Outcomes of Pharmacological and Surgical Treatment for Obesity

Outcomes of Pharmacological andSurgical Treatment for ObesityJennifer Cerulli and Margaret MaloneDivision of Pharmacy Practice, Albany College of Pharmacy, Albany, New York, USA

ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2691. Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2702. Aetiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2703. Health Risks and Quality of Life in Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2704. Social and Economic Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2725. Treatment of Obesity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

5.1 Pharmacotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2745.2 Surgical Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

Abstract The purpose of this article is to review the data from pharmacotherapeutic andsurgical intervention studies for the management of obesity. Clinical outcomesassessed include weight changes over time and the effects of weight loss on bloodpressure, serum lipid profiles and blood glucose control. Quality of life and eco-nomic data have been incorporated where available. Double-blind, randomisedcontrolled trials were used preferentially over shorter term open studies. Theliterature evaluation was based on a Medline search of published data betweenJanuary 1990 and January 1998.

Obesity affects 65 million adults in the US. Estimates based on 1990 datasuggest that obesity and comorbid illness contributed to $US46 billion in directcosts and $US23 billion in indirect costs in the US.

Obesity is a chronic condition which requires long term management. The riskof developing cardiovascular disease, hypertension, type 2 (non–insulin-depend-ent) diabetes mellitus, osteoarthritis, Pickwickian syndrome and cancer is in-creased in the obese population, resulting in excess morbidity and mortality.There are no long term prospective studies that have demonstrated that weightreduction in obese patients improves survival. However, on the basis of epidemio-logical data using the prevalence of disease and associated body mass index, it isgenerally accepted that weight reduction of 5 to 10% in obese patients is associ-ated with significant health benefits.

Current treatment modalities include diet and behaviour modification, exerciseand, where indicated, pharmacological intervention. Surgical intervention is re-served for the clinically severe obese patient [body mass index (BMI) > 40 kg/m2].

Many studies have demonstrated weight loss and improved metabolic fitnessover 6 to 12 months. Few studies have been conducted over a longer period.

REVIEW ARTICLE Pharmacoeconomics 1998 Sep; 14 (3): 269-2831170-7690/98/0011-0269/$07.50/0

© Adis International Limited. All rights reserved.

Limited data are available regarding reduced morbidity and mortality, improvedquality of life and functional or employment status and even fewer have incor-porated any economic assessments of the impact of medical or surgical interven-tion. Although prospective data have demonstrated reduced morbidity followingsurgical intervention, only retrospective data have demonstrated reduced mortality.

Studies of new drugs and interventions under development should demon-strate long term safety and efficacy in terms of sustained weight loss and sub-sequent weight maintenance. Future studies should incorporate assessment ofpatient perceived satisfaction with weight loss, health status and quality-of-lifeevaluations and pharmacoeconomic data to aid clinicians in the decision-makingprocess in terms of weight management of their obese patients.

1. Epidemiology

Obesity is a chronic health condition which affects65 million adults in the US.[1] The National Healthand Nutrition Examination Survey III (NHANESIII, 1988-1994) estimated the prevalence of obesityin the US adult population to be 22.5%, an increasefrom the previous NHANES II survey (1976-1980),which revealed a prevalence of 14.5%.[2]

Obesity is usually defined in terms of body massindex (BMI) which is derived from the bodyweightin kilograms divided by the square of the height inmeters (kg/m2). A goal BMI for adults aged 19 to34 years is between 19 and 25 kg/m2, and 21 to 25kg/m2 for those over 35 years. Overweight is de-fined as a BMI greater than 25 kg/m2 and pre-obe-sity as 25.0 to 29.9 kg/m2. Clinically significantobesity is classified as a BMI value ≥30 kg/m2. Thedegree of obesity can be further subdivided as fol-lows: Class I = 30 to 34.9 kg/m2, Class II = 35.0 to39.9 kg/m2, Class III ≥40 kg/m2.[2] Waist to hipratio (WHR) may also be used as a measurement todescribe the distribution of body fat. A high WHR(>0.8 in women and >0.95 in men) indicates a morecentral or abdominal adipose distribution which ismost commonly associated with the comorbiditiesof obesity. Individuals with a lower WHR (<0.8 inwomen and <0.95 in men) display a gynoid distri-bution of adiposity and are less likely to developsecondary cardiovascular risk factors.[3]

2. Aetiology

Obesity originates from an imbalance betweenenergy intake and energy expenditure. Although the

exact aetiology of the energy imbalance is not com-pletely understood, it is considered to be secondaryto environmental, genetic and metabolic factors.The human body regulates weight through com-plex interactions between the endocrine, gastroin-testinal and central nervous systems, which aremore fully described elsewhere.[4]

3. Health Risks and Quality of Lifein Obesity

Dietary factors and sedentary activity patternscontribute to 300 000 obesity-related deaths annu-ally in the US and are the second most preventablecause of death following cigarette smoking.[5] Obe-sity is associated with an increase in all-cause mor-tality, and life expectancy is reduced proportion-ally to the severity of the condition.[5,6] The healthrisks associated with obesity resulting in increasedmorbidity and mortality include cardiovasculardisease (including hypertension), type 2 (non–insu-lin-dependent) diabetes mellitus, osteoarthritis,Pickwickian syndrome, and cancer (breast, endo-metrial, prostate).[7] Wolf and Colditz[8] estimatedthe population-attributable risk percentage whichdetermines the proportion of disease which can beattributed to an increased BMI over 21 kg/m2. Forpatients with a BMI ≥29 kg/m2, 61% of type 2 di-abetes mellitus, 17.3% of coronary heart disease,17.1% of hypertension and 21.0% of gallbladderdisease can be attributed to excess weight.

In addition to the increased risk of morbidityand mortality, obese individuals have a decreasedquality of life and report more physical impairment

270 Cerulli & Malone

© Adis International Limited. All rights reserved. Pharmacoeconomics 1998 Sep; 14 (3)

than the general population. Fontaine et al.[9] ex-amined health-related quality of life (HR-QOL) in312 patients seeking outpatient medical treatmentfor obesity. The HR-QOL was measured by theMedical Outcomes Study Short Form-36 HealthSurvey (SF-36) and scores were adjusted for socio-demographic factors and comorbidities (includingdepression). The SF-36 is a general health statusmeasure divided into 8 domains: physical func-tioning; role limitations due to physical healthproblems; bodily pain; general health perception;vitality; social functioning; role limitations due toemotional problems and mental health. Comparedwith general population norms, obese patients(mean BMI 38.1 kg/m2) reported significantlylower scores in all 8 domains. Morbidly obese pa-tients (mean BMI 48.7 kg/m2) reported signifi-cantly lower scores in the areas of physical func-tioning, role limitations due to physical problems,bodily pain, general health, vitality, and socialfunctioning than those patients with moderate obe-sity (mean BMI 34.5 kg/m2) or mild obesity (meanBMI 29.2 kg/m2).

The authors[9] compared their findings withpublished data from patients with congestive heartfailure and symptomatic patients positive for humanimmunodeficiency virus. These 312 patients whowere obese reported significantly greater bodilypain than either of the latter 2 groups (p < 0.005).It is unclear whether the SF-36 scores of patientsstudied were compared with age- and gender-matched controls. This could potentially alter theconclusions if these important variables were ig-nored. A second confounding problem may lie inthe selection of the patients who were self referredto the obesity management programme, perhapssecondary to increased bodily pain and dysfunc-tion, and may not accurately represent the generalobese population. In addition, the majority of par-ticipants were White (77%) and women (74%).Most (94%) had graduated from high school and56% were college graduates. Although this maynot represent the general obese population, the dataindicate that in this population obesity was associ-

ated with a decreased HR-QOL demonstrating aninverse relationship with increasing excess weight.

Lean et al.[10] evaluated the effect of an increas-ing waist circumference or central adiposity onhealth outcomes which included the symptoms andrisks of chronic disorders. Waist circumference re-flects total and abdominal fat accumulation, is notgreatly influenced by height and is easy to measurein the home or outpatient clinic. The cross-sectionalstudy of 5887 men and 7018 women in the Nether-lands evaluated respiratory insufficiency, low backpain, the presence of type 2 diabetes mellitus andcardiovascular risk factors. The degree of physicalfunction, as assessed by 10 items from the SF-36,represented the measure for HR-QOL. Patientswere grouped according to waist circumference ac-tion levels which the authors used to estimate thedegree of risk from obesity-related comorbiditiesand to indicate when treatment was required. Indi-viduals with a waist circumference lower than ac-tion level 1 (<94.0cm in men, <80.0cm in women)were used as the reference group. Action level 1 to2 (94.0 to 101.9cm in men, 80.0 to 87.9cm inwomen) was used to indicate when patients wouldexperience increased obesity-related health risksand should avoid weight gain. An action level >2(≥102.0cm in men, and ≥88.0cm in women) wasused to indicate when professional advice shouldbe sought to lose and maintain weight.

Results of the study[10] showed that with in-creasing waist circumference, patients reported anincreased frequency of symptoms such as low backpain which was found to affect daily life activities,hinder daily business and increase absence fromwork. Patients exceeding action level 2 were 4times as likely to develop type 2 diabetes mellitus.Quality of life as assessed by physical functioningscores significantly declined with increasing waistcircumference, and patients above action level 2were twice as likely to have difficulty with physi-cal functioning. Waist circumferences over actionlevel 2 increased the likelihood of developingsymptoms and chronic disease and decreased pa-tients’ quality of life. The study authors suggestedthat waist action levels could be used for health

Outcomes Assessment in Obesity 271


promotion activities since it is an easy parameterto measure within the home setting.

4. Social and Economic Costs

The physical and psychosocial impairments as-sociated with obesity are likely to have economicimpacts. Although limited data are available, thesocial and economic (table I) costs of obesity aresubstantial. Estimates made on the basis of 1990data have suggested that if obesity had been pre-vented, the US could have saved approximately$US45.8 billion, or 6.8% of the total healthcareexpenditure for that year.[11] This estimate includedthe direct costs of treating the proportion of diabe-tes mellitus, cardiovascular disease, cancer andmusculoskeletal diseases which would be attri-butable to obesity. The estimated indirect costsassociated with obesity included 52.5 million daysof lost productivity, costing US employers ap-proximately $US4 billion, and mortality costs of$US18.9 billion secondary to loss of lifetime earn-ings.[11]

Compared with the general Swedish population,the annual number of sick-leave days and the fre-quency of disability pensions has been reported tobe doubled in obese individuals (bodyweight 100to 200 kg), independent of age and gender.[12] Ab-sence from work due to obesity was estimated tocost 6 billion Swedish kronor (SEK) per year(equivalent to approximately $US1 million) per 10000 inhabitants assuming 100% employment.[12]

However, the actual rate of unemployment in Swe-den was approximately 10%, therefore this esti-mate should be reduced. Preliminary data fromSweden estimated that the direct and indirect costsattributable to obesity were around 7 to 8% of the

total healthcare costs which is comparable to USdata.[12] In the Netherlands the direct cost of over-weight (BMI >25 kg/m2) and obesity (BMI >30kg/m2) was estimated to be approximately 1 billionDutch guilders (NLG), equivalent to roughly 4%of the total costs of healthcare.[13] The hidden coststo society of premature obesity-related morbidityand mortality, while difficult to document, shouldbe considered due to decreased productivity andfewer tax payments contributed by the individual.

Gorsky et al.[15] used an incidence-based analy-sis to estimate the excess costs associated withoverweight in women who remained overweightbetween the ages of 40 to 65 years.[15] The 25-yeardirect healthcare costs and medication costs wereestimated for 3 cohorts, each consisting of 10 000women entering the cohort at the age of 40 years.The 3 cohorts included the referent cohort (BMI 21to 24.9 kg/m2), the moderately obese cohort (BMI25 to 28.9 kg/m2), and the severely obese cohort(BMI ≥29 kg/m2). The cost of health outcomes in-cluded the cost of fatal and nonfatal heart disease,symptomatic gallstones treated by surgery, osteo-arthritis treated by outpatient joint lavage, the an-nual medication costs for the treatment of hyper-tension, and all annual costs associated withdiabetes. Over a 25-year period, the severely obesecohort incurred $US53 million in excess costs and497 excess deaths when compared with the referentcohort. The moderately obese cohort incurred$US22 million in excess costs and 212 excessdeaths compared with the referent control cohort.This model is likely an underestimation of the eco-nomic impact of obesity due to the exclusion of thecosts associated with obesity-related cancers, medi-cations used to manage osteoarthritis, physician

Table I. Economic costs of obesity

Study Country Direct cost of obesity Comment

Wolf & Colditz[11] USA $US45.8 billion (1990 values) 6.8% of total healthcare costs

Sjöström et al.[12] Sweden 7 to 8% of total healthcare costs

Seidell[13] The Netherlands NLG1 billion 4% of total healthcare costs

Quesenberry et al.[14] USAa $US220 million (1994 values) 6% of total healthcare costs for HMO

a Study included 17 118 US HMO patients.

HMO = health maintenance organisation; NLG = Dutch guilders.



visits for hypertension and management of thecomplications of hypertension.

Using computerised databases, Quesenberry etal.[14] quantified the association between BMI andhealth services use and the direct cost of providingthose health services in a large health maintenanceorganisation (HMO). Respondents to a member-ship health survey in 1993 (n = 17 118) were in-cluded in the analysis and grouped according toBMI: 20 to 24.9 kg/m2, 25 to 29.9 kg/m2, 30 to 34.9kg/m2, and ≥35 kg/m2. The survey elicited infor-mation regarding chronic and acute illnesses, theuse of prescription and over-the-counter medica-tions, height and weight, race or ethnicity, educa-tional level, alcohol consumption and cigarettesmoking. The overall response rate was 58.2%.Comorbidities associated with obesity (depres-sion, diabetes, hypertension, musculoskeletal pain,high serum cholesterol level, coronary heart dis-ease) were identified from questionnaires, chart re-view and prescription records. Health services usewas obtained through computerised records ofhospitalisations, outpatient visits, pharmacy, labo-ratory and radiology services, claims and referralsfor services outside the plan for the year 1993.

Results of the analysis[14] showed an apparentpositive association between BMI and each of the6 obesity-related comorbidities. Respondents witha BMI of 30 to 34.9 kg/m2 and ≥35 kg/m2 experi-enced a 17% and 24% higher rate of outpatientvisits, respectively, when compared with respon-dents who had a BMI of 20 to 24.9 kg/m2. Therewas a strong association between BMI and annualhospital days; respondents who had a BMI of 30 to34.9 kg/m2 and ≥35 kg/m2 sustained a 34% and74% higher rate of inpatient hospitalisation, re-spectively, relative to those who had a BMI of 20to 24.9 kg/m2. Compared with patients who had aBMI of 20 to 24.9 kg/m2, pharmacy costs were 23,60 and 78% higher in each increasing BMI cate-gory. The data revealed a strong association be-tween BMI and total costs in 1994 dollars with aslight J-shaped curve. Relative to those with a BMIof 20 to 24.9 kg/m2, total costs for respondentswith a BMI of 25 to 29.9 kg/m2 were 5% less.

Those with a BMI of 30 to 34.9 kg/m2 had a 25%increase in total costs and those with a BMI ≥35kg/m2 displayed a 44% elevation in total costs. Thestudy authors estimated that the cost of obesity totheir HMO was $US220 million, or approximately6% of the total healthcare for all members in 1994dollars. The study was limited in that the extremelyill patients were unlikely to participate in the sur-vey and the costs associated with morbid obesity(BMI >40 kg/m2) were not evaluated separately.Although self-reported heights and weights wereused, these figures are generally under-reported bypatients making these estimates conservative.

In summary, obesity has a large economic im-pact on society in terms of healthcare costs, mor-bidity and mortality. The management of obesityand its comorbidities could save an estimated 4 to7% of total healthcare costs.

5. Treatment of Obesity

The management of obesity involves either areduction in energy intake and/or an increase inenergy expenditure. Current treatment modalitiesinclude a combination of diet, exercise, behaviourmodification and, where indicated, pharmacothera-peutic or surgical intervention. Diet, exercise andbehaviour modification are the primary interven-tions for all obese patients and may be the solerequired treatment option for those individualswith pre-obesity (BMI 25 to 29.9 kg/m2 with nocomorbidities). Patients with a BMI ≥30 kg/m2, orthose with a BMI >27 kg/m2 and obesity-relatedcomorbidities, are potential candidates for drugtherapy in conjunction with diet and exercise.[16]

Surgical intervention may be appropriate for pa-tients with a BMI >40 kg/m2 who have failed med-ical intervention.[17] The decision for treatmentshould be based on the patient’s previous attemptsat weight loss with diet and exercise, the nature andseverity of their obesity-related comorbidities andthe presence or absence of other medical condi-tions.[16]

When evaluating treatment modalities for themanagement of obesity, it is essential to recognisethe health benefits of weight reduction. Evaluations



of weight loss strategies assume a morbidity andmortality benefit based on the knowledge that epi-demiological studies report decreased morbidityand mortality with a 5 to 10% intentional reductionin bodyweight. However, few prospective data areavailable to conclude that chronic bodyweight re-duction will reduce mortality. In a prospectiveevaluation, Williamson et al.[18] reported that innever-smoking women with pre-existing obesity-related conditions, intentional weight loss was as-sociated with a 20% reduction in all-cause mortali-ty. A loss of only 5 to 10% of bodyweight has beenreported to improve the health risks associated withobesity such as insulin resistance, hypertension,and hyperlipidaemia.[19,20] In addition, controlledtrials of pharmacotherapeutic agents have shownthat a 5 to 15% weight reduction improves obesity-related risk factors such as blood pressure, dys-lipidaemia and insulin resistance.[16] However, thesetrials are generally of short duration and weight isoften regained after discontinuation of the agents.The benefits of weight loss are only maintained ifthe reduction is sustained. Therefore, weight main-tenance or sustained weight loss are important pa-rameters to assess in clinical trials in addition toother clinical outcomes such as reduction in bloodpressure, lipid levels and insulin resistance.

The following subsections review data frompharmacotherapeutic and surgical interventionstudies for the management of obesity. The litera-ture evaluation was based on a Medline search ofpublished data between January 1990 and January1998 to identify English literature regarding themanagement of obesity. The search was supple-mented by a review of the bibliographies of the iden-tified literature. Double-blind, placebo-controlledtrials of greater than 6 months’ duration which in-cluded clinical outcome measures in addition toweight loss were utilised preferentially. These datawere evaluated for study design, patient popula-tion, results, dropout (or withdrawal) rate and theincidence of adverse events. The clinical outcomesassessed included bodyweight lost or regained,changes in blood pressure, lipid profiles and insulin

resistance. Quality-of-life and economic data havebeen incorporated where available.

5.1 Pharmacotherapy

The pharmacological agents currently availableor awaiting approval for the management of obesityreduce energy intake either through modulation ofappetite or by altering nutrient absorption. Appetiteis modulated by several neurotransmitters in theCNS. Food intake is increased by stimulation of theserotonin-1a receptor (5-hydroxytryptamine1a; 5-HT1a) and α2 adrenergic receptors, and is decreasedby stimulation of 5-HT1B/D, 5HT2C, α1 and β2 adre-nergic receptors.[21] Anorexiant agents suppressappetite through enhancement of either adrenergicactivity (phentermine), serotonergic activity (fen-fluramine, dexfenfluramine) or through modula-tion of both neurotransmitters (sibutramine). Or-listat, an inhibitor of pancreatic and gastric lipase,decreases fat absorption resulting in decreased en-ergy intake and weight loss. It is awaiting finalFood and Drug Administration (FDA) approval inthe US. β3 adrenergic receptor agonists which en-hance thermogenesis and increase energy expendi-ture are under investigation for the management ofobesity. Future pharmacological agents may de-crease appetite through modulation of hormoneswhich control appetite in the CNS. Antagonists ofneuropeptide-Y (NPY), a hormone which stimu-lates appetite, and agonists of leptin, a protein se-creted by adipocytes to reduce food intake, are un-der development.[4]

5.1.1 Adrenergic AgentsAdrenergic agonists enhance catecholamine

neurotransmission to increase sympathetic activityand reduce appetite. The first generation of norad-renergic agents (amphetamine, methamphetamine,and phenmetrazine) are no longer recommendedfor the management of obesity due to the risk ofabuse associated with their use.[16] Chemical ma-nipulation of the amphetamine structure led to thedevelopment of agents which maintain their an-orexiant properties through modulation of thenoradrenergic system. However, they exert a re-duced effect on dopaminergic transmission, there-



fore, decreasing their potential for abuse.[22] Theseagents include benzphetamine, phendimetrazine,diethylpropion, mazindol, phenylpropanolamineand phentermine. There are few reported phar-macoeconomic evaluations of these agents. Stud-ies evaluating outcomes in terms of bodyweightreduction have been reviewed elsewhere.[16,22]

In 1978, Dahms et al.[23] conducted a cost-benefitassessment in 144 obese patients who wererandomised to receive either behavioural therapy,placebo, mazindol 1mg orally 3 times daily, ordiethylpropion 25mg orally 3 times daily for 14weeks in addition to a calorically restricted diet(approximately 800 kcal for most patients). All pa-tients were found to have similar weight loss re-sults and behaviour modification was determinedto be 6 to 9 times as cost efficient as the medicationtreated group. However, there are several limita-tions to this analysis. Only 33 patients completedthe 14-week study and the cost assessment wasbased solely on the cost of the healthcare profes-sionals’ time. The behaviour modification groupcontaining 8 to 10 patients was seen by the dietitianwhose salary was estimated to be one-third of thesalary of a physician. The drug treatment groupswere all managed by a physician. Since this factorcontributed the major portion of the cost variabi-lity, clearly the group which did not require physi-cian intervention was less expensive to manage.The cost of the medication was not included in theanalysis.

5.1.2 Serotonergic AgentsSerotonergic agents, which include fenfluramine,

dexfenfluramine and the selective serotonin reup-take inhibitors (SSRIs), decrease food intake byincreasing serotonin levels in the synapse. SSRIsare not indicated for the management of obesityand can not be recommended for use.[16] Presently,fenfluramine and dexfenfluramine have been vol-untarily withdrawn from the market due to con-cerns of valvular heart disease associated with theiruse. The initial reported case series which identi-fied this problem involved 24 women treated withthe phentermine-fenfluramine combination for amean of 12.2 ± 6.8 months.[24] Echocardiography

in all patients demonstrated unusual valvular mor-phology and regurgitation, and newly documentedpulmonary hypertension was detected in 8 women.Cardiac surgical intervention was necessary in 5patients. Subsequently, further cases of cardiacvalvular abnormalities have been reported to theFDA.[25]

Another concern regarding the long term use ofanorexiant agents involves the possibility of thedevelopment of primary pulmonary hypertension(PPH). PPH is a rare disease with an annual inci-dence estimated at 1 per 200 000 to 1 to 2 per mil-lion.[26] The International Primary Pulmonary Hy-pertension Study Group conducted a case controlstudy involving 95 patients with PPH and 335 age-and gender-matched controls.[27] Their results in-dicated an increased odds ratio of 23.1 (95% CI,6.9 – 77.7) when anorectic drugs were used for aperiod of longer than 3 months. In an attempt toplace the risk in perspective, Manson and Faich[28]

estimated that 1260 excess lives are lost per millionwomen per year as a consequence of an average of13kg of excess weight per patient. When comparedwith 28 cases per million person-years of PPH as-sociated with appetite suppressant drugs, they cal-culated a benefit:risk ratio of 20:1 associated withappetite suppressant drug therapy, if a 50% mortali-ty rate is associated with PPH. On the basis of thisestimation, the use of these agents could result in280 lives saved as compared with 14 deaths causedby the drugs per million person-years of treatment.

Despite the withdrawal of these agents from theworld market, data from the largest long term clini-cal studies using the phentermine-fenfluraminecombination and dexfenfluramine have been in-cluded in this discussion.[29-37] The publication ofthese landmark studies began a new era in the man-agement of obesity now recognised as a chronicdisease requiring long term management. The body-weight loss and clinical outcomes reported in thesestudies established a benchmark by which weightloss interventions are measured. The limitations ofthese studies served as starting points for furtherobesity drug-related investigations.



FenfluramineFenfluramine is the most widely studied sero-

tonergic drug for the management of obesity.Fenfluramine partially inhibits the reuptake of se-rotonin and releases serotonin into the synapticcleft. Both actions enhance serotonergic activitywhich reduces food intake. Prior to withdrawalfrom the market, fenfluramine was frequently pre-scribed in combination with phentermine. The useof each of these agents as monotherapy was limiteddue to their adverse effects which included excessCNS stimulation by phentermine, and lethargy anddrowsiness associated with fenfluramine. In com-bination, these adverse effects were largely nega-ted and the resultant weight loss was significant.

The largest double-blind, placebo-controlledstudy conducted in 121 obese patients evaluated theeffect of an extended-release preparation of fen-fluramine 60 mg/day combined with phentermineresin 15 mg/day versus placebo in conjunction withbehaviour modification, a calorically restricteddiet, and an exercise programme.[29-36] The studywas conducted in several phases with the maxi-mum follow-up period being 210 weeks. Thereader is referred to the original data for a completediscussion of the study design. Study participantswere 130 to 180% (154 ± 1.2%, mean ± SEM) abovetheir ideal bodyweight, with no obesity-relatedcomorbidities and were not taking long term medi-cation for other illnesses. This study series indi-cated the importance of continuous drug treatmentas diet, exercise and behaviour modification alonewere not enough to sustain weight loss after medi-cation discontinuation. The authors indicated thatmaintenance of lost weight, without further weightloss, while remaining on anorexiant medicationsdid not indicate tolerance but displayed continuedefficacy. A large number of patients (70/121)dropped out before week 190 limiting the interpre-tation of the study findings; however, this high rateof recidivism is common in obesity studies. Pa-tients with obesity-related comorbidities such ascardiovascular disease and diabetes mellitus wereexcluded, making these data difficult to extrapolateto the general obese population. Outcomes such as

the impact on health status and quality of life werenot addressed. However, these data provided in-sight into the long term safety and efficacy of ano-rexiant therapy and the need for long term chronictreatment.

DexfenfluramineDexfenfluramine is the dextro rotatory stereo-

isomer of fenfluramine. Dexfenfluramine was in-dicated for the management of obesity in patientswith a BMI >30 kg/m2, or 27 kg/m2 with risk factors(hypertension, diabetes mellitus, hyperlipidaemia).For an in depth discussion of dexfenfluramine thereader is referred to 2 comprehensive reviews.[38,39]

There are several short and long term studiesevaluating the use of dexfenfluramine in the obesepopulation.[37,40] The largest study to demonstratethe efficacy and long term safety of dexfenflura-mine was a 12-month, randomised, double-blind,placebo-controlled multinational trial of 822 obesepatients.[37] Patients received either dexfenflura-mine 15mg twice daily (n = 404) or placebo (n =418) in addition to a calorie-restricted diet. After12 months, the treatment group lost significantlymore weight than the placebo group [9.82kg ± 0.5(SEM) vs 7.15kg ± 0.49; p < 0.001]. Twice as manypatients in the treatment group compared with theplacebo group lost greater than 10% of their body-weight (34.9% vs 17%; p < 0.001). Most of theweight loss occurred in the first 6 months and thenreached a plateau. Patients in the placebo group hada significant tendency to regain weight (p < 0.05).The effect on blood pressure and serum lipids wasnot evaluated. At the conclusion of the study, 189(45%) patients had withdrawn from the placebogroup and 150 (37%) from the dexfenfluraminegroup.

Dexfenfluramine has been evaluated in severalshort term, double-blind, controlled studies for themanagement of obese patients with diabetes melli-tus.[41-43] Significant reductions were reported inclinical outcomes including bodyweight, BMI,haemoglobin A1c concentration and blood pressureand improved insulin sensitivity and glucose dis-posal.[41-43] No studies using dexfenfluramine for



weight reduction have reported pharmacoecono-mic benefits or improvements in health status.

5.1.3 Adrenergic/Serotonergic Agent: SibutramineSibutramine is a tertiary amine with 2 active

metabolites which are potent inhibitors of bothnoradrenaline (NA) and serotonin reuptake.[44] Inanimal models, it has been shown to dose-depend-ently reduce food intake by enhancement of satietythrough a synergistic interaction between en-hanced NA and serotonin activity in the CNS. Thedrug has recently been approved by the FDA forthe management of obesity. Dose-ranging studiesand clinical trials have demonstrated dose-relatedweight loss which was maintained for up to 2.5years.[45-48]

Sibutramine was compared with placebo in a26-week double-blind, placebo-controlled trial in-volving 173 healthy obese patients.[46] Patientswith diabetes mellitus and those receiving medica-tion for hypertension were excluded. In addition toa standardised exercise prescription, men wereprescribed a 1500 kcal/day diet and women wereprescribed a 1200 kcal/day diet. Following a 2-week single-blind run-in period, patients wererandomised to receive either placebo or one of 6doses of sibutramine (1, 5, 10, 15, 20 or 30 mg/day)for 24 weeks. After 24 weeks, weight loss in thepatients receiving the 1 or 5 mg/day doses was notsignificantly different from placebo. The mean(±SEM) weight loss was 5.8 ± 0.96kg for the 10mg/day dose; 6.4 ± 0.96kg for the 15 mg/day dose;7.2 ± 0.96kg for the 20 mg/day dose; 8.3 ± 0.94kgfor the 30 mg/kg dose; and 0.8 ± 0.96kg in theplacebo group. When the drug was discontinued,weight was regained in all groups with the mostrapid gain in those patients who had lost the mostweight. Modest increases in blood pressure andpulse rate were noted and small changes in bloodglucose and serum cholesterol profiles occurred(data not shown). The dropout rate was 15% for allpatients during the study. Investigators withdrew14 patients from the study; 1 due to hypertension,2 secondary to tachycardia, 3 due to emotional orpsychiatric symptoms due to either the medicationor the rapid weight loss and 8 due to various other

reasons. Long term data, currently available in ab-stract form, indicate safety and efficacy for 2.5years.[48] Currently there are no pharmacoecono-mic data or reports of improvements in health sta-tus or quality of life.

5.1.4 Lipase Inhibition: OrlistatLow fat diets are often recommended for the

treatment of obesity in an attempt to decrease en-ergy and cholesterol intake. The ready availabilityof high fat foods and the poor palatability of lowfat foods make compliance with these diets diffi-cult. Orlistat, a natural product of Streptomycestoxytricini, is a chemically synthesised derivativeof lipstatin.[49] It is a potent and irreversible inhi-bitor of gastric lipase, pancreatic lipase and pan-creatic carboxylester lipase. By inhibition of theseenzymes, administration of the drug results in adecrease in the systemic absorption of dietary fatand excretion of unabsorbed triglyceride and cho-lesterol. The resulting decrease in bodyweight andreduction in plasma cholesterol levels may helpdecrease obesity-related risk factors.

Several short term trials have shown orlistat tobe safe and effective in reducing weight, total andlow density lipoprotein (LDL) cholesterol lev-els.[50,51] Preliminary data have been reported froma single centre which participated in a 1-year multi-centre study designed to assess the drug’s efficacyand tolerability.[52] In a double-blind, randomised,parallel-group, placebo-controlled study, orlistat120mg 3 times daily was compared with placeboin 46 patients. Patients were placed on a 30% fatdiet designed to cause an energy deficit of approxi-mately 600 kcal/day. After a 4-week single-blind,placebo run-in period, patients were randomised todrug treatment (n = 23) or placebo (n = 23). Maxi-mum weight loss was observed at 6 months, whenthe reduction in bodyweight was 8.6 ± 5.4kg (SD)in the treatment group (8.4% reduction) and 5.5 ±4.5kg (SD) in the placebo group (5.7% reduction,statistical significance not reported). At 1 year, pa-tients in the orlistat group sustained an 8.4% weightreduction from their initial bodyweight, whilethose in the placebo group relapsed maintainingonly a 2.6% reduction in bodyweight. Although



mean total cholesterol did not change in the treat-ment group, the placebo group had a 9.7% increase.Serum LDL-cholesterol levels in the orlistat groupdeclined (4.2%) while increasing in the placebogroup (4.8%). Serum levels of high density lipo-protein (HDL) cholesterol showed small to modestincreases in both groups. Minor increases in dias-tolic blood pressure were noted for both groups.Weight loss, which occurred in both groups, wasgenerally associated with a decrease in blood pres-sure, therefore the increase noted could not be ex-plained by the study authors.

There were 10 adverse events leading to with-drawal from the study, 6 occurred in the placebogroup (2 depression, 1 oesophagitis, 2 dyspepsia, 1hiatal hernia) and 4 in the orlistat group (2 vomit-ing, 1 bradycardia, 1 episode of syncope). Symp-toms relating to the intestinal effects of the drug(loose stools, anal leakage) did not lead to with-drawal from the study. Serum levels of α-tocophe-rol and β-carotene declined in the treatment groupwith 6 patients requiring fat soluble vitamin sup-plementation as a result of 2 consecutive serumvitamin levels falling outside the reference range.Vitamin A levels were not affected and vitamin Dlevels fell but not sufficiently to reach levels ofconcern. Analysis of prothrombin time, as a mea-sure of vitamin K activity, revealed no alterations.A 65-item self administered quality-of-life (QOL)questionnaire was completed by patients at the be-ginning of the run-in period, at the start of the treat-ment period and at weeks 24 and 52 to evaluatechanges in their quality of life following treatment(the questionnaire was developed and validated forthe pharmaceutical manufacturer). The primarymeasures of outcome were patient’s scores on 3scales; overweight distress, depression, and satis-faction with treatment. Although the data were notpresented, the authors reported that QOL indicesshowed modest improvements in both the placeboand the control group. Interestingly, the intestinaladverse effects reported in the orlistat group did notaffect their QOL responses. No economic data werereported.

5.2 Surgical Management

Clinically severe obesity is defined as a BMI>40 kg/m2 which is roughly equivalent to 100pounds overweight for an average adult man. Forpatients with severe obesity, or in selected patientswith a BMI >35 kg/m2 and obesity-related comor-bidities, surgical management provides a reason-able alternative after conventional treatment hasfailed.[17] A number of surgical procedures havebeen investigated; however, the vertical bandedgastroplasty (VBG) and the gastric bypass (GB)are the most common techniques.[17] The VBG in-volves the construction of a small pouch in thestomach with a restricted outlet along the lessercurvature of the stomach. Intestinal integrity ismaintained and the method of weight loss involvesdecreased energy intake. Gastric bypass which isthe preferred procedure involves construction of aproximal gastric pouch whose outlet is a limb ofthe jejunum. The remainder of the stomach is at-tached normally to the duodenum. In this proce-dure the weight loss is a result of reduced intakeand malabsorption as the contents from the smallgastric pouch empty directly into the jejunum.Compared with VBG, GB provides greater sus-tained weight loss possibly due to the constant nega-tive reinforcement associated with the ingestion ofexcess carbohydrates which results in dumpingsyndrome.

Weight reduction surgery has been reported toproduce long term weight loss with the majority ofweight loss occurring in the first few months;weight loss reaches a plateau at 18 to 24 months.Pories et al.[53] reported the benefits of weight re-duction surgery in 608 patients who were morbidlyobese and underwent GB with a maximum follow-up of 14 years. The average maximum weight lossof 70% of excess weight occurred 2 years after sur-gery. After 14 years, the average weight loss was49% of excess bodyweight. The perioperative mor-tality rate (within the first 30 days) was 1.5% rep-resenting 9 of 608 patients (5 sepsis, 3 pulmonaryembolus, 1 unknown). The major causes of peri-operative morbidity included minor wound infec-tions (8.7%), wound seromas (5.8%), severe wound



infections (3.0%), and anastomotic stenosis (3.0%).The total mortality over 14 years of follow-up was34 of 608 patients with 9 perioperative and 25 latedeaths. Late complications included vitamin B12deficiency (40%), anaemia (39%), hospital read-mission (38.1%) incisional hernia (23.9%), depres-sion (23.4%), staple line failure (15.1%), gastritis(13.2%), cholelithiasis (11.4%) and bile reflux(8.7%).

The improvement of comorbid conditions is anadditional purported benefit of weight reductionsurgery. Pories et al[53] reported the effects ofweight reduction surgery in a subset of their pa-tients with type 2 diabetes mellitus (n = 146) andimpaired glucose tolerance (n = 152). 121 (82.9%)of the patients with diabetes mellitus maintainednormal fasting blood glucose levels and glyco-sylated haemoglobin values post operatively (aver-age years of follow-up not reported). Among thepatients with impaired glucose tolerance, followedfor an average of 7.6 years after surgery, 2 patients(1.3%) progressed to type 2 diabetes mellitus, and150 patients (98.7%) reverted to euglycemia. Priorto surgery, 58.1% of all patients had hypertension,the incidence of which was reduced to 14% aftersurgery. The authors also reported improvementsin cardiopulmonary status, sleep apnoea, asthmaand peptic reflux and fewer limitations of physicalactivity from arthritis.

The same investigators compared the outcomeof 154 morbidly obese patients with type 2 diabetesmellitus who underwent GB with that of 78 mor-bidly obese patients with diabetes mellitus who didnot undergo the procedure due to personal reasonsor lack of insurance coverage.[54] Outcomes as-sessed included mean blood glucose levels, theneed for medical management of their diabetes andsurvival. The 2 groups were comparable at baselinewith regard to age, weight, BMI, gender and thepercentage with hypertension. The surgical grouphad a maximum weight reduction of 62.4% of ex-cess bodyweight at 1 year after surgery and main-tained an average excess weight loss of 50% for upto 14 years of follow-up. Due to the retrospectivenature of the study, bodyweight at follow-up and

mean blood glucose levels for the control groupwere not obtained. In the surgical group the meanblood glucose level fell from 187 mg/dl preopera-tively to less than 140 mg/dl postoperatively for upto 10 years of follow-up. In addition, the percent-age of surgical patients requiring oral hypoglycae-mics preoperatively was 31.8% which decreased to8.6% at the last follow-up (p = 0.001). In contrast,the need for medical management for diabetes inthe control group increased from 56.4% at the initialcontact to 87.5% at the time of last contact (p =0.0003). Despite a shorter mean length of follow-up in the control group (6.2 years) the mortality ratewas 28% compared with 9% in the surgical groupafter a mean follow-up of 9 years (p < 0.0003).

The reduction in mortality was mainly attribu-table to a reduction in the number of cardiovasculardeaths which were responsible for 54.5% of thedeaths in the control group and 14.3% in the surgi-cal group. Perioperative mortality in 4 surgical pa-tients accounted for 28% of the deaths in thatgroup. The overall perioperative mortality rate forthe 154 patients was 2.6%. Obtaining informationregarding the medical management of diabetes andsurvival by telephone interview in the controlgroup was a limitation. Regardless, the data areextremely valuable revealing decreased morbidityand mortality among obese individuals with diabe-tes who underwent a GB. Noting the limitations,the authors suggest the only ethical way to conductan evaluation would be to use a case-controlledcomparison with adequate follow-up in both groups.Such an evaluation is being conducted by the Swe-dish Obese Subjects (SOS) study.

The SOS study is an on-going study consistingof a nationwide registry and an intervention pro-gramme.[12] The intervention study involves a sur-gically treated group (gastric banding, VBG and GB)and a conventionally treated control group. Bothgroups will be followed for 10 years to determinewhether the negative effects of severe obesity onmorbidity and mortality are reduced by sustainedweight loss using a non-randomised matched-group design. Preliminary data revealed the 2-yearincidence of new cases of diabetes was 0.5% in the



surgical group and almost 7% in the obese controlgroup.[12] Among patients who had hypertension atbaseline, 43% in the surgical group became normo-tensive compared with only 22% in the controlgroup.

In addition to weight loss and reduced comorbidconditions, psychosocial improvements have beenreported following weight reducing operative pro-cedures. The SOS study has reported the prelimi-nary results of the effect of surgery on HR-QOL in487 surgical patients and 487 conventionally treatedmatched controls followed for 2 years.[55] Patientdemographics were similar at baseline with the ex-ception that the surgical patients were significantlymore obese than controls. After 2 years, 477 surgi-cal patients and 397 patients in the control groupremained in the study. The average reduction inBMI in the surgical patients was 9.3 (CI 8.5 to 10.1)kg/m2 in men and 9.8 (CI 9.2 to 10.3) kg/m2 inwomen. The average weight loss in the controlgroup was minor resulting in small changes in BMI(0.2 to 0.4 kg/m2). Prior to intervention, surgicalpatients displayed lower overall HR-QOL than pa-tients in the control group, suggesting that obeseindividuals who chose surgical management expe-rienced greater psychosocial burden. An analysisof the effects of treatment on HR-QOL showed sig-nificant improvements in the surgical patients after24 months on all measures which were related tothe magnitude of weight loss. Patients who weresurgically treated and sustained a weight reductionof >20kg (n = 325) reported significantly betterHR-QOL scores than those patients with lessweight loss (<20kg, n = 152).

Van Gemert et al.[56] evaluated the quality of lifeof 62 patients who were morbidly obese and under-went bariatric surgery and compared their resultsto 20 preoperative controls. Surgical patients had 1of 3 operative procedures, Roux-en-Y GB (n = 18),VBG with a Marlex band (n = 14) or VBG with aDacron band (n=30). Quality of life was assessedby the Nottingham Health Profile and a visual ana-logue scale (VAS). The control group was found tohave poorer quality of life when compared with thesurgical patients whose scores were similar to ref-

erence values. In the surgical group a lower BMIand higher percentage of bodyweight loss resultedin better HR-QOL scores. No significant differencein HR-QOL scores was found between the type ofsurgical procedure performed. Despite the lack ofmatched controls, the data showed that beforebariatric surgery, patients who were obese experi-enced reduced quality of life which improved afterthe procedure to near reference values and was cor-related with the amount of weight lost.

Isacsson et al.[57] compared HR-QOL scores ofobese patients (n = 102, mean BMI after surgery 31kg/m2, range 22 to 39 kg/m2) 1.5 to 5.5 years afterVBG for morbid obesity with those of age-matchedpatients (n = 74, mean BMI 26.6 kg/m2, range 19.5to 29.6 kg/m2) who underwent elective cholecys-tectomy in the same institution during the sametime period. A questionnaire was sent to patients bymail which contained 82 questions aimed at inves-tigating quality of life. Questions were pooled into6 categories: subjective health and symptoms, so-cial life, family life, clothes, leisure activities, andeating. Portions of the questionnaire were pre-viously validated in a prior study. There was nodifference in the overall quality of life between thegroups, and patients who underwent VBG reporteda profound change in their life. Most of thesechanges reflected improvements in self esteem, re-lations with partners and social networking. Thedata currently available suggest that surgical can-didates differ from the general obese population atbaseline and that psychosocial functioning im-provements after surgery correlate with the magni-tude of the weight lost.

As with any medical or surgical intervention,the benefits of weight reduction surgery must beweighed against the potential risks. The NationalBariatric Surgery Registry (NBSR) data reportsperioperative mortality to be 0.1% with 5 deathsoccurring within 40 days of operation in 5178 pa-tients.[58] The causes of death included gastric/duo-denal rupture, intestinal vascular insufficiency,acute cerebrovascular disease, pulmonary embo-lism, and intestinal perforation. Severe complica-tions included deep vein thrombosis (0.4%) and



gastrointestinal leaks (0.6%). The most frequentlyoccurring perioperative complications involvedrespiratory complications in 4.5% of patients.Older, heavier patients were found to be at thegreatest risk of developing complications.

Martin et al.[59] examined the long term costsand outcomes of gastric bypass (n = 201) comparedwith medical therapy (n = 161). Medical therapyincluded 12 weeks of a very low calorie diet(VLCD) and 18 months of weekly behaviouralmodification meetings. Patients were included ifthey underwent GB or consumed the VLCD in ad-dition to completion of at least 4 months ofbehavioural therapy. The surgical group was ini-tially heavier than the medical group (mean ±SEM; BMI = 49.3 ± 0.6 kg/m2 vs 41.2 ± 0.7 kg/m2,p < 0.01). The total cost of medical therapy was$US3000 and included physician visits, laboratorymonitoring, electrocardiogram (ECG), x-ray,behavioural therapy and the liquid diet. The totalcost of surgical therapy was $US24 000 and in-cluded physician visits, laboratory monitoring,ECG, x-ray and surgery costs, and 6-day hospitalcharges. The cost per pound of weight lost for allpatients successfully monitored was calculated forat least 2 years after medical treatment or surgery.Patients without complete monitoring at 2 years(lost to follow-up) were considered treatment fail-ures and were included in the cost analysis. At 2years, the cost per pound of surgical treatment ex-ceeded that of medical treatment; however, after 6years the cost per pound of medical therapy ex-ceeded that of surgical therapy due to the long termsuccess of the GB procedure (actual dollar valuesfor the cost per pound were not reported). After 6years, 88% of the surgical patients were successfulat maintaining weight loss versus 29% of the pa-tients who were medically treated (p < 0.05). Therewere no estimates made for additional costs whichcould be incurred due to surgical complications orcost reduction due to the decreased need for medi-cations to treat obesity-related comorbid condi-tions. These data indicated that without attemptingto estimate savings based on improvement in mor-

bidity and mortality, the cost per pound was lessfor surgical patients at 6 years following GB.

Despite the potential benefits of weight reduc-tion, it has not been prospectively proven that itdecreases the previously obese individual’s risk ofdeath to the level of never obese individuals. Ten-year follow-up data from the SOS study shouldhelp to determine whether sustained weight loss inobese individuals reduces morbidity and mortality.These data will also provide direct and indirectcosts on the management of obesity and allow fora thorough economic evaluation of the surgical andmedical management of obesity. In addition toweight loss, complications, and quality of life, thisstudy is also evaluating the impact of obesity andweight loss on hospitalisation days, sick days, phy-sician visits and annual income. Preliminary dataindicate that although surgical patients spent 6times longer in the hospital than obese control pa-tients in the first year, the second postoperativeyear shows a hospitalisation rate similar to theobese control rate.[12]

6. Conclusion

Obesity is now recognised as a chronic diseaserequiring long term intervention. The goal of treat-ment is to maintain a sustained weight loss in anattempt to reduce the morbidity and mortality risksassociated with this disease. There are few thera-peutic options currently available and limited dataregarding quality of life or pharmacoeconomicbenefits of weight loss. Current pharmacotherapyproduces modest weight loss which is often re-gained after treatment is discontinued and may beassociated with adverse effects. Future pharmaco-logical options need to incorporate sustained, longterm weight loss with minimal adverse effects al-lowing for life-long treatment. For morbidly obeseindividuals, surgical treatment is a viable option aspreliminary data reveal improvements in obesity-related comorbidities and reductions in mortality.Future studies should incorporate assessment ofpatient perceived satisfaction with weight loss,health status and QOL evaluations and pharmaco-economic data to aid clinicians in the decision-



making process for the weight management of theirobese patients.

References1. Williamson DF. Descriptive epidemiology of body weight and

weight change in U.S. adults. Ann Intern Med 1993; 119 (7Pt 2): 646-9

2. Flegal KM, Carroll MD, Kuczmarski RJ, et al. Overweight andobesity in the United States: prevalence and trends, 1960-1994. Int J Obes 1998; 22: 39-47

3. Bray GA. Pathophysiology of obesity. Am J Clin Nutr 1992; 55Suppl.: 488S-94S

4. Rosenbaum M, Leibel RL, Hirsch J. Obesity. N Engl J Med1997; 337 (6): 398-407

5. McGinnis M, Foege WH. Actual causes of death in the UnitedStates. JAMA 1993; 270 (18): 2207-12

6. Manson JE, Willett WC, Stampfer MJ, et al. Body weight andmortality among women. N Engl J Med 1995; 333 (11): 677-85

7. Pi-Sunyer FX. Medical hazards of obesity. Ann Intern Med1993; 119 (7 t 2): 655-60

8. Wolf AM, Colditz GA. Social and economic effects of bodyweight in the United States. Am J Clin Nutr 1996; 63 Suppl.;466S-9S

9. Fontaine KR, Cheskin LJ, Barofsky I. Health-related quality oflife in obese persons seeking treatment. J Fam Pract 1996; 43(3): 265-70

10. Lean MEJ, Han TS, Seidell JC. Impairment of health and qualityof life in people with large waist circumference. Lancet 1998;351: 853-6

11. Wolf AM, Colditz GA. The cost of obesity: the US perspective.Pharmacoeconomics 1994; 5 Suppl. 1: 34-7

12. Sjöström L, Narbro K, Sjöström D. Costs and benefits whentreating obesity. Int J Obes 1995; 19 Suppl. 6: 9S-12S

13. Seidell JC. The impact of obesity on health status: some impli-cations for health care costs. Int J Obes 1995; 19 Suppl. 6:13S-6S

14. Quesenberry CP, Caan B, Jacobson A. Obesity, health servicesuse, and health care costs among members of a health main-tenance organization. Arch Intern Med 1998; 158: 466-72

15. Gorsky RD, Shaffer PA, Pamuk E, et al. The 25-year health carecosts of women who remain overweight after 40 years of age.Am J Prev Med 1996; 12 (5): 388-94

16. National Task Force on the Prevention and Treatment of Obe-sity. Long-term pharmacotherapy in the management of obe-sity. JAMA 1996; 276 (23): 1907-15

17. National Institutes of Health (NIH). Gastrointestinal surgery forsevere obesity. Ann Intern Med 1991; 9 (1): 25-7

18. Williamson DF, Pamuk E, Thun M, et al. Prospective study ofintentional weight loss and mortality in never-smoking over-weight US white women aged 40-64 years. Am J Epidemiol1995; 141 (12): 1128-41

19. Goldstein DJ. Beneficial health effects of modest weight loss.Int J Obes 1992; 16: 1-19

20. Stunkard AJ. Current views on obesity. Am J Med 1996; 100:230-6

21. Bray GA. The new era of drug treatment. Pharmacologic treat-ment of obesity: symposium overview. Obes Res 1995; 3Suppl. 4: 415S-7S

22. Bray GA. Use and abuse of appetite-suppressant drugs in thetreatment of obesity. Ann Intern Med 1993; 119 (7 Pt 2): 707-13

23. Dahms WT, Molitch ME, Bray GA, et al. Treatment of obesity:cost-benefit assessment of behavioral therapy, placebo, andtwo anorectic agents. Am J Clin Nutr 1978; 31: 774-8

24. Connolly HM, Crary JL, McGoon MD, et al. Valvular heartdisease associated with fenfluramine-phentermine. N Engl JMed 1997; 337 (9): 581-8

25. Centers for Disease Control. Cardiac valvulopathy associatedwith exposure to fenfluramine or dexfenfluramine: U.S. De-partment of Health and Human Services interim public healthrecommendations, November 1997. MMWR Morb MortalWkly Rep 1997; 46 (45): 1061-6

26. The International Primary Pulmonary Hypertension Study(IPPHS). Chest 1994; 105 Suppl. 2: 37S-41S

27. Abenhaim L, Moride Y, Brenot F, et al. Appetite suppressantdrugs and the risk of primary pulmonary hypertension. N EnglJ Med 1996; 335 (9): 609-16

28. Manson JE, Faich GA. Pharmacotherapy for obesity: do thebenefits outweigh the risks? N Engl J Med 1996; 335 (9):659-60

29. Weintraub M, Sundaresan PR, Madan M, et al. Long-termweight control study I (weeks 0 to 34): the enhancement ofbehavior modification, caloric restriction, and exercise byfenfluramine plus phentermine versus placebo. Clin Phar-macol Ther 1992: 51 (5): 586-94

30. Weintraub M, Sundaresan PR, Schuster B, et al. Long-termweight control study II (weeks 34 to 104): an open label studyof continuous fenfluramine plus phentermine versus targetedintermittent medications as adjuncts to behavior modification,caloric restriction, and exercise. Clin Pharmacol Ther 1992;51 (5): 595-601

31. Weintraub M, Sundaresan PR, Schuster B, et al. Long-termweight control study III (weeks 104 to 156): an open-labelstudy of dose adjustment of fenfluramine and phentermine.Clin Pharmacol Ther 1992; 51 (5): 602-7

32. Weintraub M, Sundaresan PR, Schuster B, et al. Long-termweight control study IV (weeks 156-190): the second double-blind phase. Clin Pharmacol Ther 1992; 51 (5): 608-14

33. Weintraub M, Sundaresan PR, Schuster B, et al. Long-termweight control study V (weeks 190-210): follow-up of parti-cipants after cessation of medication. Clin Pharmacol Ther1992; 51 (5): 615-8

34. Weintraub M, Sundaresan PR, Cox C. Long-term weight controlstudy VI: individual participant response patterns. Clin Phar-macol Ther 1992; 51 (5): 619-33

35. Weintraub M, Sundaresan PR, Schuster B. Long-term weightcontrol study VII (weeks 0 to 210): serum lipid changes. ClinPharmacol Ther 1992; 51: 634-41

36. Weintraub M. Long-term weight control study: conclusions.Clin Pharmacol Ther 1992; 51 (5): 642-6

37. Guy-Grand B, Apfelbaum M, Crepaldi G, et al. Internationaltrial of long-term dexfenfluramine in obesity. Lancet 1989; 2:1142-4

38. McTavish D, Heel RC. Dexfenfluramine. Drugs 1992; 43 (5):713-33

39. Davis R, Faulds D. Dexfenfluramine: an updated review of itstherapeutic use in the management of obesity. Drugs 1996; 52(5): 696-724

40. Guy-Grand B. Clinical studies with dexfenfluramine: from pastto future. Obes Res 1995; 3 Suppl. 4: 491S-6S

41. Scheen AJ, Paolisso G, Salvatore T, et al. Improvement of insu-lin-induced glucose disposal in obese patients with NIDDMafter 1 week treatment with d-fenfluramine. Diabetes Care1991; 14 (11): 325-32

42. Willey KA, Molyneaux LM, Overland JE, et al. The effects ofdexfenfluramine on blood glucose control in patients withtype 2 diabetes. Diabet Med 1992; 9: 341-3



43. Stewart GO, Stein GR, Davis TME, et al. Dexfenfluramine intype II diabetes: effect on weight and diabetes control. MedJ Aust 1993; 158: 167-9

44. Stock MJ. Sibutramine: a review of the pharmacology of anovel anti-obesity agent. Int J Obes 1997; 21 Suppl. 1: 25S-9S

45. Lean MEJ. Sibutramine: a review of clinical efficacy. Int J Obes1997; 21 Suppl. 1: 30S-9S

46. Bray GA, Ryan DH, Gordon D, et al. A double-blind random-ized placebo-controlled trial of sibutramine. Obes Res 1996;4 (3): 263-70

47. Weintraub M, Rubio A, Golik A, et al. Sibutramine in weightcontrol: a dose-ranging, efficacy study. Clin Pharmacol Ther1991; 50 (3): 330-7

48. Ryan DH, Bray GA, Wilson JK, et al. Treatment for 21⁄2 yearswith sibutramine [abstract]. Obes Res 1996; 4 Suppl. 1: 31S

49. Guerciolini R. Mode of action of orlistat. Int J Obes 1997; 21Suppl. 3: 12S-23S

50. Drent ML, van der Veen EA. Lipase inhibition: a novel conceptin the treatment of obesity. Int J Obes 1993; 17: 241-4

51. Drent ML, Larsson I, William-Olsson T, et al. Orlistat (Ro 18-0647), a lipase inhibitor, in the treatment of human obesity: amultiple dose study. Int J Obes 1995; 19: 221-6

52. James WPT, Avenell A, Broom J, et al. A one-year trial to assessthe value of orlistat in the management of obesity. Int J Obes1997; 21 Suppl. 3: 24S-30S

53. Pories WJ, Swanson MS, MacDonald KG, et al. Who wouldhave thought it? An operation proves to be the most effectivetherapy for adult-onset diabetes. Ann Surg 1995; 222 (3):339-52

54. MacDonald KG, Long SD, Swanson MS, et al. The gastricbypass operation reduces the progression and mortality onnon-insulin-dependent diabetes mellitus. J Gastrointest Surg1997; 1: 213-20

55. Karlsson J, Sjostrom L, Sullivan M. Swedish obese subjects(SOS): an intervention study of obesity. Two-year follow-upof health-related quality of life (HRQL) and eating behaviorafter gastric surgery for severe obesity. Int J Obes 1998; 22(2): 113-26

56. van Gemert WG, Adang EM, Greve JWM, et al. Quality of lifeassessment of morbidly obese patients: effect of weight-re-ducing surgery. Am J Clin Nutr 1998; 67: 197-201

57. Isacsson Å, Frederiksen SG, Nilsson P, et al. Quality of life aftergastroplasty is normal: a controlled study. Eur J Surg 1997;163: 181-6

58. Mason EE, Renquist KE, Jiang D. Perioperative risks and safetyof surgery for severe obesity. Am J Clin Nutr 1992; 55 Suppl.:573S-6S

59. Martin LF, Tjiauw-Ling T, Horn JR, et al. Comparison of thecosts associated with medical and surgical treatment of obe-sity. Surgery 1995; 118 (4): 599-607

Correspondence and reprints: Dr Jennifer Cerulli, Divisionof Pharmacy Practice, Albany College of Pharmacy, 106New Scotland Ave, Albany, NY 12208, USA.E-mail: [email protected]



Outcomes of Pharmacological and Surgical Treatment for Obesity

Documents

Transcript of Outcomes of Pharmacological and Surgical Treatment for Obesity