascular disease (Figure 2a). However, subjects in the sibutramine-treated arm showed
higher rates of primary outcome events in all three categories of cardiovascular disease
severity compared to placebo-treated subjects.
In subjects randomised to placebo following the 6-week sibutramine Lead-in Period and who
achieved and maintained a weight loss of 5 kg at Month 12, there was a reduction in the risk
of primary outcome events from 9.1% (for no weight change) to 8.3%; in comparison, for
similar subjects who continued to receive sibutramine after randomisation, the risk reduction
was 10.9% to 9.9%. Although an additional average weight loss of approximately 3 kg
achieved in the initial 6-week period appeared to offset the subsequent increase in primary
outcome event rates for the subjects randomised to long-term sibutramine (Figure 1c),
greater weight loss (approximately 10 kg) was required at Month 12 in the sibutramine arm to
offset the higher rate of primary outcome events and match those seen in placebo subjects
(Figure 2a). Intermediate results were seen with analyses at Month 6 (data not shown).
There was a U-shaped relationship between cardiovascular mortality and weight change in
subjects with severe pre-existing cardiovascular disease (Figure 2b); increased mortality was
demonstrated in subjects who either gained weight, or lost more than approximately 10 kg.
Similar results were observed irrespective of treatment arm post-randomisation.
Sensitivity analyses performed to assess whether different degrees of weight loss in the Lead-
in Period or to Month 12 in relation to total primary outcome events or to all cardiovascular
mortality from the time of the start of the trial i.e. the Lead-in Period baseline showed similar
relationships to those presented (data not shown).
The initial results from SCOUT12 showed that in subjects at high risk for cardiovascular
outcome events long-term sibutramine treatment plus lifestyle modification, independent of
weight loss, resulted in more non-fatal cardiovascular events compared with those subjects
who after a short 6-week course of sibutramine remained on placebo for the remainder of the
trial. There was no difference in all-cause mortality between the two treatment groups. These
initial results did not consider the potential importance of weight change itself (either loss or
gain) or the initial severity of cardiovascular disease. In some other studies and with other
treatments13-15, but not in randomised controlled trials, it has been suggested that weight loss
may be associated with a reduced overall mortality but without documenting the impact on
cardiovascular morbidity or mortality.
In this analysis, the data show that with modest weight loss after 6 weeks, or after
approximately 12 months, there is a reduced risk of both primary outcome events and
cardiovascular mortality. These findings are novel and substantiate the epidemiological
suggestions that therapeutic or intentional weight loss could reduce cardiovascular disease
risk. Even the modest degree of weight loss achieved during the first 6 weeks of the SCOUT
trial was associated with benefit irrespective of further drug therapy or weight change. This
early weight loss of just over 2 kg reduced the incidence of cardiovascular events and
mortality over the subsequent 5 years - a similar so-called “memory” effect has been shown
in diabetes treatment and prevention trials18-20. This provides strong confirmatory evidence
that interventional weight loss has long term benefit.
Nevertheless, the primary results from SCOUT did show an overall deleterious effect of
treatment with sibutramine in this population with an increase in non-fatal myocardial
infarction and non-fatal stoke events in subjects with a history of cardiovascular disease. The
exact cause(s) of the increased risk was not apparent from the initial analyses12. The data
presented here suggest that continued treatment with sibutramine may reduce primary
outcome events as long as subjects achieve sufficient weight loss.
How then are these findings from SCOUT that appear to contradict the original findings to be
explained? The pre-specified initial analyses focused on the relationship between the two
randomised arms of the trial after both groups had lost weight with the aid of sibutramine
during the 6-week Lead-in Period. The overall impact of weight loss on outcomes was not
considered. The weight loss achieved in the Lead-in Period was accompanied by falls in
blood pressure that would be predicted to improve cardiovascular morbidity and mortality. It
is possible that the observed differences in blood pressure and pulse that occurred with
continued sibutramine treatment may have driven the selective relative increase in non-fatal
events in the randomised sibutramine group, particularly in subjects with a history of
cardiovascular disease; however, additional analyses will be necessary to further explore this
hypothesis. Other possible explanations might relate to the direct serotonergic, rather than
noradrenergic, actions of sibutramine, the potential of bias produced by changes or choices in
medication by the treating physicians, or to a direct effect of extant cardiovascular disease
Clear evidence of a reduction in cardiovascular mortality has only been shown to be
significant and substantial following bariatric surgery9,10. In contrast, there is some evidence
that weight loss may be associated with increased mortality particularly in patients with pre-
existing cardiovascular disease, e.g. heart failure, coronary artery disease or peripheral
arterial occlusive disease22. The data from SCOUT may support these facts as subjects who
entered with the most severe cardiovascular disease and who lost very large amounts of
weight had the worst outcomes in terms of cardiovascular deaths (Figure 2b) even if their
actual rates of myocardial infarctions and stroke declined (Figure 2a). Whether the
relationship between marked weight loss and increased cardiovascular mortality is causal (i.e.
weight loss driving the excess mortality) or an effect of marked unintentional weight loss in
subjects whose cardiovascular function has decompensated was not specified in previous
studies and was not ascertained here. Both weight gain and marked degrees of weight loss
were associated with increased mortality in those with pre-existing cardiovascular disease
(Figure 2b) and that seems, in part, to substantiate the suggested “obesity paradox” but only
in relation to marked weight loss (rarely achieved in routine clinical practice) in those with
pre-existing severe cardiovascular disease.
There are limitations to the interpretation of these findings. The trial was designed to meet
regulatory requirements in the form of a classical cardiovascular outcome trial but there was
no true control group as all subjects received sibutramine for 6 weeks with lifestyle
modification. There are no historical controls for similar demographics that could have been
used for comparison. Some subjects who experienced increased blood pressure and/or pulse
and should have been discontinued per the protocol were, in fact, randomised at the
investigators’ discretion. Subjects who failed to lose weight were required to continue on
study medication and the subjects could not be blinded to their weight changes. The precise
analytic techniques used in this analysis were not pre-specified, although exploration of the
effect of weight loss and its association with clinical outcome was part of the analysis plan
Nevertheless, there are important implications from these findings for the care of overweight
and obese patients with cardiovascular disease. Modest intentional weight loss, if
maintained, reduces the risk of both primary cardiovascular outcome events and
cardiovascular death. Excessive weight loss could be harmful or more likely in this
population serve as a marker of ill health or possibly cardiac decompensation. The weight of
subjects with known cardiovascular disease and diabetes, therefore, should be the focus for
active management. A stronger emphasis on lifestyle intervention is needed in all treatment
plans and will require changing the way chronic care is delivered. Treatment with
sibutramine induced extra weight loss, but the benefits were offset and, in part, masked by the
requirement to maintain subjects on the drug even when they were not losing weight. An
important consideration in the determination of the effectiveness of future weight loss drugs
will be to assess the validity of the required weight loss threshold for clear reductions in both
cardiovascular events and mortality.
The authors thank the patients and the SCOUT National Leaders and Investigators for their
participation in conducting this study and members of the SCOUT Team from Abbott and
MDS Pharma Services, Inc for site monitoring, study coordination and help with statistical
1. Van Gaal LF, Mertens IL, De Block CE. Mechanisms linking obesity with
cardiovascular disease. Nature 2006; 444 (7121):875-80.
2. Organization for Economic Cooperation and Development Health Data 2010:
Statistics and Indicators; available at
_1,00.html. Accessed 21 October 2010.
3. Prospective Studies Collaboration, Whitlock G, Lewington S, Sherliker P, Clarke R,
Emberson J, Halsey J, Qizilbash N, Collins R, Peto R. Body-mass index and cause-
specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies.
Lancet 2009; 373: 1083-96.
4. Nathan C. Epidemic inflammation: Pondering obesity. Mol Med. 2008 Jul-Aug;14(7-
5. Higgins M, D'Agostino R, Kannel W, Cobb J, Pinsky J. Benefits and adverse effects
of weight loss. Observations from the Framingham Study. Ann Intern Med
6. Lee IM, Paffenbarger RS. Change in body weight and longevity. JAMA 1992; 268:
7. Andres R, Muller DC, Sorkin JD. Long-term effects of change in body weight on all-
cause mortality. Ann Int Med 1993; 119: 737–43.
8. Yaari S, Goldburt U. Voluntary and involuntary weight loss: Associations with long
term mortality in 9,228 middle-aged and elderly men. Am J Epidemiol 1998; 148:
9. Adams TD, Gress RE, Smith SC, Halverson RC, Simper SC, Rosamond WD,
LaMonte MJ, Stroup AM, Hunt SC. Long-term mortality after gastric bypass surgery.
N Engl J Med 2007;357:753-61.
10. Sjöström L, Narbro K, Sjöström CD, Karason K, Larsson B, Wedel H, Lystig T,
Sullivan M, Bouchard C, Carlsson B, Bengtsson C, Dahlgren S, Gummesson A,
Jacobson P, Karlsson J, Lindroos A-K, Lönroth H, Näslund I, Olbers T, Stenlöf K,
Torgerson J, Ågren G, Carlsson LMS, for the Swedish Obese Subjects Study. Effects