This article was downloaded by: [Simon Fraser University]On: 18 November 2014, At: 03:06Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Food Additives & Contaminants: Part APublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tfac20

Determination of diuretics and laxatives as adulterantsin herbal formulations for weight lossAna Paula Lançanova Moreiraa, Monique Jung Mottab, Thaís Ramos Dal Molinb, Carine Vianaab

& Leandro Machado de Carvalhoac

a Graduate Programme in Pharmaceutical Sciences, Federal University of Santa Maria(UFSM), Campus universitário, CEP 97105-900, Santa Maria-RS, Brazilb Center of Health Sciences (CCS), Federal University of Santa Maria (UFSM), Campusuniversitário, CEP 97105-900, Santa Maria-RS, Brazilc Department of Chemistry, Federal University of Santa Maria (UFSM), Santa Maria-RS, BrazilAccepted author version posted online: 03 May 2013.Published online: 20 Jun 2013.

To cite this article: Ana Paula Lançanova Moreira, Monique Jung Motta, Thaís Ramos Dal Molin, Carine Viana & LeandroMachado de Carvalho (2013) Determination of diuretics and laxatives as adulterants in herbal formulations for weight loss,Food Additives & Contaminants: Part A, 30:7, 1230-1237, DOI: 10.1080/19440049.2013.800649

To link to this article: http://dx.doi.org/10.1080/19440049.2013.800649

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Determination of diuretics and laxatives as adulterants in herbal formulations for weight loss

Ana Paula Lançanova Moreiraa, Monique Jung Mottab, Thaís Ramos Dal Molinb, Carine Vianaa,b andLeandro Machado de Carvalhoa,c*aGraduate Programme in Pharmaceutical Sciences, Federal University of Santa Maria (UFSM), Campus universitário, CEP 97105-900,Santa Maria-RS, Brazil; bCenter of Health Sciences (CCS), Federal University of Santa Maria (UFSM), Campus universitário, CEP97105-900, Santa Maria-RS, Brazil; cDepartment of Chemistry, Federal University of Santa Maria (UFSM), Santa Maria-RS, Brazil

(Received 5 February 2013; final version received 25 April 2013)

A new method is described for the determination of the most common diuretic and laxative adulterants found informulations of anorexics and antidepressants. The method is based on the separation of furosemide, hydrochlorothiazide,chlorthalidone and amiloride (diuretics), phenolphthalein (laxative), amfepramone (anorexic) and fluoxetine and paroxetine(antidepressants) by capillary zone electrophoresis with capacitively coupled contactless conductivity detection. The methodshowed a precision ranging from 1.9% to 6.9% for a concentration of 25 mg/L, 0.6% to 5.3% for a concentration of 50 mg/L and 1.6% to 6.0% for a concentration of 100 mg/L for all analytes. The accuracy was 99% for amiloride, 102% forchlorthalidone, 101% for hydrochlorothiazide, 101% for furosemide, 94% for phenolphthalein, 105% for fluoxetine, 114%for paroxetine and 117% for amfepramone. The method allowed the drugs to be determined in the formulations atconcentrations higher than 5.1 mg/kg for amiloride, 7.7 mg/kg for chlorthalidone, 6.8 mg/kg for hydrochlorothiazide,10.7 mg/kg for furosemide, 8.4 mg/kg for phenolphthalein, 11.0 mg/kg for fluoxetine, 9.4 mg/kg for paroxetine and11.0 mg/kg for amfepramone. Three of the 26 analysed herbal formulations were found to be adulterated (not declared onthe label) with the diuretic hydrochlorothiazide. Five other samples contained diuretics declared on the label on theformulation. Thus, a total of eight samples, which were marketed as natural products, contained diuretics (declared ornot) on the formulation.

Keywords: adulterants; herbal formulations; capillary zone electrophoresis; conductivity detection

Introduction

Obesity is a worldwide problem that currently affectsmillions of people and is considered a public health pro-blem. Moreover, it is associated with diseases such asdietary disorders, heart disease, high blood pressure andhypercholesterolemia, which increase the public healthconsequences. A global change in diet and a tendencynot to practice physical exercise are considered the mainfactors that are responsible for the incidence of obesity.

Consequently, there is a growing demand for alternativeweight loss treatments that use natural products. The increas-ing demand for phytotherapeutics can be well understoodmainly due to the false notion that because they are naturalproducts, they do not cause adverse side effects or healthdamage. However, recent studies have demonstrated thepresence of non-declared synthetic substances in the formu-lations of so-called “natural products” that show high effi-cacy in the treatment of obesity (Almeida et al. 2000; Ernst &White 2000; Ernst 2001; Corns & Metcalfe 2002; Ernst2002; Sombra et al. 2005; Bogusz et al. 2006; Liang et al.2006; Al-Safi et al. 2008; Cianchino et al. 2008; Kanan et al.2009; Müller et al. 2009; RIVM 2009; Carvalho et al. 2010;Carvalho, Moreira, et al. 2011; Ernst 2012).

The adulteration of dietary supplements for weightloss makes use of many resources used in the therapeu-tic of obesity, involving basically those substances thatreduce appetite, decrease anxiety and increase intestinalmotility and urine flow. Hence, the most commonclasses of contaminants in these products are appetitesuppressants, antidepressants, anxiolytics, diuretics andlaxatives (Carvalho, Martini, et al. 2011). The latter twocontaminants are intentionally added to give a falsesensation of weight loss, yet these drugs only reducebody water content and not body fat, resulting in dehy-dration and electrolyte imbalance. Furthermore, manycompounds from plants have intrinsic diuretic and laxa-tive properties as well as adverse effects that are similarto those caused by conventional synthetic drugs. Thus,the undeclared addition of synthetic diuretics and laxa-tives in herbal-based medicines and dietary supplementsmay exacerbate impaired health, either because of theirisolated effects or because of interactions with othersubstances present in the formulation. Consideringthese facts, the identification and quantification of anadulterant is rather important, from either a clinical or atoxicological point of view. Thus, the development of

*Corresponding author. Email: lemacarvalho@gmail.com

Food Additives & Contaminants: Part A, 2013Vol. 30, No. 7, 1230–1237, http://dx.doi.org/10.1080/19440049.2013.800649

© 2013 Taylor & Francis

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

analytical methods that allow the detection of possibleadulterants is of great relevance for the quality controlof these formulations. In previous studies (Carvalhoet al. 2010, 2012), we reported the development andapplication of a new method for the determination ofanorexics and antidepressants in herbal-based formula-tions that were commercialised for weight loss purposesin different regions of Brazil. Thus, eight possible adul-terants were selectively identified and quantified byusing the proposed capillary zone electrophoretic(CZE) method with capacitively coupled contactlessconductivity detection (C4D) method in which sibutra-mine and fenproporex were determined to be non-declared adulterants in some analysed formulations.Considering that the association of anorexics withother drug classes is a common practice in adulterationcases worldwide, we extended our studies to develop amethod for the determination of the most commonlyused diuretics and laxatives in these formulations. It isnoteworthy that both of these drug classes are normallyassociated with anorexics in order to improve the effectof the herbal formulations on weight loss.

Capillary electrophoresis (CE) is a commonly used ana-lytical technique among separation techniques that are avail-able for the determination of adulterants inphytotherapeutics. Some advantages of CE, such as thehigh-resolution power, short analysis time and low consump-tion of chemicals and samples, make it an attractive methodfor this type of investigation (Tavares 1997; Silva 2003).Because CE has a high separation capacity, it can be appliedto complex mixtures. This is an important feature for theanalysis of herbal-based formulations that may have not onlynatural constituents but also synthetic organic compounds.However, compared with chromatographic methods, CE hasnot been frequently applied to the analysis of adulterants inthese formulations. In addition, electrochemical detectionmethods have been investigated as a means of overcomingthe limitations of optical detectors. Contactless conductivitydetection is more easily implemented and can be used for theanalysis of inorganic and organic compounds, includingpharmaceuticals (Kubán & Hauser 2008).

Considering the increasing demand for phytotherapeuticformulations and the need for analytical methods for theselective identification and quantification of possible adul-terants, we developed a method based on separation by CZEwith C4D for evaluating diuretics (furosemide, hydrochlor-othiazide, chlorthalidone and amiloride) and laxatives (phe-nolphthalein) in the presence of anorexics (amfepramone)and antidepressants (fluoxetine and paroxetine) that can bepresent as adulterants in herbal formulations commercialisedfor weight loss (Al-Safi et al. 2008; Carvalho, Martini, et al.2011, Carvalho, Moreira, et al. 2011). The present methoduses phosphate buffer (pH 9.2) containing methanol as elec-trolyte additive. Under these conditions, the drugs are mostlyanalysed at their anionic forms, so that the other drug classes

(anxiolytics, anorexics and antidepressants) investigated inour previous work (Carvalho et al. 2010, 2012) did notinterfere in the determination of the studied diuretics andlaxatives. The proposed method was used to evaluate theadulterants in samples of herbal formulations commercia-lised by compounding pharmacies in the Brazilian market.

Materials and methods

Instrumentation and apparatus

The capillary electrophoretic measurements were per-formed on a homemade system equipped with a C4D asdescribed elsewhere (Carvalho et al. 2009, 2010). Theseparations were performed by using an uncoated fused-silica capillary tube of 68 cm × 75 μm innerdiameter × 360 μm outer diameter (Polymicro, Phoenix,AZ, USA) with 15 kV of applied voltage. Indirect detec-tion by contactless conductivity was performed at thecathode side (48 cm from the injection side) by using asinusoidal wave generator operating at 400 kHz frequency.The solutes were injected in the hydrodynamic mode fromthe anodic compartment by pressure (gravity) at a com-partment elevation of 20 cm for 60 s. All experimentswere conducted at 25°C.

Reagents and solutions

All chemicals used in this investigation were of analytical-grade purity. Water was purified by using a Milli-Q UltraPure Water System (Millipore Synergy® UV, Bedford,USA). Amiloride, chlorthalidone, hydrochlorothiazide,furosemide, fluoxetine, paroxetine and amfepramonewere of pharmaceutical grade, as verified by a certificateof analysis. Phenolphthalein and phosphoric acid (85%)were obtained from Merck (Darmstadt, Germany).Methanol and acetonitrile were obtained from Tedia(Fairfield, OH, USA). Sodium hydroxide was obtainedfrom Vetec (Rio de Janeiro, Brazil).

Methanol stock solutions of 1 g/L of diuretics (furo-semide, hydrochlorothiazide, chlorthalidone and amilor-ide), laxatives (phenolphthalein), antidepressants(fluoxetine and paroxetine) and anorexic (amfepramone)were made. Working solutions of 100 mg/L were preparedby diluting these solutions with ultrapure water (Milli-Q).

The working electrolyte solutions were prepared daily andconsisted of 30% (v/v) methanol in 20 mmol/L phosphatebuffer. The pHof theworking electrolyte solutionwas adjustedwith 1.0 mol/L sodium hydroxide solution. The working elec-trolyte solutions were filtered through a 0.45-μm membranefilter (Sartorius, Göttingen, Germany) prior to use.

Samples

The phytotherapeutic formulation samples analysed in thiswork were acquired from pharmacies in different regions

Food Additives & Contaminants: Part A 1231

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

of Brazil and were received by mail. Although therequested natural products were intended for weight lossuse, some samples contained synthetic drugs that weredeclared on the package label, including the diureticsfurosemide and hydrochlorothiazide. The samples hadthe following active ingredients and composition, asdescribed in Table 1.

Analytical procedure

At the beginning of each day, the silica capillary wasrinsed with 0.1 mol/L sodium hydroxide solution for15 min and with water for 15 min before equilibratingwith the working electrolyte solution for 30 min. Aftereach electrophoretic separation, the capillary was rinsedfor 5 min with the working electrolyte solution.

For optimisation of the CE method, working standardsolutions of furosemide, hydrochlorothiazide, chlorthali-done, amiloride, phenolphthalein, fluoxetine, paroxetineand amfepramone were each prepared at a concentrationof 100 mg/L. The following parameters were varied foroptimisation: the working electrolyte composition,

working electrolyte pH, separation potential, operationfrequency of the C4D and sample injection by gravitypressure.

For the CE determination of the adulterants in theformulations, the average weight of 20 capsules wasobtained, and a sample pool was prepared. The equivalentweight of 1 capsule was then dissolved in 25 mL ofmethanol in a volumetric flask before filtration through acotton membrane and then a cellulose acetate membrane(0.45 μm). This solution was then injected into the CEsystem. The identification and quantification of adulterantsin the samples was accomplished by using the standardaddition method (n = 3).

Results and discussion

CZE separation of drug classes as adulterants

The first step of the method optimisation involved usingsolutions (10–50 mmol/L) of phosphate, acetate andborate buffer as working electrolytes. Among the investi-gated buffers, 20 mmol/L phosphate was the chosen work-ing electrolyte solution because it allowed the separation

Table 1. Samples analysed with their respective compositions declared on the label.

A Fucus vesiculosus 80 mg, Centella asiatica 80 mg, Spirulina maxima 80 mg, Passiflora sp. 50 mg, Rhamnus purshiana 100 mg,caffeine 30 mg and hydrochlorothiazide 10 mg

B Cassia augustifolia, Fucus vesiculosus, furosemide, Rhamnus purshiana, Garcinia cambogia and Cyamopsis sp.C Hydrochlorothiazide, Fucus vesiculosus, Rhamnus purshiana, Cassia augustifolia, Centella asiatica, ranitidine, triptophan and

Cordia EcalyculataD Gelidium corneum, vitamin C, vitamin E, Phaseolus vulgaris, Chlorella pyrenoidosa, gelatine and furosemideE Amorphophallus konjac, Cynara scolymus, hydrochlorothiazide, Rhamnus purshiana, Centella asiatica and Ptychopetalum olacoidesF Rhamnus purshiana, Centella asiatica, Cynara scolymus, Baccharis trimera, Fucus vesiculosus, Equisentum sp., Cassia augustifolia,

spirulina maxima and Passiflora sp.G SlendestaTM and Caralluma fimbriataH Caralluma fimbriata 200 mg, Cassia nomame 100 mg, chromium picolinate 10 µg, Camelia sinensis (white tea) 100 mg, Phaseolus

vulgaris 200 mg, Plantago psyllium L. 200 mg, Fucus vesiculosus 100 mg and Equisentum sp. 100 mgI Formulation undeclaredJ Centella asiatica, Cynara scolymus, carboxymethylcellulose, Rhamnus purshiana and Spirulina maximaK Cynara scolymus, Rhamnus purshiana, Garcinia cambogia, Cyamopsis sp., Gymnema sylvestre, Passiflora sp. and Cassia

augustifoliaL Camelia sinensis (red tea)M Gymnema sylvestre, Phaseolus vulgaris, Camelia sinensis (green tea), Citrus aurantium, Caralluma fimbriata and chitosanN Rhamnus purshiana, Spirulina maxima, magnesium stearate, Fucus vesiculosus, carboxymethylcellulose and Centella asiaticaO Caralluma fimbriata, Citrus aurantium and Phaseolus vulgarisP Garcinia cambogia 250 mg, Citrus aurantium 200 mg and chromium picolinate 100 µgQ Caralluma fimbriata, Phaseolus vulgaris, Garcinia cambogia, Gymnema sylvestre and Citrus aurantiumR Garcinia cambogia, Rhamnus purshiana, Fucus vesiculosus, Cynara scolymus, Equisentum sp., Amorphophallus konjac, Cassia

angustifolia, Centella asiatica, Ginkgo biloba L. and Passiflora sp.S Cordia ecalyculata Vell, Garcinia cambogia and Camelia sinensis (green tea)T Citrus aurantium, Fucus vesiculosus, Centella asiatica and Camelia sinensis (green tea)U Formulation undeclaredV Camelia sinensis (green tea) 250 mgW Caralluma fimbriataX Cordia ecalyculata VellY Citrus aurantium, carnitine, Camelia sinensis (green tea) and chitosanZ Cynara scolymus, Centella asiatica, Garcinia cambogia, Equisentum sp., Rhamnus purshiana, Fucus vesiculosus, Spirulina maxima

and Gelidium corneum

1232 A.P.L. Moreira et al.

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

and detection of the largest number of adulterants andyielded a stable baseline in the electropherogram. Theinfluence of methanol and acetonitrile as additives ofphosphate buffer was also investigated at concentrationsbetween 5% and 50% (v/v), and it was observed that thegreater percentages of methanol added to the electrolyteresulted in increased migration times for the drugs. Theopposite effect was observed when adding acetonitrile tothe buffer, which reduced the migration time of the drugs.After choosing 20 mmol/L phosphate buffer and 30% (v/v) methanol as the optimal working electrolyte, a wide pHrange was investigated after considering the variable pKa

values of the drugs. The pH values varied between 2.0 and11.0 and directly influenced the appearance of the electro-phoretic drug peaks, as well as their migration times(Figure 2). Taking into account the pKa values of thedrugs, which ranged from 8.7 to 9.6, we first tested thepossibility of separating the drugs in their protonatedforms. However, the electrophoretic peaks of the drugscould not be observed at pH values from 2.0 to 7.0.Conversely, the peaks for amiloride, furosemide, fluoxe-tine and paroxetine were observed at pH values higherthan 7.0, with the peaks for hydrochlorothiazide, chlortha-lidone and phenolphthalein being observed only at pHvalues higher than 9.0 (Figure 2). Considering that thestudied drugs have several groups that are hydrogen ionacceptors, as observed in their chemical structures inFigure 1, fluoxetine, paroxetine, amiloride and

amfepramone may be still in a protonated form at pHvalues between 8.5 and 10.0 (Figure 2), which mayexplain the observed separation at pH 9.2 in phosphatebuffer with detection at the cathode side of the CE system.However, chlorthalidone, phenolphthalein, hydrochlor-othiazide and furosemide are detected in their anionicforms because they migrate after the electro-osmoticflow (EOF) peak (Figure 3). Furthermore, it is wellknown that the magnitude of the EOF is high above pH8.0. Consequently, the EOF is directed to the cathode, andboth the cationic and anionic drugs can also be carried intothe capillary by the existent EOF towards the cathode.Therefore, pH 9.2 for the phosphate buffer was chosento be optimal for the CZE separations because it permittedthe detection of all the drugs.

In this optimised carrier buffer, separation potentialsvaried in the range 8–20 kV, with the best resolution of thepeaks at 15 kV, in addition to a baseline with little noiseand shorter migration times for all the adulterants. Thefrequency range of 200–1000 kHz was studied, and400 kHz provided the best signal-to-noise ratio for detect-ing the adulterants by indirect contactless conductivity.Finally, considering that the CE equipment used in thiswork contains a system based on hydrodynamic injectionusing gravity pressure, the injection time was varied from100 to 250 mm over 60 to 120 s. Injections made afterelevating the anodic side by 200 mm for 60 s resulted inbetter defined peaks without broadening the bands and

Figure 1. Diuretic, laxative, anorexic and antidepressant drugs studied as adulterants in this work.

Food Additives & Contaminants: Part A 1233

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

maintained good sensitivity. Figure 3 shows the electro-pherogram for the separation of the eight studied adulter-ants by the proposed method under all of theaforementioned optimised conditions.

Figures of merit

The optimised method was validated by considering themajor analytical validation parameters for the studied

adulterants. The LODs and LOQs were calculated fromthe 3σ and 10σ values, respectively. The standard devia-tion was an average of seven measurements of the back-ground noise divided by the slope of the respectivecalibration function obtained for each adulterant. The pre-cision was expressed by the variation coefficients(expressed as RSD) of the results obtained in triplicatefor three different concentrations of each analyte. For theaccuracy calculation, the standard addition method (n = 3)was used, where three different concentrations of thestandard solution were added to the sample prior to theextraction and filtration process. The results obtained forthe linear range, LODs/LOQs, precision and accuracy areshown in Table 2.

Considering that the adulteration of natural productsintended to be used for weight loss can involve many differ-ent drug classes, other compounds were tested as possibleinterferents in the optimised method. It is important to notethat the initial purpose of this work was to develop a methodfor analysing diuretics (furosemide, hydrochlorothiazide,amiloride and chlorthalidone) and laxatives (phenolphtha-lein) as adulterants. However, after adding other drug classessuch as anorectics (amfepramone, fenproporex and sibutra-mine), antidepressants (fluoxetine, paroxetine, sertraline andbupropion) and anxiolytics (diazepam, alprazolam, fluraze-pam, chlordiazepoxide, lorazepam, bromazepam and clona-zepam) as possible interferences in the analysis,amfepramone, fluoxetine and paroxetine were detectedalong with diuretics and laxatives and their presence didnot interfere with the CE peaks. The other drugs studied aspossible interferents were not detected under the conditionspreviously optimised, most likely because at pH 9.2 they are

Figure 2. Effect of pH of the working electrolyte on the migration times of the synthetic drugs. Working electrolyte: H3PO4 20 mmol/L:NaOH 40 mmol/L: 30% methanol (v/v); potential of separation: –15 kV; conductivity detection (C4D) operating at 400 kHz and 2 Vpp;temperature: 25°C; sample injection by gravity: 20 cm for 60 s.

Figure 3. Electroferogram of the adulterants (126.58 mg/kgeach): (a) fluoxetine, (b) paroxetine, (c) amiloride, (d) amfepra-mone, (e) chlorthalidone, (f) phenolphthalein, (g) hydrochlor-othiazide and (h) furosemide. Working electrolyte: H3PO4

20 mmol/L, NaOH 40 mmol/L1 and 30% methanol (v/v) at pH9.2; other conditions as described in Figure 2. EOF, electro-osmotic flow.

1234 A.P.L. Moreira et al.

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

not well ionised and/or co-migrate with the EOF in theirneutral form. Other possible organic interferences that couldbe extracted from the plant structure in methanol during thesolubilisation process did not interfere in the electrophero-grams obtained for the studied drugs probably because of thesame reason.

Analytical application in dietary supplements

The proposed method was used for the analysis of 26samples of herbal-based supplements acquired from com-mercial compounding pharmacies in the Brazilian market.For each sample, an analytical screening was first carriedout to identify the possible adulteration of the sampleswith any of the studied drugs. Samples with electrophore-tic peaks compatible with the migration times of thestudied adulterants were confirmed by adding the refer-ence standards. In the samples that were confirmed to beadulterated, the standard addition method was applied tocalculate the concentration of the identified adulterant.Figure 4 shows the electropherograms obtained from two

different samples containing furosemide and hydrochlor-othiazide as the declared components on the product label.In addition to the five samples that declared diuretics onthe formulation label, three other samples were identifiedas adulterated with hydrochlorothiazide (not declared onthe label). The samples analysed for adulterants (declaredor not) are shown in Table 3. As shown in Table 3, theconcentration found for hydrochlorothiazide in sample A(22.2 mg/g) was rather accurate considering the 20.0 mg/gamount declared by the compounding pharmacy. Theother concentrations found for furosemide and hydrochlor-othiazide were not declared by the manufacturers.Regardless of whether the presence of diuretics wasdeclared on the label of the formulations, the recom-mended doses of three to four capsules per day wouldresult in daily minimal consumptions of 47–86 mg furo-semide and 1.8–67.5 mg hydrochlorothiazide, accordingto the concentrations described in Table 3. Considering therecommended doses of furosemide (20–40 mg) and hydro-chlorothiazide (12.5–25 mg), it is evident that the amountsof these diuretics in the formulations are, in some cases,

Table 2. Figures of merit of the CZE method developed for the determination of drugs in herbal formulations.

Adulterants (mg/L) LOD (mg/kg) LOQ (mg/kg) Linear range (mg/kg) r Precision (RSD) Accuracy (%)

Fluoxetine 3.30 11.01 10–500 0.999725.0 4.85 104.6050.0 1.10

100.0 3.64Paroxetine 2.81 9.37 10–500 0.9965

25.0 6.96 114.2050.0 1.04

100.0 6.00Amiloride 1.54 5.14 10–500 0.9955

25.0 3.02 98.6050.0 0.62

100.0 1.63Amfepramone 3.29 10.96 10–500 0.9993

25.0 3.50 116.8050.0 2.52

100.0 3.26Chlorthalidone 2.32 7.72 10–500 0.9995

25.0 2.33 101.8050.0 5.30

100.0 2.50Phenolphthalein 2.48 8.42 10–500 0.9984

25.0 1.93 93.6050.0 4.47

100.0 2.81Hydrochlorothiazide 2.03 6.78 10–500 0.9987

25.0 2.21 101.0050.0 2.65

100.0 1.87Furosemide 3.19 10.66 10–500 0.9994

25.0 6.67 101.0050.0 4.68

100.0 3.23

Note: CZE, capillary zone electrophoresis.

Food Additives & Contaminants: Part A 1235

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

below the recommended doses. However, in some formu-lations, a significant excess of diuretic is being ingested bythe users, for example, 67.5 mg of hydrochlorothiazideand 86 mg of furosemide. This disharmonious consump-tion is strongly related to the uneven distribution of theadulterant drugs in the capsules during the manipulationprocess of the supplements. According to the currentpractices of manipulation of pharmaceutical formulations,there is always an error of dose expected when preparingherbal-based capsules by compounding pharmacies.

Conclusions

Adulteration is only one of the problems linked to the poorcontrol of herbal products commercialised as dietary

supplements. Thus, more effective means for increasingand improving control over the production and marketingof natural products are needed. This is possible with theavailability of analytical methods that are capable of iden-tifying the presence of prohibited substances in herbal-based medicines as well as worldwide standardisation ofthe laws applied in the production, commercialisation anduse of phytotherapeutics.

Considering the LOQs achieved for the adulterants,the method is able to quantify the concentrations from5.1 mg/kg for amiloride, 7.7 mg/kg for chlorthalidone,6.8 mg/kg for hydrochlorothiazide, 10.7 mg/kg for furo-semide, 8.4 mg/kg for phenolphthalein, 11.0 mg/kg forfluoxetine, 9.4 mg/kg for paroxetine and 11.0 mg/kg foramfepramone. These LOQs for the adulterants are much

Figure 4. Electropherograms of samples screened by the method containing synthetic drugs declared on the formulation label. (a)Sample A containing hydrochlorothiazide. (b) Sample B containing furosemide; other conditions as described in Figure 2. EOF, electro-osmotic flow.

Table 3. Synthetic drugs found as labelled or no labelled component in the studied herbal formulations.

Sample Labelled sample composition Synthetic drugDetermined

contenta (mg/g)

A Fucus vesiculosus 80 mg, Centella asiatica 80 mg, Spirulina maxima 80 mg,Passiflora sp. 50 mg, Rhamnus purshiana 100 mg, caffeine 30 mg andhydrochlorothiazide 10 mg

Hydrochlorothiazide 22.2

B Cassia augustifolia, Fucus vesiculosus, furosemide, Rhamnus purshiana, Garciniacambogia and Cyamopsis sp.

Furosemide 31.6

C Hydrochlorothiazide, Fucus vesiculosus, Rhamnus purshiana, Cassia augustifolia,Centella asiatica, ranitidine, triptophan and Cordia Ecalyculata

Hydrochlorothiazide 45.0

D Gelidium corneum, vitamin C, vitamin E, Phaseolus vulgaris, Chlorella pyrenoidosa,gelatine and furosemide

Furosemide 57.4

E Amorphophallus konjac, Cynara scolymus, hydrochlorothiazide, Rhamnus purshiana,Centella asiatica and Ptychopetalum olacoides

Hydrochlorothiazide 2.4

T Citrus aurantium, Fucus vesiculosus, Centella asiatica and Camelia sinensis(green tea)

Hydrochlorothiazide 4.2

X Cordia ecalyculata Vell Hydrochlorothiazide 1.2Y Citrus aurantium, carnitine, Camelia sinensis (green tea) and chitosan Hydrochlorothiazide 14.0

Note: aRSD (n = 3): 2.0–5.2%.

1236 A.P.L. Moreira et al.

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014

lower than the minimal daily doses recommended forthese drugs and permit the determination of the adulterantseven at concentrations much lower than the prescribedtherapeutic doses.

Three of the 26 samples studied in this work werefound to be adulterated (not declared on the label) withthe diuretic hydrochlorothiazide. This drug has a lowincidence of side effects, but it is associated with severalmetabolic complications when administered in high doses.This is precisely the problem of adulterated products, asconsumers are not aware of the qualitative and quantitativecontent of the formulation. Adulterated formulations andformulations that are called “natural” but contain syntheticdiuretics that are declared on the product’s label are inef-fectively regulated. Among the 26 analysed formulations,a total of 8 samples (30.77%) were found to containdiuretics in the final composition (declared or not).

AcknowledgementsThe authors express their grateful thanks to CAPES (PROCADno. 098/2007), CNPq and FAPERGS for the financial support ofthis work.

ReferencesAlmeida AE, Ribeiro ML, Polese L. 2000. Determination of

amfepramone hydrochloride, fenproporex and diazepam inso-called “natural” capsules used in the treatment of obesity.J Liq Chrom Relat Technol. 23:1109–1118.

Al-Safi SA, Ayoub NM, Ayoub AM, Al-Momany E, Al-DoghimI, Al-Balas M, Alkofahi AS, Aboul-Enein FH, Aboul-EneinBH. 2008. Public awareness of the abuse of herbs and drugsto decrease body weight: a novel survey in Jordan. Am JPublic Health. 16:205–213.

Bogusz MJ, Hassan H, Al-Enazi E, Ibrahim Z, AL-tufail M.2006. Application of LC-ESI-MS-MS for detection of syn-thetic adulterants in herbal remedies. J Pharm Biomed Anal.41:554–564.

Carvalho LM, Cohen P, Silva CV, Moreira APL, Falcão T, DalMolin T, Zemolin G, Martini M. 2012. A new approach todetermining pharmacologic adulteration of herbal weightloss products. Food Addit Contam: Part A. 29:1661–1667.

Carvalho LM, Martini M, Moreira APL, Garcia SC, NascimentoPC, Bohrer D. 2010. Determination of synthetic pharmaceu-ticals in phytotherapeutics by capillary zone electrophoresiswith contactless conductivity detection (CZE-C4D).Microchem J. 96:114–119.

Carvalho LM, Martini M, Moreira APL, Lima APS, Correia D,Falcão T, Garcia SC, Bairros AV, Nascimento PC, Bohrer D.2011. Presence of synthetic pharmaceuticals as adulterants in

slimming phytotherapeutic formulations and their analyticaldetermination. Forensic Sci Int. 204:6–12.

Carvalho LM, Moreira APL, Martini M, Falcão T. 2011. Theillegal use of synthetic pharmaceuticals in herbal formula-tions: an overview of adulteration practices and analyticalinvestigations. Forensic Sci Rev. 23:73–90.

Carvalho LM, Spengler C, de Lima APS, Raabe A, Martini M,do Nascimento PC, Bohrer D. 2009. Determinaçãosimultânea de cátions empregando eletroforese capilar comdetecção sem contato em equipamento construído emlaboratório. Quim Nova. 32:2203–2208.

Cianchino V, Acosta G, Ortega C, Martínez LD, Gómez MR.2008. Analysis of potential adulteration in herbal medicinesand dietary supplements for the weight control by capillaryelectrophoresis. Food Chem. 108:1075–1081.

Corns C, Metcalfe KJ. 2002. Risks associated with herbal slim-ming remedies. J Roy Soc Promot Health. 122:213–219.

Ernst E. 2001. Toxic heavy metals and undeclared drugs in Asianherbal medicines. Trends Pharmacol Sci. 23:136–139.

Ernst E. 2002. Adulteration of Chinese herbal medicines withsynthetic drugs: a systematic review. Eur J Intern Med.252:107–113.

Ernst E. 2012. Risks associated with complementary therapies.In: Dukes MNG, Aronson JK, editors. Meyler’s side effectsof drugs. Amsterdam: Elsevier; p. 1649–1681.

Ernst E, White A. 2000. The BBC survey of complementarymedicine use in the UK. Complement Ther Med. 8:32–36.

Kanan S, Abu-Yousef IA, Gunasekar C, Abd N, Narasimhan S.2009. Detection and quantification of synthetic drugs inherbal slimming formula. Eur J Sci Res. 34:348–357.

Kubán P, Hauser PC. 2008. A review of the recent achievementsin capacitively coupled contactless conductivity detection.Anal Chim Acta. 607:15–29.

Liang Q, Qu J, Luo G, Wang YW. 2006. Rapid and reliabledetermination of illegal adulterant in herbal medicines anddietary supplements by LC/MS/MS. J Pharm Biomed Anal.40:305–311.

Müller D, Weinmann W, Hermanns-Clausen M. 2009.Chinese slimming capsules containing sibutramine soldover the internet – a case series. Dtsch Arztebl Int.106:218–222.

RIVM Report 370030002. 2009. Trends in drug substances detectedin illegal weight-loss medicines and dietary supplements. A2002–2007 survey and health risk analysis [Internet]. [cited2013 June 12]. Available from: http://www.rivm.nl/bibliotheek/rapporten/370030002.pdf

Silva JAF. 2003. Detecção em eletroforese capilar. Quim Nova.26:56–64.

Sombra LL, Gómez MR, Olsina R, Martínez LDM, Silva MF.2005. Comparative study between capillary electrophoresisand high performance liquid chromatography in “guarana”based phytopharmaceuticals. J Pharm Biomed Anal.36:989–994.

Tavares MFM. 1997. Mecanismos de separação em eletroforesecapilar. Quim Nova. 20:493–510.

Food Additives & Contaminants: Part A 1237

Dow

nloa

ded

by [

Sim

on F

rase

r U

nive

rsity

] at

03:

06 1

8 N

ovem

ber

2014