Quality characteristics and HPLC detection of phosphodiesterase inhibitors in some herbal capsules indicated for male sexual disorders in Ghana

Background Herbal medicines used for male sexual disorders are widespread across the globe and have been noted as likely candidates for adulteration. To assure access to safe and quality herbal products, this study aimed at sampling some herbal capsules indicated for male vitality for quality analytical checks. Methods Herbal capsules sampled from two regions were subjected to physicochemical assessment, using pharma-copeial and regulatory standards. Microbial quality analysis was also performed on the selected products. A validated HPLC method for simultaneous analysis of sildenafil citrate and tadalafil was used to detect and quantify the possible presence of phosphodiesterase inhibitors in the selected products. Results Out of a total of 57 herbal capsules, 19 were indicated for male sexual related illnesses of which only 8 fulfilled the inclusion criteria of being used exclusively for male vitality. The majority (62.5%) of the herbal products failed the weight uniformity test but all of them disintegrated within the acceptable limit of 30 min. The moisture and heavy metal contents were within limits of acceptability. Majority (87.5%) of the herbal products failed the test for the aerobic bacteria count; however, none of the isolated bacteria was pathogenic. Sildenafil citrate was detected in the majority (87.5%) of the herbal products in concentrations of 0.98–33.95 mg/g. Two products contained both sildenafil and tadalafil. High batch-to-batch variability was recorded in some physicochemical parameters and pharmaceutical adulterants. Conclusion High incidence of quality fails and adulteration of herbal capsules for sexual disorders observed means there is a real risk of unintended dosing with phosphodiesterase inhibitors. This represents a public health issue and necessitates tighter monitoring of production standards by the regulator.


Background
According to the World Health Organization (WHO), herbal medicines are medicinal preparations made with plant materials including the leaves, stems, flowers, roots and seeds which may be fragmented or cut and used alone or in combinations [1].The practice of herbal medicine is the oldest form of health care which has been used for decades in developing and developed countries.Today, up to 65% of the people who live in developed countries utilize herbal medicines as complementary and alternative medicine (CAM), whereas 80% of the population in low-and middle-income countries heavily rely on them as their main source of healthcare products [2].In Ghana, 70% of patients prefer herbal remedies over conventional therapy for various reasons including affordability, cultural acceptance, accessibility, and availability [3].Global acceptability and demand for herbal medicinal products have been on the ascendency in recent times.In 2020, the global share of the herbal medicine market was USD 185.86 billion and is projected to increase to USD 430.05 billion at a compound annual growth rate (CAGR) of 11.32% by the end of 2028 [4].The driving force propelling the increase in use of herbal medicines indicated for sexual dysfunction is attributed to a high prevalence of male sexual disorders.The Ghanaian market is similarly flooded with herbal aphrodisiacs, which are substances used to enhance sexual desire or performance [5].
There is this notion among some herbal medicine producers that plant-based products are natural with little or no toxicity or side effects and therefore require less rigorous and comprehensive quality, safety, and efficacy evaluations.Poor regulatory measures, adulterations, lack of adherence to good manufacturing practices, herb-drug, and herb-food interactions may also bring about adverse events in their use [6].
Standardization and quality control are a major weakness of the herbal medicine industry [7].These underpin the usual high batch-to-batch variation observed in the same herbal medicinal product.High batch-to-batch quality variations in medicinal products implies that consumers are not guaranteed of same benefit from the use of the same product.Standardization and quality control activity covers the harvesting and processing of raw materials; correct identification of source plant species; the selection of appropriate parts used in the production of herbal medicines; the methods and production process of the herbal medicinal products.
In Ghana, herbal medicines and related products are not expected to be introduced onto the market without any mandatory evaluation and approval by the Food and Drugs Authority (FDA).Prior to granting marketing authorization, the FDA inspects the manufacturing premises to ensure compliance with current codes of good manufacturing practices (cGMP) in addition to evaluation of pharmacological, toxicological, and clinical trial data on the herbal medicinal products.Despite this rigor, substandard and falsified herbal products still exist on the market.
Herbal medicines claiming aphrodisiac properties have been noted as a vulnerable therapeutic area for adulteration [8] as pieces of evidence have shown their adulteration with phosphodiesterase (PDE) inhibitors including sildenafil citrate, tadalafil, vardenafil and their unapproved analogs [9][10][11].PDEs are a class of medications used in the treatment of erectile dysfunction and other conditions such as pulmonary arterial hypertension.Sildenafil and tadalafil are synthetic medications which dilate blood vessels in the penis and enhance erection.The extent of usage and adulteration as well as the quality characteristics of these herbal aphrodisiacs need to be assessed to provide an assurance of the safety and efficacy of the products on the Ghanaian market.
The presence of a plethora of substances in herbal products implies that selective and sensitive analytical methods such as high-performance liquid chromatography (HPLC) will have to be employed to detect the presence of the adulterants.This ensures the separation, identification and quantification of the adulterants in a one-step analysis.In this study, an already developed and validated HPLC method was used to detect adulteration.The project represents the first of its kind in Ghana, assessing both the physicochemical qualities of male sexual enhancing herbal supplements and the simultaneous detection of sildenafil and tadalafil as possible adulterants.

Research design and study area
The study was structured in two parts: a market survey to identify and sample herbal capsules sold for male sexual disorders in the Bono Region of Ghana, and laboratory analysis of the product quality, including possible adulteration with PDE inhibitors.The Bono region was selected because its market is accessible to people from all districts within three other regions of Ghana, namely Bono East, Ahafo and Ashanti [12].Due to Ghana's welldeveloped distribution and retailing sector, we believe that herbal products found in this region are very likely to be encountered on the markets of other regions.To measure quality variations in the herbal products sold on the market, one batch of herbal products were purchased from the Bono region (designated group A products) and another batch from the Ashanti region (designated group B products).

Market survey
A cross-sectional study was conducted to obtain information on herbal capsules offered for sale on the market between January and February 2021.Wholesale and/or retail sales outlets where herbal medicinal products were displayed and offered for sale were purposively sampled.They were stratified into pharmacies (P), over-the-counter chemical sellers (OTCMs) and herbal medicinal products sellers (HMPs) sale outlets.The snowball sampling technique was used to scout for brands of male sexual disorder herbal capsules on the market.This method is a form of non-probability sampling which is useful when the population under study is difficult to reach [13].In a situation where previously identified products were found; interviewees were asked to name a neighboring outlet where other brands may be found.

Sample size determination
At a confidence level of 80%, the Epi Info StatCalc Sample Size Calculator [14] was used to determine the sample size (n) of participants to include in the survey from a population of 86 pharmacies, 546 OTMCs and 30 HMPs scattered in the Bono region.The sample size (n), using the Dobson's formula, was computed using Eq.1: Sample size calculator formula: where n = required sample size; p = guestimate proportion of herbal capsules sold on the Ghanaian market = 50%; z = level of confidence at 80% = 1.28; d = desired degree of precision or acceptable margin of error = 10%.A total of 83 sale outlets sample size comprising of 28 pharmacy shops, 38 OTC, and 17 HMPs were computed as the sample size (n T) to be used for the survey.However, due to resource constraints, a total of 37 outlets including 12 pharmacies, 15 OTC, and 10 HMPs were visited for the survey (limitation).Source of data for the distribution of the pharmacies, OTCMS and herbal shops were obtained from the Food and Drugs Authority, Sunyani-Bono Region [15].Due to ethical reasons, selected product samples were coded as HC-001, HC-008, HC-021, HC-027, HC-034, HC-037, HC-039, and HC-041; where HC referred to "herbal capsule" and the three-digit numbers were arbitrary numbers (Supplementary Tables 1S and 2S).

Ethical considerations
This article does not contain any studies involving animals or human participants.Instruments used to collect data about the herbal products sold on the market were approved by the research team.Verbal permission was sought from owners and contact persons of outlets before data extraction.All information about herbal products obtained from the points of sale or the contact persons was treated confidentially.Personal identifiers such as brand names and manufacturer's details were coded with unique identifiers which were used during analysis and reporting.Data extracted from product labels were stored on the researcher's computer under a password to prevent unauthorized access.

Determination of physicochemical properties and adulteration
Important pharmacopoeial and non-pharmacopoeial quality parameters were determined.These include capsule weight variation and disintegration, moisture content, pH, toxic metal content, microbial load and presence of adulterants.All of these contribute to the overall quality, safety and efficacy of the medicinal products.

Weight variation
This was based on the uniformity of mass for single-dose preparations (Sect.5.2) from the International Pharmacopoeia [16].Briefly, using the analytical balance, the individual gross weight of some randomly selected 20 capsules was taken, after which the net content weight of each capsule was determined by subtracting the weight of shells after emptying its content.The average content weight for the 20 capsules was then calculated and compared with the content weight of each individual capsule previously determined.

Capsule disintegration
This was based on the disintegration test for tablets and capsules (Sect.5.3) from the International Pharmacopoeia [17].One dosage unit of six randomly selected capsules was placed in each of the six tubes of the basket rack assembly suspended in the immersion fluid of the disintegration apparatus.The apparatus was then operated until complete disintegration where only fragments of insoluble capsule shell remained on the screen of the test apparatus.The time for complete disintegration of the capsule was noted and reported.According to the International Pharmacopoeia, the recommended time for the disintegration of the capsule dosage unit is 30 min.

Moisture content
The loss on drying method based on the Articles of Botanical Origin (561) in the USP was used [18].The contents of four (4) randomly selected capsules were emptied and thoroughly mixed.About 1.0 g of test sample was weighed onto a tarred porcelain dish to obtain the initial weight of the test sample.The test sample was then placed in a hot air oven to dry at a temperature of 105 °C.Drying and weighing were done at one hourly interval until a constant weight was obtained.The dried sample and porcelain dish were weighed to obtain weight of the dried plant material.The moisture content of the product was then calculated by dividing the difference in sample weight before and after drying (weight loss) by the initial weight of the test sample.

pH
The pH of the samples was measured based on the determination of pH (Sect.1.13) of the International Pharmacopoeia [19].The contents of about 3-4 herbal capsules were emptied into a dish and mixed thoroughly.An amount of 1.0 g sample was accurately weighed into a beaker containing 100 mL distilled water The mixture was thoroughly stirred and allowed to stand for 30 min at room temperature of 25 °C ± 0.02 for stabilization after which the sensor was immersed and the pH value read.This was done in triplicate for each of the batch's A and B.

Toxic metal analysis (As, Cd, Pb, Hg)
The weight of about 3-4 herbal capsules, equivalent to about 1.00 g, was taken and dropped into the Kjeldahl digestion tube.This was then digested with 30.0 mL hydrochloric acid and 10.0 mL nitric acid at 450 °C for about 60 min.The digestate was allowed to cool and then filtered through filter paper (Watman paper number 3) into a 100-mL volumetric flask and made up to the 100-mL mark with freshly prepared distilled.This was analyzed in the atomic absorption spectrophotometer (air pressure = 50-60 psi, acetylene pressure = 10-15 psi, and voltage = 208-240 V).Calibration curves were plotted for each heavy metal (As, Cd, Pb, Hg) to be analyzed using the stock standard elements solution covering six concentrations in the ranges of 0.00-30.00mg/L for As; 0.00-1.02mg/L for Cd; 0.00-18.00mg/L for Pb and 0.00-60.00mg/L for Hg.The unknown elemental concentration of the test solution was read using the various metals at the specified wavelengths of 193.7 nm, 228.9 nm, 217.0 nm and 253.7 nm for As, Cd, Pb, and Hg, respectively.The heavy metal content in the product samples was then calculated using Eq. 2 as follows: here nominal volume = 100 mL, sample weight = weight (g) of herbal capsules digested.Limits for heavy metals specified in the Association of Southeast Asia Nations (ASEAN) guidelines on limits of contaminants for traditional medicines were then applied [20]. (2)

Microbial load analysis
Test procedures described in the European Pharmacopoeia microbial examination of nonsterile pharmaceutical preparations and substances for medicinal use were followed for microbial quality analysis [21].Briefly, the contents of 3-4 herbal capsules were aseptically emptied into a disinfected (sterile) dish and mixed thoroughly from which 1.00 g sample was taken and dissolved in 9 mL normal saline contained in previously sterilized tubes to obtain a stock solution.The stock solution was labeled as 1-in-10 dilution (10 -1 ).Using a sterile micropipette, 1.0 mL of stock solution was measured and diluted in another 9.0 mL of normal saline contained in a sterile test tube labeled as 1-in-100 dilution (10 -2 ).The stock solutions were further diluted tenfold down to 10 -6 .Using a sterile micropipette and the procedure for the pour plate technique, 1.0 mL of each solution was transferred into their respective labeled petri dishes, after which 10.0 mL molten casein soya bean digest agar cooled to 45 °C was poured into the set of dishes for enumeration of total aerobic bacteria.The same volume and condition of molten Sabouraud-dextrose agar were also poured into the other sets of labeled petri dishes for total combined yeasts and molds enumeration.The plates were gently swirled in a circular motion to achieve uniform distribution of microorganisms.The agar was allowed to solidify after which the plates were inverted and incubated at 37 °C for 48 h for growth to occur.Quebec colony counter was used to enumerate colonies.The arithmetic mean per culture medium of the counts was computed and the number of colony-forming units (CFU) per mL or gram of product sample was then calculated.The total aerobic microbial count (TAMC) was equal to the number of CFU found on casein soya bean digest agar, whereas the total combined yeasts and mold count (TYMC) was equal to the number of CFU found using Sabouraud-dextrose agar medium.The colonyforming unit per mL (cfu/mL) of colony counts was determined using Eq. 3. The dilution factor was calculated from Eq. 4: Qualitative test for the presence or absence of bile tolerant gram-negative enterobacterium was done using casein soya bean digest broth as diluent inoculated with 1.0 g of product sample and incubated at 37 °C for 24 h.Ten (10)

High pressure liquid chromatographic (HPLC) analysis for PDE inhibitors
HPLC analytical method for mixed sildenafil citrate and tadalafil as reported by Al-Tahami [9] was employed in this study.The HPLC method was verified according to the ICH guideline Q2R1 on validation of analytical procedures [22].The HPLC gradient elution technique was employed for the analysis.The mobile phase composed of HPLC grade methanol (A) and 0.05% trifluoroacetic acid (B) solvents.The elution process was as follows; 0-3 min: isocratic with 90.0%A; 3-23 min: linear from 90%A to 10.0%A; 23-28 min: isocratic with 10.0%A; 28-28.1 min: linear from 10%A to 90.0%A; 28.1-33.1 min: isocratic with 90.0%AThe flow rate of the mobile phase was set to 1.0 mL/ min, and injection volume of 10 µL was used.
The HPLC photodiode array (PDA) detector at a wavelength of 285 nm was used to monitor, detect, and quantify both sildenafil citrate and tadalafil in the product samples.Stock solutions of sildenafil citrate and tadalafil reference standards were prepared by dissolving 10 mg powder in 10 mL methanol with sonication for 10 min and then filtered using 0.45-µm membrane filters for analysis.Six (6) different concentrations (0.5-50 mg/L) of the working standards sildenafil citrate and tadalafil were prepared for calibration.Stock reference solutions were then subjected to UV-VIS spectroscopy using HPLC grade methanol solvent as reference scanned from 200 to 400 nm.A calibration curve was obtained from the peak areas obtained from running the six calibration solutions of the sildenafil and tadalafil mixture.Verification was conducted in respect of selectivity, precision, linearity, the limit of detection (LOD), and the limit of quantification (LOQ).This is because the original method was already validated.Selectivity was determined from the HPLC chromatograms showing well resolved peaks.Linearity was expressed as regression coefficients (R 2 ) and was measured by analyzing working standards of sildenafil citrate and tadalafil at six concentration levels in the range of 0.5-50 mg/L and plotting peak areas against concentration.Precision (repeatability) was assessed using 3 replicates of 3 concentrations within the linear range.The relative standard deviation (RSD) was subsequently determined.The limits of detection and quantification for Sildenafil citrate and Tadalafil were estimated by using data from the linear regression equations of the calibrated curves.LOD and LOQ were calculated using Eq. 5 and 6, respectively: where SD = standard deviation of the intercepts of the regression lines.
To analyze the herbal capsules for the presence of sildenafil and tadalafil, the contents of four (4) herbal capsules were then emptied into a dish and thoroughly mixed.A hundred milligrams (100 mg) of the powder was accurately weighed and transferred into a 10-mL volumetric flask and topped up with a solvent mixture of 1:1 methanol (96%) and water with sonication for 10 min.The mixture was then filtered through a 0.45 µm nylon membrane filter into 2-mL HPLC vials for injection and analyzed under the stated chromatographic conditions.The final concentration of PDE inhibitor in herbal products was then calculated by considering the dilution factor, amount of sampling, and the average weight of capsule content.Product samples with chromatograms containing peak retention times corresponding to sildenafil citrate and tadalafil were noted as herbal capsules containing pharmaceutical adulterants.

Data and statistical analysis
Data obtained from experiments were reported as means ± standard deviations (SD) and standard errors of means (SEM) of three replicates of each test.Twoway ANOVA was used to estimate how the mean of the dependent quantitative output variable changed according to the levels of two independent categorical variables, ie, the herbal products and the product groups (group A and group B).The ANOVA was followed with a multiple comparison test to check for consistency in the mean difference of quantitative variables observed between Group A and Group B variables.One sample t-test was then employed to compare the sample mean to acceptable quality specifications or standards.GraphPad prism (version 8 for windows, San Diago, USA) was used for the statistical analysis.P values less than 0.05 were considered statistically significant.

Market survey and product sampling
A total of nineteen (19) herbal capsules indicated for use in male sexually related illnesses such as "delayed release of sperms, " "low libido, " "low virility, " or "low male vitality" were found on the market (Supplementary data, Table 1S).Most of these herbal capsules were also indicated for use in other illnesses that did not relate to sexual disorders.However, only eight (8) satisfied the inclusion and exclusion criteria applied (Supplementary data, Table 2S).

Physicochemical analysis Weight variation of herbal capsule
The net content by weight of the herbal products varied between 0.1955 ± 0.024 and 0.4341 ± 0.034 g.The products HC-034, HC-037, and HC-041 passed the weight variation test (Table 1).The products HC-001, HC-008, and HC-027 with an average net content weight of less than 300 mg failed the weight variation test because the content of less than 18 individual capsules deviated within ± 10% of the product's average content weights (Table 1).A batchto-batch analysis of the weight variation compliance rate showed consistency in all except the product HC-021 (Table 1).Batch A of HC-021 complied with acceptance criteria, but batch B failed because the contents weights of a total of 15 individual dosage units deviated within ± 7.5% of the average content weight of 0.3575 g (Table 1).This falls short of a minimum of 18 dosage units required by the International Pharmacopoeia [29].

Capsule disintegration time
All herbal capsules analyzed fragmented between 5.06 ± 0.007 and 19.43 ± 0.035 min.This was within the disintegration time of 30 min recommended by the International Pharmacopoeia.A two-way ANOVA was used to analyze the effect of herbal product batches on the disintegration time revealed significant interaction between the disintegration time and the batch (F (7,16) = 4077, P < 0.0001).A significant difference (P < 0.05) in disintegration time for batch A and batch B was observed for the majority (6/8; 75.00%) of the products (Table 2).

Moisture content and pH of herbal capsules
The results of the moisture content and pH obtained in this study are shown in Table 2.The moisture content of the products was in the range of 3.11% and 6.67%.In the pH measurement, aqueous solution of the herbal capsules recorded values in the range of 4.76 ± 0.085 to 6.39 ± 0.187.No statistically significant difference was observed in the pH and moisture content of the two batches (Table 2).

Heavy metal content analysis
Different concentrations of arsenic, cadmium, lead and mercury were found in each of the eight herbal capsules analyzed (Table 3).The concentrations were within acceptable limits prescribed by the ASEAN limits of heavy metal contaminants in all the herbal medicinal products (Table 3).In these herbal products, significant pairwise differences (P < 0.05) were seen between the mean concentrations of arsenic, cadmium, lead, and mercury (Supplementary data, Table 3S).

Microbiological load of herbal products
Bacterial and fungal colonies were found in all (8/8; 100%) herbal capsules analyzed.The total aerobic microbial colonies (TAMC) counted varied between 7.03 × 10 2 cfu/mL and 2.55 × 10 6 cfu/mL.Almost all the herbal products failed to comply with the TAMC for non-aqueous oral dosage forms.None of the bacteria  were pathogenic.Total yeast and mold colonies counted in all the herbal products were within acceptable limits of the European Pharmacopoeia (Table 4).

Method validation
HPLC chromatograms in Figs. 1, 2, and 3 shows clearly resolved peaks confirming the selectivity of the method.The precision (intraday) of the method was expressed in the RSD of three concentrations in the linear range.For sildenafil citrate, these were 0.23, 0.55 and 0.39% for concentrations of 5, 10 and 25 mg/L (Table 5S).For tadalafil, these were 0.98, 0.66 and 0.31%, respectively (Table 6S).The values obtained were all less than the upper limit of 2% stated by the ICH.This confirmed the precision of the method.The detection limits (LOD) of sildenafil citrate and tadalafil at a concentration range of 0.5-50.0mg/L were found to be 0.081 mg/L and 0.134 mg/L, respectively, whereas their limits of quantification (LOQ) were 0.243 mg/L and 0.403 mg/L correspondingly (Table 4S).The calibration curves of both standards had lines of best fit (R 2 ) above 0.99 (Fig. 1S).

Table 4 Microbial load of herbal products indicated for male vitality
European Pharmacopoeia [21] microbial limits for non-aqueous oral dosage forms of herbal medicinal preparations for which antimicrobial pretreatment is not feasible The mixed sildenafil citrate and tadalafil HPLC analysis employed in this study detected sildenafil citrate in almost all the herbal products while tadalafil was also  found in many of the herbal products analyzed (Table 5).The amount of sildenafil citrate measured in the adulterated herbal products was in the range of 0.12-33.95mg/g, whereas the highest concentration of 0.82 mg/g of tadalafil was recorded in the products (Table 5).

Market survey
A high proportion of the herbal capsules found on the market were promoted and used for the treatment, management, prevention, and cure of various signs and symptoms suggestive of and involving sexual disorders (Table 1S).This observation lends support to reports of the high demand for herbal remedies for sexual disorders in communities where these are sold [23].

Physicochemical quality parameters
Three products passed the weight uniformity test, whereas the rest failed.This means that the amount of drug substances (dose) contained in each dosage unit is not uniform, as such each capsule unit may contain more than or less than the average or effective dose of drug substances needed to achieve the product's therapeutic objective.Studies have shown that the type and size of capsule shells used, the type, speed, and settings of the capsule filling machine together with some properties  inherent to the powder characteristics such as density, flowability, and particle size, influence the average capsule fill weight and hence weight variability [24,25].It is common knowledge that a good number of local herbal manufacturers fill capsule shells manually and thus variations in weight would be evident as observed in the study.Manufacturers of herbal capsules should pay attention to parameters such as the speed and setting of the capsule filler and factors that influence powder flowabilities such as powder density, particle size, and moisture content that may introduce a high variability to the contents of the capsules.Another important quality parameter measured was the disintegration time, which shows how quickly the shell fragments to release its content for dissolution and subsequent absorption when taken by the patient [26].The results obtained were found to be within the prescribed time limit of 30 min in the WHO recommendation on stability testing for active pharmaceutical ingredients and finished pharmaceutical products (Annex 10) [27].This variation may be a result of the differences in the prevailing environmental, storage, and handling conditions surrounding each product as they were sourced from two different geographic locations.Other physicochemical parameters such as the moisture content and pH, were found to be consistent with regulatory requirement.Moisture content influences the growth of microbes in pharmaceutical products.The guidelines and regulations issued by many global and national drug regulatory authorities specify moisture content as a routine stability testing requirement for the capsule dosage form of herbal medicinal products [28].The moisture content values obtained were below the water content range of 10-16% of many gelatinous capsule shells [29], and below WHO recommended level of 10% for powdered herbal materials [27].The moisture content also affects the texture and flowability of blended powders by forming stronger interparticle liquid bridges or increases powder flowability by acting as a lubricant [30], thus affecting mean capsule fills and the encapsulation process.For finished capsule dosage forms, moisture content accounts for the mechanical strength, solubility, and overall shelf-life stability of the medicinal product.To maintain product quality, it is important to pack and store them under conditions that would ensure the maintenance of optimal moisture content in the shell.pH is not a routine regulatory quality parameter requirement for finished medicinal products in capsule dosage forms [28] but it may influence the drug product's rate of decomposition, and microbiological quality.
The herbal capsules were also found to contain toxic metals whose concentrations were all within acceptable limits prescribed by the ASEAN limits for heavy metal contaminants [20].The heavy metals Arsenic, Cadmium, Lead, and Mercury are ranked priority metals of public health importance due to their degree of toxicity even at lower levels of exposure.Plants by their ability to withstand toxic ions can grow on metalliferous soils where they accumulate heavy metals in biological tissues [31].A simple main effects analysis showed that the type of herbal products did not have a significant effect on the heavy metal contents in each product.Since herbal products may be contaminated through some bad manufacturing practices, the low levels of heavy metals recorded in the herbal products and the result of the main effects analysis strike out the possibility of their deliberate addition to herbal products.Instead, the inconsistencies observed in their average concentration levels may be attributed to the different concentrations of heavy metals accumulated in the type of plant species and plant parts used in product formulation.It is known that some medicinal plants can accumulate some heavy metals more than others when grown under natural conditions.For instance, Senecio coronatus (Thunb.)Harv.(Asteraceae) is a nickel hyperaccumulating plant in Africa whereas Clausena anisata (Rutaceae) and Rauwolfia vomitoria (Apocynaceae) are known accumulators of iron [32]

Microbiological quality of the products
Apart from heavy metals, adverse reactions from herbal products could be traced to the presence of pathogenic microbes.However, the results from the present did not show any bile tolerant gram-negative bacteria, Escherichia coli, and Salmonella spp. in any of the herbal products (Table 4).These pathogenic bacteria reside in the gut of animals and their presence may be indicative of exposure to fecal contaminants and poor personnel hygienic practices.Bacterial and fungal colonies were found in all (8/8; 100%) herbal capsules analyzed.Total yeast and mold colonies counted in all the herbal products were within European Pharmacopoeia microbial acceptable standards (Table 4).Almost all the herbal products failed to comply with the TAMC for non-aqueous oral dosage forms.However, none of the bacteria were pathogenic.There are several reports on the microbiological quality of herbal medicinal products in various regions of Ghana and some countries in Africa [33][34][35].The unacceptable levels of microorganisms in herbal medicines can reduce or inactivate the therapeutic activity of the product and thus may present health risks to patients taking these medications.Due to the high risk associated with human consumption of contaminated products, a continuous microbiological quality monitoring program must be developed to assure safety.

HPLC detection of PDE inhibitors in herbal products
Both sildenafil and tadalafil were detected in some of the herbal capsules (Table 5).Sildenafil was found in 87.5% of the products in batch A and 62.5% of those in batch B. Thus, an adulterant (sildenafil) detected in a product in group A was absent in product replica in group B, and vice versa.This observation could be a deliberate attempt to evade regulatory scrutiny where chances of picking a spiked product in post market surveillance may be say 50%.The recommended dose of sildenafil citrate for most patients with erectile dysfunction is 50 mg daily as a single dose and usually one hour before sexual intercourse.For most of the herbal capsules in this study the recommended dosage is two capsules two times daily (4 capsules per day).From the dose uniformity data (Table 1,) the average net content weight of herbal medicine in herbal capsule HC-008 was determined as 0.2856 g.This implies that a client complying with the dosage indicated on the product label would consume a total of 1.1424 g of herbal medicines which could contain as much as 37.84 mg sildenafil citrate in a day which is lower than the recommended daily dose.These findings are corroborated by several studies on herbal products that showed evidence of adulteration with conventional sexual enhancing agents.Park and Ahn, [36] measured 4.3-453.2mg of sildenafil citrate and 2.2-40.4mg of tadalafil in some fake herbal medicinal products distributed in Korea.Elsewhere, Al-Tahami [9] used TLC and HPLC methods to record high levels of sildenafil citrate, tadalafil, and vardenafil in dietary supplements sold on the Yemeni market.The practice of adding synthetic pharmaceutical ingredients to herbal preparations to enhance therapeutic values is unacceptable, illegal and an offense that is punishable by law [37] The safety and efficacy of tadalafil in combination with other conventional sexual enhancing agents such as sildenafil citrate have not been investigated.However, an adulterated product may cause injury to the health of the patient when the product is used.Patients with health conditions contraindicated for sildenafil citrate and tadalafil may naively be exposed to unwarranted health risks as the contents of these adulterants are usually not declared on the product's label to inform consumers.

Conclusion
A high degree of variability in the quality characteristics of herbal medicinal products is indicative of inconsistency in the quality of herbal products and lack of standardization in their production processes.Herbal medicine manufacturers should standardize their manufacturing processes, develop, and maintain robust quality control/assurance programs to ensure the consistent production of safe and quality medicinal products.In the meantime, it is highly recommended that the Food and Drugs Authority institutes a recall action for all the contaminated herbal capsules indicated for male vitality.Beyond the recall, a more robust regulatory framework in respect of herbal aphrodisiacs is recommended to safeguard public health and safety.This should include sustained market surveillance and development of sensitive on-site detection technologies.

Limitations of the study
The study focused on the capsule dosage form of herbal medicines used in male sexually related illnesses marketed in Sunyani and Kumasi cities of Ghana.Other non-solid oral dosage forms sold which may also be adulterated with PDE inhibitors or contaminated with unacceptably high levels of microorganisms and heavy metals were not evaluated.Additionally, the number of shops sampled fell below the 10% threshold due to resource constraints.

3
Total aerobic microbial count (TAMC) = 2 × 10 4 cfu/mL Total combined yeast and mold count (TYMC) = 2 × 10 2 cfu/mL Absence of Salmonella ssp. in 10.0 g or 10.0 mL product sample Absence of E. coli in 1.0 g or 1.0 mL product sample Bile tolerant gram-negative bacteria: NMT 2 × 10 2 cfu/mL in 1 g or 1 mLNoProduct Total aerobic microbial count (TAMC) cfu/mL Mean (N = 3) Total combined yeast and mold count (TYMC) cfu/mL Mean (N = min (Fig.1).Chromatograms of products with peaks and retention times matching sildenafil citrate and tadalafil were flagged (Figs.2 and 3; Figs.2S-7S).Sildenafil citrate was found in 7/8 of products in group A and 5/8 products in group B. Products HC-034 and HC-039 had sildenafil in batch A but not in B. Product HC-037 did not contain sildenafil.

Fig. 1
Fig. 1 HPLC chromatogram of mixed sildenafil citrate and tadalafil reference samples

Fig. 3
Fig. 3 HPLC chromatogram of product HC-039 showing sildenafil and tadalafil peaks Mossel after which it was incubated at 37 °C for 48 h.Inoculated enterobacteria enrichment broth-Mossel was then sub-cultured on plates of violet-red bile glucose agar and incubated at 37 °C for another 24 h for growth to occur.The product sample was reported as compliant if no growth of colonies of gram-negative bacteria was detected on the plate.To test for the presence of Escherichia coli, 1.0 mL of inoculated casein soya bean digest broth was transferred to 100 mL of MacConkey broth and incubated at 37 °C for 48 h.The inoculated Mac-Conkey broth was then sub-cultured on a plate of Mac-Conkey agar at 37 °C for 72 h for growth to occur.The growth of colonies on plates of MacConkey agar indicated the presence of Escherichia coli in herbal capsules.To test for Salmonella spp.1.0 mL of inoculated casein soya bean digest broth was transferred to 100 mL of Salmonella enrichment broth and incubated at 37 °C for 24 h, then sub-cultured on plates of xylose, lysine, deoxycholate agar and incubated at 37 °C for 48 h.The growth of well-developed, red colonies, with or without black centers indicated the presence of Salmonella species.
(3)(which is equivalent to 1 g of product sample) of inoculated casein soya bean digest broth(3)

Table 1
Weight uniformity of herbal capsules indicated for male vitalityIf capsule have average net content by weight less than 300 mg, the product passed the weight variation test if the weight of the content of a minimum of 18 individual capsules fall within ± 10% deviation of the product's average weight of content, as well as a maximum of two (2) deviates more than two times the specified deviation.If the average weight content was more than 300 mg, then the capsule passed the variation test if the contents of not less than 18 individual capsules varied within ± 7.5% of the products average weight content, and a maximum of two individual capsule weight content deviated more than two times ± 7.5% of the product's average weight of content Key: "-": Not applicable.The average content weight of an herbal capsule is not less than 300 mg."P" = passed; "F" = failed

Table 2
Disintegration time, moisture content and pH for herbal capsules used in male sexual disorders

Table 3
Concentrations of heavy metals in herbal capsules indicated for male vitality

Table 5
Concentration of sildenafil citrate and tadalafil measured per gram of herbal capsule