Microsoft word - erah et al

African Journal of Biotechnology Vol. 2 (10), pp. 384-389, October 2003 Available online at http://www.academicjournals.org/AJB ISSN 1684–5315 2003 Academic Journals Plasmodium falciparum malaria resistance to
chloroquine in five communities in Southern Nigeria
Patrick O Erah*, Gertrude Arienmughare and Augustine O Okhamafe
Pharmacotherapy Group, Faculty of Pharmacy, University of Benin, Benin City, Nigeria. Chloroquine is still a first-line antimalarial drug in uncomplicated falciparum malaria. Increasing
resistance to chloroquine has been reported in many parts of Nigeria. Clinical and parasitological
responses and classes of resistance to chloroquine in falciparum malaria in five communities in Delta
region, southern Nigeria were assessed. Chloroquine was administered to 218 patients with
uncomplicated P. falciparum malaria. The levels of parasitemia, clinical response and classes of
resistance were monitored for 7 days. High levels of therapeutic failures of chloroquine in P. falciparum
malaria were recorded in the region. The frequencies of clinical and parasitological failure of
chloroquine were 25.7% and 55%, respectively. These frequencies were significantly lower in children
below 5 years than older people. R2 and R3 resistance occurred in 37.2% and 17.4% of the patients,
respectively. The therapeutic failure of chloroquine was not gender dependent. We conclude that
chloroquine is still effective in the treatment of uncomplicated P. falciparum malaria in some
communities in Delta region of Nigeria. However, resistance to chloroquine is likely. These results may
be used as an important indicator of the significant level of therapeutic failure of uncomplicated P.
falciparum malaria to chloroquine in Nigeria.

Keywords: Chloroquine, falciparum malaria, resistance, clinical failure, parasitological failure.


INTRODUCTION
Malaria is the second leading health problem (the first
Malaria in Africa is mainly due to a blood parasite, being HIV/AIDS) in sub-Saharan Africa accounting for Plasmodium falciparum. Chloroquine, which was over 1 million deaths yearly in the region (Rathod et al., introduced in the 1940s, and for many decades served as 1997). These deaths are primarily among children under a cheap and reliable drug, is becoming ineffective against 5 years of age and pregnant women (Phillips, 2001). The P. falciparum in most tropical areas (Rathod et al., 1997; disease can increase the risk of major problems including Warhurst, 2001). Resistance to chloroquine developed in anaemia, premature births, still-births, abortion and low Southeast Asia and South America at the end of the birth weight. Health and socio-economic problems 1950s and in Africa by the late 1970s (Warhurst, 2001). caused by malaria in sub-Saharan Africa are enormous, Since the emergence of chloroquine resistant strains of and the development and spread of malaria parasite P. falciparum, the rate of resistance has been increasing resistance to drugs have led to global recognition and and limiting adequate treatment of malaria (Peters, 1998; commitment to public health problems of malaria. Neequaye et al., 1986). Despite the growing resistance of P. falciparum, chloroquine remains the first line antimalarial drug in Nigeria and many other African countries mainly on account of cost and effectiveness in *Corresponding author; E-mail: [email protected], Tel: +234 uncomplicated cases of malaria. Resistance to 802 336 0318; 805 526 3622, Fax: +234 52 602257. chloroquine by P. falciparum has prompted many studies within the last decade in different parts of Nigeria (Molta, of Delta State and were selected because of their 1995; Umotong et al., 1991; Salako et al., 1990; proximity to the study site. Apart from Agharho (a village Sowunmi et al., 1990; Sowunmi and Salako, 1992). In the whose inhabitants are mainly farmers), the communities study carried out in northeastern Nigeria, for example, are urban communities whose inhabitants are mainly most strains of P. falciparum were found to be fully petty traders, civil servants and workers in the oil sensitive to chloroquine (Molta, 1995). However, strains industry. They are served by the Central Hospital, Warri with reduced sensitivity and resistant strains have and Shell Petroleum Hospital, Warri as well as private emerged with parasitological failure increasing from clinics, herbal homes, pharmaceutical stores and patent 18.7% in 1988 to 24.5% in 1995 (Molta, 1995). medicine stores. Like other parts of southern Nigeria, the Consistently high resistance to chloroquine by malaria climate is equatorial, with 2 seasons: rainy season from parasite has also been reported in southeastern Nigeria May to October and dry season from November to April. The goal of chemotherapy in malaria therapy is often to effect a clinical cure or parasitological clearance or to Patients and study protocol
limit the development of drug resistance. Resistance to chloroquine is often manifested in subtle ways which are After ethical approval from the Delta State Ministry of only apparent when the treatment is accompanied by Health, Asaba, 500 males and females patients (age: 6 detailed follow-up. Accurate and effective surveillance months and above) with history of fever (axillary systems for monitoring antimalarial drug efficacy have temperature ≥ 37.5oC) in the last 48 h before attending been recognized as an essential basis for decisions on the hospital (Central Hospital, Warri) were clinically the use of drugs (Malaria Foundation International, 1998). examined and screened for malaria parasites using both Surveillance is often carried out at different levels of clinical examination and laboratory data. Of the 500 sophistication, from basic drug efficacy tests through to patients, 218 patients (age range, 6 months to 48 years) more detailed in vitro characterization of drug resistant met the criteria for inclusion in the data analysed. These parasite strains and clinical information. The World patients were selected from the 5 communities using Health Organization has recognized that malaria simple random sampling as earlier described (WHO, treatment failure rate of 5% to 14% signals an alert phase 1992). The sample was evenly distributed among the in malaria monitoring while failure rates of 15% to 24% communities selected and the sample size was signals the need for action (WHO, 2002a). Among other satisfactory in respect of the calculated sample size of factors, reports of chloroquine resistant P. falciparum patients based on the population of the communities at treatment failure rate of up to 53.6% in southeastern 95% confidence interval, 5% precision, and expected Nigeria (Umotong et al., 1991) and up to 37% in proportion of treatment failure of 5% – 14 % (WHO, southwestern part of Nigeria (Salako et al., 1990; Sowunmi et al., 1990) with a drop to 15% in 1992 The criteria for inclusion in the study included the (Sowunmi and Salako, 1992) justifies the need for presence of P. falciparum malaria as determined by continuous monitoring of resistance to chloroquine. clinical examination and laboratory data (parasite count; The main purpose of this study was to ascertain the 2,000 to 200,000 per µl blood) and informed consent. therapeutic efficacy of treatment of uncomplicated P. Other criteria for inclusion were: (1) ability to be followed falciparum malaria with chloroquine in the Delta region of up for 7 days, (2) presence of febrile conditions caused Nigeria. Specifically, the parasitological and clinical by P. falciparum malaria, (3) absence of general danger responses of P. falciparum to chloroquine and the signs or signs of severe and complicated P. falciparum classes of resistance are evaluated. The data obtained malaria, and (4) the fact that the patient has not taken from this study would provide useful information for future any antimalarial drug within 2 week period preceding the management of uncomplicated P. falciparum malaria in study. However, patients with clinical symptoms compatible with severe or complicated malaria or with any other symptoms or signs of non-malarial etiology were excluded from the study and referred to appropriate health services. Also excluded were patients with Study area
repeated vomiting, diarrhoea, malnutrition, pregnancy, difficulty in complying with drug treatment, and inability to This study was carried out in a government-owned comply with the stipulated follow-up visits on days 3 and Central Hospital, Warri which is a 300-bed tertiary health care facility. It involved patients drawn from five The 218 patients (113 males and 105 females communities in Delta State, Southern Nigeria namely, including 78 children below 5 years) were routinely Agbarho, Aladja, Effurun, Ekpan and Enerhen with a treated with chloroquine orally: adults were treated with population of over 1 million inhabitants. These 600 mg chloroquine base stat and 24 h later followed by communities are located in malaria endemic oil rich area 300 mg next day; children received 10 mg/kg stat and 24 386 Afr. J. Biotechnol.
Table 1. Age and sex distribution of patients with P. falciparum malaria treated with chloroquine.
aχ2 = 0.015, df = 1, p = 0.903 h later, then 5 mg/kg next day, as appropriate. Each with Chi square test or Fisher’s Exact test, using a patient was also given appropriate doses of paracetamol computer software, Instat (GraphPad Inc., USA). and multivitamin orally. The qualities of drugs Probability (p) was considered at 95% confidence administered were evaluated and found to have met the interval, and 2 – tailed p-values less than or equal to 0.05 British Pharmacopoeia (BP, 1998) standard before they were accepted to imply significant differences between were used for this study. All patients were treated as outpatients and were appropriately followed up for 7 days with clinical and laboratory examinations at 3 h, day 3 and day 7 following the start of treatment with chloroquine. Blood samples were also obtained from the patients during the follow-up days and malaria parasite Age and sex distribution of patients
levels determined using standard methods (Ejov et al., The age and sex distribution of the 218 patients who fully participated in this study are given in Table 1. This number excluded those who either defaulted or withdrew Clinical and parasitological responses to chloroquine
from the study on account of convenience. There was no significant difference between the number of males and Clinical failure was considered to have occurred if a females, and the number of children below 5 years when patient had fever or history of fever (axillary temperature compared with other patients from 5 years and above (p ≥ 37.5oC) by day 7 with parasitemia detected in blood or assessed to have been ill due to no other cause than malaria. Parasitological failure was considered to have Clinical and parasitological responses to chloroquine
occurred if parasite count in blood was ≥ 25% of count on The clinical and parasitological responses to chloroquine are shown in Table 2. Chloroquine produced a clinical success of 74.3% in the eradication of malaria parasites Level of resistance to chloroquine
in the patients but parasitological success was only achieved in 45% of the patients. The proportion of We adopted the 7-day WHO in vivo test protocol in patients with clinical failure or parasitological failure were determining the class of resistance to chloroquine (Bruce, significantly higher in patients from 5 to 48 years old than 1986). Patients with marked reduction of parasitemia patients below 5 years (p ≤ 0.001). Furthermore, the (parasite count reduced by more than 75%) at 48 h but proportion of patients from 5 to 48 years old with clinical failed to clear parasites by day 7 were considered to be failure was more than twice the proportion of patients R2 resistant to chloroquine. Patients whose parasitemia below 5 years with clinical failure. Also, the proportion of did not fall by more than 75% within 48 h or occasionally patients from 5 to 48 years years old with parasitological increased by day 7 were considered to be R3 resistant to failure was nearly doubled the proportion of patients chloroquine. R1 resistance (clearance of asexual below 5 years with parasitological failure. Gender had no parasitemia on or after day 7 followed by recrudescence) significant effect on either the clinical response or was not determined because the patients were only parasitological response to chloroquine (p > 0.05). monitored for 7 days and it is not possible to determine recrudescence in areas of intense malaria transmission Level of resistance to chloroquine
Class II (R2) and class III (R3) types of resistance were Statistical Analysis of data
found to have occurred in 81 (37.2%) and 38 (17.4%) of the patients, respectively. We found no significant The data obtained were categorized based on sex and differences in the levels of resistance in males when age and reported as frequencies. Data were compared compared to those of the females (p > 0.05).
Table 2. Clinical and parasitological response of chloroquine in patients with P. falciparum malaria.
bp = 0.001; cp < 0.001

DISCUSSION
people. Malaria infections with parasite strains other than P. falciparum have been suggested to act as natural Drug resistance in malaria is the ability of the parasite vaccines preventing severe manifestations of P. strains to survive and/or multiply despite administration falciparum infections (Williams et al., 1996). If this and absorption of a drug given in doses equal to or suggestion holds, clearance of P. falciparum malaria in higher than those usually recommended but within the the patients below 5 years of age could be enhanced, as limits of tolerance of the subject (Bruce, 1986). The World none of them had severe malaria. Furthermore, long-term Health Organization (WHO) categorized resistance in exposure to chloroquine, particularly by presumptive malaria as S (sensitive), R1, R2 and R3 based on in vivo chloroquine treatment or prophylaxis, are prerequisites test (Bruce, 1986). This test was replaced in 1996 by for the emergence of chloroquine resistance WHO with the Therapeutic Efficacy Test based on clinical (Mockenhaupt et al., 2000). This situation is less likely to and parasitological criteria - a pragmatic test for National occur in children below the age of 5 years than older Malaria Control Programme (WHO, 1996). We have used people. The efficacy of chloroquine is thus more likely to these two approaches to evaluate the effectiveness of be lower in people older than 5 years than those below 5 chloroquine in the five communities studied. Our results years of age. Greater awareness on the need for early have shown that chloroquine is still effective in the and proper treatment in chlidren below 5 years of age treatment of uncompicated P. falciparum malaria in the may also be a contributing factor. Gender appears not to communities despite the high prevalence of chloroquine- play a role in therapeutic efficacy of chloroquine as we resistant malaria in the region. This is evident in the found no significant difference between the frequency of clinical success of chloroquine in 74.3% of the patients. therapeutic failures in males as compared to females. However, parasitemia was still evident in 45% of the Increasing resistance of P. falciparum malaria to patients by the 7th day after the first dose of chloroquine chloroquine has earlier been reported in some parts of was administered, indicating resistance to chloroquine. Nigeria (Umotong et al., 1991). Survellance network in We recognise that a mixed population of malaria parasite the country from mid-1987 to 1990 revealed that strains (susceptible and resistant) can be present in the chloroquine resistant P. falciparum (CRPF) was blood of the patients. In this situation, the susceptible widespread (Ekanem, 1997). In the northern part of the strains would be killed by the chloroquine in the early part country, parasitological failures increased from 18.7% in of the treatment, making way for more resistant and 1988 to 24.5% in 1995 (Molta, 1995). From 1987 to 1997, virulent strains to multiply later (Wernsdorfer et al., 1995). resistance ranging from 33% to 72% was also reported in Children under the age of 5 years are known to be the the southern part of the country (Ekanem, 1997). most susceptible group to malaria mortality (WHO, Although the parasitological failure found in our study is 2002b). The severity and prevalence of malaria are also similar to the 53.6% found in Calabar, southern Nigeria in higher in this age group than older children and adults 1991 (Umotong et al., 1991), stability of resistance to (May et al., 1999; Mockenhaupt et al., 2000). chloroquine in the area studied cannot be inferred Furthermore, at equal dosages per body weight, plasma because the two areas (Calabar and Delta region) are in chloroquine concentrations are lower in children below 5 different locations. Clinical failure of 25.7% recorded in years of age than older children and adults (Maitland et this study is also similar to the 22% reported in Ibadan, al., 1997). These indications would suggest that the southern Nigeria in 1997 (Ekanem, 1997; Oduola, 1997). therapeutic failure of chloroquine in malaria treatment is The frequencies of clinical and parasitological failures likely to be higher in children younger than 5 years than following chloroquine treatment are indicative of the older people. Conversely, our results indicate that continuous emergence of chloroquine resistance strains children younger than 5 years had much lower of P. falciparum in the communities studied and most therapeutic failure rates than older people. The reason for likely in other parts of the country. These frequencies are this is unclear. Earlier report indicates that mixed species at unacceptable levels having exceeded the 15 – 24% set infections of malaria parasites are higher in children by WHO (2002a) as action phase. Thus, the need for below 5 years (Mockenhaupt et al., 2000) than older action to be taken to address the problem in the communities and other parts of the country has become has been established as a measure to improve treatment efficacy, delay the emergence of drug resistance, and Based on WHO earlier categorisation of resistance to reduce the prevalence of gametocyte carriage which antimalarial drugs (class I, R1; class II, R2 and class III, could, in turn, reduce transmission of malaria parasites R3), our results have demonstrated high proportions of (WHO, 2001a,b). Cost considerations in many parts of R2 (37.2%) and R3 (17.4%) resistance in the the country would support the use of empirical treatment communities studied. The R3 resistance was similar to with chloroquine or an alternative first-line drug, such as that reported in 1991 in Calabar (Umotong et al., 1991). sulfadoxine-pyrimethamine (Warhurst, 2001), in patients Studies have shown that R3 resistance to chloroquine is with uncomplicated malaria. In the case of more not present in some areas in Congo (Carme et al., 1998) expensive drug regimens, the cost would have to be and Madagascar (Raharimalala et al., 1995) but present balanced against the potential savings in the cost of at a very low level in Coté d'Ivoire (Henry et al., 1998). R3 treatment. Reliable, microscopy-based diagnosis should resistance in this study is lower than the 23% reported in be mandatory in view of the financial implications when Zimbabwe (Mohamva et al., 1996) and the 22% reported much more expensive alternative drugs need to be used in Port Moresby, Papua New Guinea (Hombhanje, 1997). However, the R2 resistance in our results is much higher than the 10% reported in Zimbabwe (Mohamva et al., 1996) and the 4% reported in Papua New Guinea Conclusion
(Hombhanje, 1997). High, rather than low, doses of chloroquine is believed to induce R3 resistance to Chloroquine is still effective in the treatment of most cases of uncomplicated P. falciparum malaria in some The mechanism of resistance to chloroquine in the communities in Delta State, Nigeria. However, in many communities studied is not clear, since the mechanism of cases of uncomplicated P. falciparum malaria, resistance resistance to antimalarial drugs is still debatable to chloroquine is likely and intervention with alternative (Wellems et al., 1990; Rathod et al., 1997; Phillips, 2001). antimalarial drugs may be required for complete However, resistance to chloroquine has been reported to clearance of parasitemia in the patients. Despite the be due to reduced uptake of the drug by the infected limited communities in which this study was carried out, erythrocyte and presumably reflects a reduction in the our results may be used as an important indicator of the accumulation of the drug in the parasite’s lysosome significant level of therapeutic failure of uncomplicated P. (Phillips, 2001; Warhurst, 2001). Association between falciparum malaria to chloroquine in Nigeria. Additional chloroquine resistance and mutations in an mdr-like gene drug efficacy assessment throughout the country is (pfmdr-1) has been reported (Warhurst, 2001). There is needed to support national antimalarial drug policy. evidence to suggest that a mutation in pfcrt is required to confer a basic level of resistance before mutations in pfmdr-1 can have an effect; the basis for the suggested ACKNOWLEDGEMENTS
use of a rapid and sensitive test to detect pfcrt T76 in blood samples to ascertain the effectiveness of We are grateful to the staff and management of Central chloroquine (Warhurst, 2001; Reed et al., 2000). A Hospital, Warri and Calvary Medical Laboratory for their number of factors have been identified as playing a role support. This work was carried out with the financial in the emergence of resistance to chloroquine. These support from the Arienmughare family of Warri, Delta include poor compliance to drug therapy, host immunity, use of sub-therapeutic doses, mutation and frequency of re-infection (Hombhanje, 1997; Phillips, 2001; REFERENCES
Treatment of uncomplicated malaria in the Niger-Delta region (where this study was carried out) is frequently British Pharmacopoeia (1998). The Pharmaceutical Press, London. based on the fact that as long as chloroquine is effective Bruce CLJ (1986). Chemotherapy of Malaria. WHO Technical Report Series No. 27, Geneva, World Health Organization, Geneva. enough, there is very little incentive for basing its use on Carme B, Ndounga M, Kissila AM, Samba G, Baya TN (1998). No reliable diagnostic confirmation. As in the communities variation in chloroquine resistance (Plasmodium falciparum) from studied, resistance to antimalarial drugs is proving to be a 1986 to 1996 in semi-immune children in Brazzaville (Congo). Bull. challenging problem in malaria control in most parts of Ejov MN, Tun T, Aung S, Sein K (1999). Response of falciparum the world. A fundamental aspect of drug usage malaria to different antimalarials in Myanmar. Bull. World Health recommendations is the impact of resistance on morbidity and mortality and hence the impact of the timing and Ekanem OJ (1997). Use of antimalarial drugs in Nigeria. National nature of a policy decision on the health of the people in symposium on malaria in Nigeria, Nigeria Institute for Medical Research, Lagos, Nigeria. the community studied and the rest of the country. The Henry M, Koné M, Guillet P, Mouchet J (1998). Resistance to principle of combination drug in malaria chemotherapy chloroquine and malaria control in the Ivory Coast. Santé 8:287-291. Hombhanje FW (1997). Parasitological response of Plasmodium Sowunmi A, Salako LA, Walker O, Ogundahunsi OA (1990). Clinical falciparum infection to chloroquine treatment in malaria patients in efficacy of mefloquine in children suffering from CRPF in Nigeria. Trans. Roy Soc. Trop. Med. Hyg. 84: 761-764. Maitland K, Williams TN, Kotecka BM, Edstein MD, Rieckmann KH Sowunmi A, Salako LA (1992). Evaluation of the relative efficacy of (1997). Plasma chloroquine concentrations in young and older various antimalarial drugs in Nigerian children under 5 years of age malaria patients treated with chloroquine. Acta Trop. 66:155-161. suffering from acute uncomplicated falciparum malaria. Ann. Trop. Malaria Foundation International (1998). Report on the multilateral initiative on malaria meeting on antimalarial drug usage and Umotong AB, Ezedinachi EN, Okerengwo AA, Usanga EA, Udo JJ, resistance. http://www.malaria.org/ppt/drug.html; assessed Sept 25, Williams AI (1991). Correlation between in vivo and in vitro response of choloroquine resistant Plasmodium falciparum in Calabar, South- May J, Mockenhaupt FP, Ademowo OG, Falusi AG, Olumese PE, Eastern Nigeria. Acta Tropica. 49: 119-125. Bienzle U, Meyer CG (1999). High rate of mixed and subpatent Wellems TE, Panton LJ, Gluzman IY, do Rosario VE, Gwardz RW, malarial infections in southwest Nigeria. Am. J. Trop. Med. Hyg. 61: Walker-Jonah A, Krogstad DJ (1990). Chloroquine resistance not linked to mdr-like genes in Plasmodium falciparum. Nature 345: 202- Mockenhaupt FP, May J, Bergqvist Y, Ademowo OG, Olumese PE, Falusi AG, Gro terlinden L, Meyer CG, Bienzle U (2000). Wernsdorfer WH (1991). The development and spread of drug resistant Concentrations of chloroquine and Malaria parasites in blood in Nigerian children. Antimicrob. Agents Chemother. 44: 835-839. Wharhurst DC (2001). A molecular marker for chloroquine-resistant Mohamva AI, Peterson DE, Rakata L (1996). Chloroquine resistant falciparum malaria (editorial). New Engl. J. Med. 344: 299-302. falciparum malaria in Mutare District, eastern Zimbabue. Cent. Afr. J. Williams TN, Maitland K, Bennett S, Ganczakonski M, Peto TE, Newbold CI, Bowden DK, Weatherall DJ, Clegg JB (1996). High Molta NB (1995). Susceptibility of Plasmodium falciparum to malarial incidence of malaria in alpha-thalassaemic children. Nature. 383: drugs in North-eastern Nigeria. Trans. Roy. Trop. Med. Hyg. 89: World Health Organisation (1992). WHO/DAP/92.3, How to investigate Neequaye J, Coe-Ene J, Taelman H (1986). In vivo chloroquine drug use in communities. World Health Organization, Geneva, resistant falciparum malaria in West Africa. Lancet 1: 153-154. Oduola AMJ (1997). Susceptibility of P. falciparum to antimalarial drugs World Health Organisation (1996). WHO/MAL/96.1077, Assessment of in vitro in Nigeria. National symposium on malaria in Nigeria, Nigeria therapeutic efficacy of antimalarial drugs for uncomplicated Institute for Medical Research, Lagos, Nigeria. falciparum malaria in areas of intense transmission. World Health Peters W (1998). Drug resistance in malaria parasites of animals and World Health Organisation (2001a). WHO/CDS/RBM/2002.33, Use of Phillips RS (2001). Current status of malaria and potential for control. antimalarial drugs. Report of a WHO informal consultation. World Health Organisation, Geneva, Switzerland. Raharimalala L, Rabarison P, Pepers-Rason MD, Pepers JP, World Health Organisation (2001b). WHO/CDS/RBM/2002.35, Ramambanirina L, Rason MA, Jambou R, Roux J (1995). Drug Antimalarial drug combination therapy. Report of a WHO technical resistance and therapeutic strategy. Sante 5: 389-392. consultation. World Health Organisation, Geneva, Switzerland. Rathod PK, McErlean T, Pei-Chieh L (1997). Varaitions in frequencies World Health Organisation (2002a). Antimalarial drug resistance: of drug resistance in Plasmodium falciparum. Proc. Natl. Acad. Sci. Guidelines for surveillance and containment. World Health Reed MB, Saliba KJ, Caruana SR, Kirk K, Cowman AF (2000). Pgh1 World Health Organisation (2002b). WHO/CDS/RBM/2002.42, modulates sensitivity and resistance to multiple antimalarials in Community involvement in rolling back malaria. World Health Plasmodium falciparum. Nature 403: 906-909. Salako LA, Sowunmi A, Walker O (1990). Evaluation of the clinical efficacy and safety of halofantrin in falciparum malaria in Ibadan, Nigeria. Trans. Roy. Soc. Trop. Med. Hyg. 84: 644-647.

Source: http://www.uniben.edu/sites/default/files/abstracts/AJB_Erah_et_al2.pdf