Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh.

The proportion of Plasmodium vivax malaria among all malarias is increasing worldwide. Treatment with 8-aminoquinolines remain the only radical cure. However, 8-aminoquinolines can cause severe hemolysis in glucose-6-phosphate dehydrogenase (G6PD) deficient patients. The population of the multi-ethnic Chittagong Hill Tracts (CHT) carry the highest malaria burden within Bangladesh. As in many countries the national treatment guidelines recommend 8-aminoquinoline based radical cure without routine G6PD deficiency (G6PDd) testing to guide treatment. Aim of this study was to determine the need for routine testing within a multi-ethnic population by assessing the prevalence of G6PDd among the local population. Participants from 11 ethnicities were randomly selected and malaria status was assessed by microscopy, rapid diagnostic test (RDT) and polymerase chain reaction (PCR). G6PD status was determined by spectrophotometry and G6PD genotyping. The adjusted male median (AMM) was defined as 100% G6PD activity, participants were categorized as G6PD deficient (<30% activity), G6PD intermediate (30% to 70% activity) or G6PD normal (>70% activity). Median G6PD activities between ethnicities were compared and the association between G6PD activity and malaria status was assessed. 1002 participants were enrolled and tested for malaria. G6PD activity was measured by spectrophotometry in 999 participants and host G6PD genotyping undertaken in 323 participants. Seven participants (0.7%) had peripheral parasitaemia detected by microscopy or RDT and 42 by PCR (4.2%). Among 106 participants (32.8%) with confirmed genotype, 99 (93.4%) had the Mahidol variant. The AMM was 7.03U/gHb with 90 (9.0%) G6PD deficient participants and 133 (13.3%) with intermediate G6PD activity. Median G6PD activity differed significantly between ethnicities (p<0.001), proportions of G6PD deficient individuals ranged from 2% to 26% but did not differ between participants with and without malaria. The high G6PDd prevalence and significant variation between ethnicities suggest routine G6PDd testing to guide 8-aminoquinoline based radical in the CHT and comparable settings.

Introduction

Great efforts in the reduction of malaria have been made over the last decade, but with a much greater impact on Plasmodium falciparum (P. falciparum) than the neglected P. vivax parasite; accordingly, the proportion of malaria cases caused by P. vivax is increasing worldwide [1]. In contrast to P. falciparum, P. vivax forms dormant liver stages (hypnozoites) that reactivate weeks to months after the first infection, causing significant morbidity and mortality [2]. Relapses arising from reactivation of hypnozoites are the main contributor to the transmission of P. vivax infections and may contribute to over 80% of clinical cases [3,4]. The 8-aminoquinolines primaquine (PQ) and tafenoquine (TQ) are the only licensed drugs that effectively remove hypnozoites from the human host and, while well tolerated in most recipients, can cause severe and potentially fatal hemolysis in patients with glucose-6-phosphate dehydrogenase deficiency (G6PDd) [5-12]. G6PDd is among the most common inherited enzymopathies worldwide with more than 400 million people affected; prevalence differs by ethnicity and is highest among populations at risk of malaria, possibly the result of a protective effect of G6PDd against some forms of malaria [13-15]. The WHO recommends routine testing for G6PDd to guide 8-aminoquinoline based radical cure [16,17], a recommendation frequently not followed since drug induced hemolysis is under-estimated, the impact of vivax malaria is not recognized and fears of additional costs to national public health systems [18,19].

Malaria has declined over the last decade in Bangladesh, with the population at risk falling to 17 million across 13 districts [20,21]. The highest rates of malaria are in the multi-ethnic Chittagong Hill Tracts Districts (CHT), on the eastern border with Myanmar, where a total of 20,446 clinical cases were reported in 2016 with an estimated incidence of 1.7 per 1,000 population.

P. falciparum remains the dominant malaria species in the country, however its proportion has dropped from 90% in 2007 to less than 80% in 2016, with a rising proportion of P. vivax cases [22-25]. As in many countries, PQ is provided for the radical cure of P. vivax, but patients are not tested for G6PDd to guide treatment and no systematic assessment of the prevalence of G6PDd within the country has been conducted. Anecdotal reports from the CHT estimate the G6PDd prevalence between 20% and 40% (Khan, personal communication), however in hospital based surveys in the same region only 0.6% to 1.4% of patients were G6PDd [25,26]. The significant differences in these estimates may reflect study populations with different ethnic backgrounds, and associated differences in G6PD genetic profiles, or an underlying difference in G6PD activity in patients with and without malaria [15,27]. Understanding whether the prevalence of G6PDd differs significantly between different ethnicities living in the same area is essential in informing decision-makers on whether and how to introduce routine testing for G6PDd into national malaria treatment guidelines. The aim of this study was to assess whether different ethnicities sharing the same risk of malaria infection would have significantly different prevalence of G6PDd.

Methods

Ethics statement

The study was approved by the Ethics Review Committee of the icddr,b, Bangladesh (PR-15021), the Human Research Ethics Committee of the Northern Territory Department of Health and Menzies School of Health Research, Australia (HREC 2015–2336), and the University of Notre Dame Institutional Review Board (18-09-4875). Written informed consent was collected from all participants or their legal guardians prior to enrolment and in addition written assent was collected from all minors above the age of 11 years.

Study area

The CHT are densely forested, with limited accessibility and while the area covers approximately 10% of the country, it accounts for only 1% of the total population [28]. At least 12 major indigenous groups and a Bengali subpopulation are permanent residents of this region [29,30]. The CHT carry the highest burden of malaria within the country, incidence is seasonal with a major peak around June to August and a second smaller peak around February.

Sampling

Most villages (paras) in the CHT are mono-ethnic [30]. To sample sufficient numbers of participants from all targeted ethnic groups, individual villages were selected purposefully according to the village size, accessibility and ethnicity. The average village comprises 46 households, a total of 34 villages and 2134 households were selected [30]. Each village was mapped using Google Maps (Alphabet Inc., Mountainview, USA) and individual households were identified, numbered and a subset selected at random.

To ensure a good representation of genetic differences in G6PD alleles, one person per household above the age of one year was selected randomly and invited to participate. Randomization was done using pre-generated randomization list from random.org [31] (last accessed 24.06.2020).

Community sensitization, consent and sample collection

Selected villages were visited several days prior to participant enrolment to inform the population about the survey. Following written informed consent, a questionnaire and medical examination were completed including details on body weight, height and axillary body temperature, and a self-reported medical history. A malaria rapid diagnostic test (RDT) (FalciVax, Zephyr Biomedicals, Goa, India), a malaria microscopy slide and hemoglobin (Hb) measurement (Hemocue 201, Angelholm, Sweden) were done on site. Participants were informed of the results of their malaria RDT and Hb concentration and those with a positive malaria RDT were referred to the closest medical facility for treatment. Participants above the age of 7 years were asked to contribute a venous blood sample (maximum of 5ml) which was collected in an EDTA tube (BD, Franklin Lakes, USA). In participants less than 7 years old a capillary sample (~400μl) was collected into an EDTA Microtainer (BD, Franklin Lakes, USA) tube.

Sample processing

Slides were read at a local laboratory. Thick and thin films were stained with Giemsa and the parasite density quantified either per 500 white blood cells or 1000 red blood cells. All slides were read by two independent readers and the mean was recorded, provided that the discrepancy between readings was less than 50% of the smaller value and species identification was in accordance. The presence of gametocytes was recorded, but not quantified. Whenever discordant results were found the respective slides were shipped for reference reading to the International Centre for Diarrheal Diseases and Research, Bangladesh (icddr,b), Dhaka, and this result was considered as final.

Blood samples were stored immediately at 4°C and shipped under controlled conditions to the reference laboratory at the icddr,b in the capital Dhaka for quantitative G6PD measurements, host genotyping and Plasmodium species confirmation by qPCR. Since samples can hemolyze during transport, G6PD activity was only measured if there were no visible signs of hemolysis. G6PD measurement was done on a temperature controlled Shimadzu UV-1800 spectrophotometer (Kyoto, Japan), using kits and lyophilized controls from Randox Laboratories (Crumlin, UK), following procedures described previously, considering the Hb value measured in the field [32,33]. All samples were measured in duplicate, and a third measurement was performed if the measurements differed by more than 10% of the upper value. All procedures followed the manufacturer’s instructions with normal and deficient controls run with every test run. The measurement temperature was set to 37°C, the absorption was read twice, five minutes apart, at 340 nm. G6PD activity was calculated from the difference in absorption between the two measurements, following a formula provided by the manufacturer; no temperature correction factor was applied. A run was repeated in case either control returned a result outside the range recommended by the manufacturer.

DNA from whole blood was extracted using QIAamp DNA Blood Minikit (Qiagen, Germany) according to manufacturer’s instruction, and screened by qPCR for presence of malaria parasites. 4 μL of DNA, corresponding to 4 ml of blood, was screened by qPCR for P. falciparum using the varATS assay, and for P. vivax using the cox1 assay. P. falciparum varATS qPCR was run in a total volume of 12 μL, containing 0.4 μM of each primer and probe, 6 μL TaqMan FastAdvanced (Applied Biosystems), and 4 μL DNA. P. vivax cox1 qPCR was run in a total volume of 12 μL, containing 0.4 μM of each primer and probe, 6 μL TaqMan FastAdvanced (Applied Biosystems), and 4 μL DNA. The varATS assay targets multicopy genes and amplifies approximately 20 copies per genome. Cox1 is a mitochondrial gene and present in ~10 copies per genome [34,35]. By qPCR, samples were only screened for P. falciparum and P. vivax. Samples were not screened for P. malariae, P. ovale, or P. knowlesi.

A subset of samples was genotyped for host G6PD variants known to be present in the study area (Mahidol, Viangchan, Mediterranean, Orissa and Kalyan-Kerala) by Sanger sequencing. PCR was performed in a final reaction volume of 20μL containing 10 x PCR buffer with MgCl2 (25 mM), GC rich buffer, dNTP mixture (2.5 mM), forward (10 μM) and reverse primers (10 μM), FastStart Taq DNA Polymerase (Roche, France) and DNA template. The primer sequences are provided in S1 Table.

Thermal cycling profile for all of the included variants except the Kalyan-Kerala variant were set following reference literature with slight modifications: pre-denaturation was done at 95˚C for 15 minutes, followed by 35 cycles of denaturation at 94˚C for 45 seconds, annealing at specific temperature for 30 seconds and extension at 72˚C for 1 minute 20 seconds, and a final extension at 72˚C for 10 minutes [36]. Thermal cycling profile for Kalyan-Kerala variant: pre-denaturation at 95˚C for 15 minutes, 35 cycles of denaturation at 94˚C for 35 seconds, annealing at 63˚C for 40 seconds and extension at 72˚C for 1 minute 10 seconds, and a final extension at 72˚C for 10 minutes. To yield better DNA bands, the annealing temperatures for Mahidol and Viangchan variants were set according to reference literature and were optimized for Mediterranean, Orissa and Kalyan-Kerala variants (S2 Table).

Data management and statistical analysis

All participant and corresponding laboratory data were entered on paper forms, which were then digitalized using Epidata version 3.1 (Denmark). All analyses were done using Stata version 14 (College Station, USA).

G6PD activity, measured by spectrophotometry (in U/dL), was normalized by Hb measurement (in g/dL) to provide a result in U/gHb. The adjusted male median (AMM) was calculated and defined as 100% G6PD activity for the entire study population [37]. Individuals were categorized as being G6PD deficient if their enzyme activity was less than 30% of the AMM, G6PD intermediate if enzyme activity was between 30% to 70% activity, and G6PD normal if enzyme activity exceeded 70%. Age was assessed for normality and G6PD activity and age were then compared for significant differences between ethnic groups using the t-test, Mann—Whitney U test or Kruskal Wallis test as appropriate. Proportions of G6PD deficient, G6PD intermediate and G6PD normal participants per ethnic group were assessed for significant differences using a χ2-test or fishers exact test. Multiple regression analysis was performed to identify key variables affecting non-normalized G6PD activity. Non-normalized G6PD activity (in U/dL) was used as the dependent variable and Hb concentration included as a covariate [38]. The independent variables were backwards selected, initially considering age, gender, body weight, height and temperature, Hb, the delay between sample collection and processing, spectrophotometry result, ethnicity, and G6PD genotype. Participants were categorized by malaria status as being malaria positive by microscopy or RDT, positive by malaria PCR but not microscopy or RDT, or negative by all assays. Since categories showed strong dependency the model was repeated separately for microscopy/RDT and PCR diagnosis. Models were assessed for multi-co-linearity by calculating the variation inflation factor (VIF), with co-linearity defined when VIF was >10.

Sample size

To detect a G6PDd prevalence of 5% with 4% absolute confidence limit and assuming erroneous procedures in 10%, would require a total of 100 individuals enrolled per ethnic group [39,40]. To achieve this a total of 10 distinct ethnicities were identified within the area that could be contacted with a total sample size of 1,000 participants [30]. Host genotyping was performed on all samples with less than 70% G6PD activity by spectrophotometry as well as 10 randomly selected samples per ethnicity from G6PD normal participants.

Results

Between 13th August 2015 and 11th January 2016 a total of 1000 participants from 10 ethnic groups were enrolled into the study. In addition, two participants from the Lushai ethnic group were enrolled, bringing the total sample size to 1002 (Fig 1 and S1 File).

Fig 1

Consort chart of participant selection process.

Spectrophotometry was not performed in three individuals, resulting in G6PD data from 999 (99.7%) participants (Table 1).

Table 1

Overview on baseline data collected, stratified per ethnicity.

Ethnicity	Total	n Male (%)	n Female (%)	Mean Age in yrs. (95%CI)	Mean Hb in g/dL (95%CI)	Febrile (Temp >37.5C)n (%)	Mean body temp. in °C (95%CI)
Bawm	100	43 (43.0)	57 (57.0)	39.9 (36.9–43.0)	13.4 (13.1–13.7)	1 (1.0)	36.3 (36.3–36.4)
Bengali	99	34 (34.3)	65 (65.6)	26.2 (23.2–29.2)	13.1 (12.8–13.4)	1 (1.0)	36.4 (36.3–36.4)
Chak	100	32 (32.0)	68 (68.0)	26.8 (23.7–29.8)	12.8 (12.5–13.1)	0 (0.0)	36.3 (36.2–36.3)
Chakma	100	44 (44.0)	56 (56.0)	40.8 (37.6–44.0)	13.1 (12.8–13.4)	0 (0.0)	36.2 (36.1–36.3)
Khumi	100	45 (45.0)	55 (55.0)	31.3 (27.8–34.8)	13.2 (12.9–13.5)	0 (0.0)	36.5 (36.4–36.6)
Khyang	100	43 (43.0)	57 (57.0)	38.9 (35.2–42.6)	12.8 (12.4–13.2)	0 (0.0)	36.1 (36.0–36.2)
Lushai	2	1 (50.0)	1 (50.0)	50.5 (NA)	13.9** (13.1–14.7)	0 (0.0)	36.4 (32.9–39.9)
Marma	99	40 (40.4)	59 (59.6)	28.3 (25.0–31.6)	12.2 (11.9–12.6)	1 (1.0)	36.1 (36.0–36.2)
Mro	100	37 (37.0)	63 (63.0)	27.6 (24.1–31.1)	13.7 (13.4–13.9)	0 (0.0)	36.1 (36.1–36.2)
Tangchangya	100	49 (49.0)	51 (51.0)	28.5 (25.2–31.9)	13.2 (12.9–13.6)	1 (1.0)	36.2 (36.1–36.3)
Tripura	99	33 (33.3)	66 (66.7)	36.8 (33.1–40.5)	12.6 (12.3–12.9)	0 (0.0)	36.3 (36.2–36.4)
Total	999	401 (40.1)	598 (59.9)	32.5 (31.4–33.6)	13.0 (12.9–13.1)	4 (0.4)	36.2 (36.2–36.3)

The proportion of males within each ethnic group ranged from 40.1% to 59.9% and did not differ significantly (p = 0.279) (Table 1). Mean Hb concentration was 13.0 U/gHb (95% confidence interval (CI): 12.9 to 13.1, total range: 4.8 g/dL—19.3g/dL). A total of 11 participants were below the age of 7, the ratio of males to females did not differ between those above and below 7 years of age (p = 0.110). When comparing median G6PD activities between both cohorts, no significant difference was observed (p = 0.1418). The mean axillary body temperature was 36.3°C (95% confidence interval (95%CI): 36.2–36.3) and 0.4% of all participants (4/1002) had a temperature above 37.5°C (Table 1). Hb and body temperature differed significantly between ethnic groups, and this was also apparent after stratifying by sex (all p<0.001).

Three participants were diagnosed with P. falciparum by microscopy; all three cases were confirmed by RDT. Four microscopy-negative cases were also diagnosed as parasite-positive by RDT, including 3 P. falciparum cases and one participant diagnosed with a mixed P. falciparum and P. vivax infection; two of these participants (both P. falciparum) reported a history of malaria in the last 90 days. Malaria qPCR was undertaken successfully in all cases. Six of the seven malaria patients positive by RDT were confirmed by qPCR, one participant diagnosed with P. falciparum by RDT and microscopy returned a negative PCR result, and 36 participants were positive by PCR only. Based on PCR a total of 21 participants were positive for P.falciparum, 16 participants were positive for P. vivax and 5 participants harbored both P. falciparum and P. vivax parasites.

Relative to uninfected participants, the mean Hb concentration was significantly lower in patients diagnosed with malaria by microscopy 11.2g/dl (95%CI: 6.2 to 16.2) vs. 13.0 g/dl (5%CI: 12.9 to 13.1); p = 0.0264) or RDT (11.2 g/dL (95%CI: 9.9 to 12.5) vs. 13.0 (95%CI: 12.9 to 13.1); p = 0.0018) but not by the PCR result (p = 0.3285).

The median duration between sample collection and measurement of G6PD activity in the reference laboratory was 2 days (IQR: 0 to 4, range 0 to 7). 100% G6PD activity was defined as 7.03 U/gHb (interquartile range (IQR): 5.38–8.69, total range: 0.73–17.17) and median G6PD activity varied significantly with ethnicity when considering only males (p<0.001) or females (p<0.001). G6PD activities across the entire study population showed a bimodal distribution, which was more apparent in males than females (Fig 2).

Fig 2

Histogram of G6PD activities for the entire study population and stratified by sex.

Vertical lines from left to right indicate 10%, 30% 70% and 100% of the AMM.

The lowest AMM was detected among the Bengali sub-population (5.43 U/gHb, IQR: 4.78–6.46), however, when females were included, the Chak had the lowest median G6PD activity (4.92 U/gHb, IQR: 1.83–7.12); Fig 3, Table 2, S1 Fig.

Fig 3

Distribution of G6PD activity per ethnicity, stratified by sex.

Horizontal lines indicate 100%, 70%, and 30% G6PD activity of the AMM (from top to bottom).

Table 2

Overview on G6PD activity and genotype per ethnicity.

Ethnicity	Total	AMM in U/gHb (IQR)	Number with <30% G6PD activity (%)	Number with >30%—<70% G6PD activity (%)	Number with hemizygous / homozygous Mahidol	Number with heterozygous Mahidol	Number with hemizygous / homozygous Orissa	Number with hemizygous Kalyan Kerala	Number with no variant determined (% of all genotyped)
Bawm	100	7.12 (5.48–8.76)	8 (8.0)	9 (9.0)	5	1	0	0	21 (77.8)
Bengali	99	5.43 (4.78–6.46)	6 (6.1)	21 (21.2)	0	2	2	1	32 (86.5)
Chak	100	5.71 (1.27–7.93)	26 (26.0)	24 (24.0)	20	20	0	0	20 (33.3)
Chakma	100	8.36 (6.84–9.19)	7 (7.0)	5 (5.0)	4	2	0	0	16 (72.7)
Khumi	100	6.33 (5.04–7.98)	7 (7.0)	14 (14.0)	5	4	0	0	22 (71.0)
Khyang	100	6.70 (4.60–7.94)	6 (6.0)	15 (15.0)	2	0	4	0	25 (80.7)
Lushai	2	6.43 (0.00 to 27.52)*	0 (0.0)	1 (50.0)	0	0	0	0	1 (NA)
Marma	99	8.59 (6.78–9.64)	7 (7.1)	5 (5.1)	4	1	0	0	17 (77.3)
Mro	100	5.94 (4.38–7.23)	11 (11.0)	23 (23.0)	6	9	0	0	29 (65.9)
Tangchangya	100	7.30 (5.65–9.77)	10 (10.0)	12 (12.0)	9	3	0	0	20 (62.5)
Tripura	99	8.18 (6.65–9.77)	2 (2.0)	4 (4.0)	2	0	0	0	14 (87.5)
Total	999	7.03 (5.38–8.69)	90 (9.0)	133 (13.3)	57	42	6	1	217 (67.2)

*arithmetic mean (95%CI)

Median G6PD activity was 1.7U/gHb (24.2% of the AMM) higher among participants with peripheral malaria confirmed by microscopy or RDT compared to those with a negative result (8.6 U/gHb (IQR: 7.5 to 10.4) vs. 6.9 U/gHb (IQR: 5.2 to 8.6)), however the difference was not significant (p = 0.1894) and G6PD activities also did not differ between malaria and non-malaria participants when considering PCR instead (p = 0.6558) (Fig 4).

Fig 4

Distribution of G6PD activity and confirmed genotype per malaria diagnosis* Horizontal lines indicate 100%, 70%, and 30% G6PD activity of the AMM (from top to bottom) *one participant, positive by RDT and microscopy but negative by PCR was excluded.

Overall, 90 (9.0%) participants were identified to be G6PD deficient (<30% enzyme activity), the prevalence being 12.7% (51/401) in males and 6.5% (39/598) in females; p = 0.001. A total of 133 (13.3%) individuals had intermediate G6PD activity based on the population wide AMM, which was present in 12.2% (49/401) of males and 14.0% (84/598) of females; p = 0.465. The proportions of individuals with G6PD deficiency differed significantly with ethnicity (p<0.001) with severe G6PD deficiency (below 30%) present in 26% (26/100) of participants of the Chak ethnicity but only 2.0% (2/99) of those of Tripura ethnicity; Table 2.

A total of 323 participants were genotyped (90 G6PD deficient individuals, 133 G6PD intermediate and 10 randomly selected G6PD normal participants per ethnic group (except Lushai); Table 3. Of the 223 participants diagnosed with G6PD activities below 70%, a G6PD variant was identified in 94 (42.2%) cases, comprising 68/90 (75.6%) G6PD deficient individuals and 26/133 (19.5%) intermediate deficient individuals. A total of 2 (2.0%) males and 10 females (10%) of the randomly selected G6PD normal participants had a detectable G6PD variant. Overall, the Mahidol variant was detected in 99 individuals (93.4% of all identified variants), the Orissa variant in 6 individuals (5.7% of all identified variants) and the Kalyan—Kerala variant in 1 individual (0.9% of all identified variants).

Table 3

Genotyping results per category of G6PD activity measured by spectrophotometry.

Genotype	G6PD activity <30%	G6PD activity ≥30% to <70%	G6PD activity ≥70%	Total
Total with no variant determined (% per row)	22 (10.1)	107 (49.3)	88 (40.6)	217
Total hemi / homozygous Mahidol (% per row)	54 (94.7)	1 (3.8)	2 (3.5)	57
Total heterozygous Mahidol (% per row)	11 (26.2)	21 (50.0)	10 (23.8)	42
Total hemi / homozygous Orissa (% per row)	2 (33.3)	4 (66.7)	0 (0.0)	6
Total homozygous Kerala (% per row)	1 (100.0)	0 (0.0)	0 (0.0)	1
Total	90	133	100	323

In total, 45 males were hemizygous and 12 females homozygous for the Mahidol variant with a median G6PD activity of 0.49 U/gHb (7.0% of the AMM, range: 0.0% to 185.8%); two males and one female had G6PD activities above 30%. In the 42 females identified as being heterozygous for the Mahidol variant, their median G6PD activity was 3.72 U/gHb (52.9% of the AMM, range: 0.0% to 206.5%). 11/42 (26.2%) were G6PD deficient (activity <30%), 21/42 (50.0%) intermediately deficient, and 10/42 (23.8%) had normal G6PD activity (Table 3).

The median G6PD activity among the 3 males hemizygous and 3 females homozygous for the Orissa variant was 2.77U/gHb (39.4% of the AMM; range 3.8%-59.6%) and 2.07 U/gHb (29.4% of the AMM) in one male hemizygous for the Kalyan-Kerala variant (Fig 5).

Fig 5

Distribution of G6PD activity per genotype. Horizontal lines indicate 100%, 70%, and 30% G6PD activity of the AMM (from top to bottom).

A multivariate regression model was generated to predict non-normalized G6PD activity (F(17, 981) = 73.12, p<0.001, adjusted R2 = 0.5513). Significant independent variables in order of contribution that predicted G6PD activity were genotype (except for the Kalyan Kerala variant), Hb concentration, ethnicity, sex and weight. Malaria diagnosis, irrespective of method, was not a significant covariate (S3 Table). There was no significant multi-co-linearity in either the initial or final models.

Discussion

In the first study of G6PDd in the multi-ethnic CHT population, our results revealed that almost 10% of the population had G6PD activities of less than 30%, however the prevalence differed significantly between ethnic groups. Among identified G6PD variants the Mahidol variant accounted for 93% of all cases, similar to earlier reports from the area.

Differences in the prevalence of G6PDd among different ethnicities living in the same area and sharing the same risk of malaria infection had been reported earlier from other sites, however not to the same extent as reported here. PQ based radical cure is contraindicated in patients presenting with G6PD enzyme activity of less than 30% [16]. Using this criterion, only 74% of Chak would be eligible for treatment, compared to 98% of those from the Tripura tribe. Despite a high prevalence of G6PD deficiency and thus associated risk of drug induced hemolysis, there are few reports of PQ induced hemolytic reactions either from the local medical staff (C.S. Phru, personal communication) or in the literature [8,18]. The reasons for this are unclear, but could reflect low prescription of PQ by healthcare providers, underreporting of adverse events due to limited access to healthcare, poor treatment adherence to a prolonged course of PQ, or G6PDd individuals having partial protection against malaria [15,41,42]. In earlier hospital-based malaria surveys conducted in the same area, less than 2% of malaria patients were G6PDd, considerably less than the 9% prevalence we found within the mostly healthy population of this survey. While not statistically significant, the median G6PD activity was almost 25% higher (1.7U/gHb) among patients with malaria diagnosed by microscopy or RDT compared to participants with a negative result. Malaria status was not a significant predictor of G6PD activity irrespective of diagnostic assay applied, however our study was neither designed nor powered to this effect.

Although G6PD genotype is a key determinant of enzyme activity [43], we observed poor correlation between G6PD genotype and phenotype, a phenomenon that reflect both assay variability as well as host and environmental factors [44-49]. Almost 25% of females heterozygous for a known local G6PD variant had activities above 70%, likely due to lyonization patterns in favor of the G6PD normal allele [50]. Interestingly, 5% of participants, hemi- or homozygous for the Mahidol variant, that confers very low G6PD levels, had either intermediate or normal G6PD activities. Although not quantified, study participants may have had recent blood loss due to injury, or substance or pathogen induced hemolysis, thereby replacing older erythrocytes with reduced G6PD activity with younger RBCs with higher activity [51]. Neither can we exclude the possibility that concomitant hemoglobinopathies or high leukocyte counts, may have also confounded measured G6PD activity [37].

The study had some inherent limitations. The panel of five G6PD variants that were genotyped was based on earlier reports from the area [8,26,52]. However, a known G6PD variant was only present in 42% of patients with less than 70% G6PD activity. Given that none of the samples failed PCR, it is likely that the chosen panel did not include all variants present in the region. And even in those in whom a G6PD variant was identified, the correlation between genotype and phenotype was not good, suggesting either erroneous results from spectrophotometry, possibly arising from delays in sample transport to the reference laboratory or procedural errors. However, all samples were tested within seven days, the recommended maximum interval between venipuncture and measurement of G6PD activity [33]. Furthermore, standardized commercial controls were run with every measurement, all samples were tested in duplicate, and results only accepted if both measurements did not differ by more than 10%.

In conclusion, there was a high prevalence of G6PDd in the CHT, but this varied significantly between ethnic groups. Whilst testing should ideally be made widely available to identify individuals with G6PDd, in practice this is often unavailable or unreliable due to financial and logistical constraints. Priority should be given to ensuring the availability of G6PDd testing in communities where the prevalence of G6PDd and thus hemolytic risk is greatest. In areas where testing is not available, consideration should be given to administering primaquine with additional measures to mitigate severe adverse reactions by detecting early signs of hemolysis. Determining the prevalence and variants of G6PDd in malaria endemic areas with different ethnic groups will inform public health interventions to ensure that primaquine radical cure can be provided safely and effectively.

Primer sequences for polymerase chain reactions.

(DOCX)

Click here for additional data file.

Annealing temperatures, post PCR product base pair and restriction enzymes for the PCR-RFLP protocol.

*Optimized temperature.

(DOCX)

Click here for additional data file.

Multivariate regression model to predict non-normalized G6PD activity (n = 999, F17, 981) = 68.27, p<0.001, adjusted R2 = 0.5513).

(DOCX)

Click here for additional data file.

Village names / ethnicity.

(XLSX)

Click here for additional data file.

Underlying database.

(CSV)

Click here for additional data file.

G6PD activity per ethnicity. Horizontal lines indicate 100%, 70%, and 30% G6PD activity of the AMM (from top to bottom).

(TIF)

Click here for additional data file.

28 May 2020

Dear Dr Ley,

Thank you very much for submitting your manuscript "Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh" for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

The manuscript "Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh" was evaluated by 3 reviewers who issued the opinions below.

Assessment of phenotypes for G6PD is really valuable. Unfortunately, the quality of the data as presented so far adds limitations to the manuscript. The limitations need to be discussed to avoid these findings might be then generalized and used inappropriately in the future. To be even more clear, the idea that there is not good correlation between phenotypes and genotypes in G6PD is most often caused by poor laboratory technique or poor sample handling. Technical limitations of the manuscript need to be highlighted.

I recommend to present full data set in an universal format such as csv.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Wuelton Marcelo Monteiro, Ph.D.

Associate Editor

PLOS Neglected Tropical Diseases

Mary Lopez-Perez

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

The manuscript "Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh" was evaluated by 3 reviewers who issued the opinions below.

Assessment of phenotypes for G6PD is really valuable. Unfortunately, the quality of the data as presented so far adds limitations to the manuscript. The limitations need to be discussed to avoid these findings might be then generalized and used inappropriately in the future. To be even more clear, the idea that there is not good correlation between phenotypes and genotypes in G6PD is most often caused by poor laboratory technique or poor sample handling. Technical limitations of the manuscript need to be highlighted.

I recommend to present full data set in an universal format such as csv.

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: Describe the selection of participants for genotyping in the methods section?

Reviewer #2: Yes (please see comments below)

Reviewer #3: (No Response)

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: “… Hill Tracts (CHT) carry the highest malaria burden …” better say people living in the Hill tracts?

Table 1: the column “mean body temp. …” is not helpful and could be omitted or replaced with number of febrile participants detected?”. Same is true for the sentence “The mean axillary body temperature was 36.3°C (95% confidence interval (95%CI): 36.2 – 224 36.3) …”

A figure illustrating the assembly of the participants (consort chart) could be useful?

The regression model is difficult to understand. Please revise the Table 5 so as to show the affected/non affected number of participants for each row in a separate column?

There is something missing in the sentence: “…almost 10% of the population of the had G6PD activities of less than 30%...”

Reviewer #2: Partly please see comments below)

Reviewer #3: (No Response)

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: The authors speculate regarding the low number of disease/symptom presented by G6PD deficient individuals. “Despite a high proportion of the population being at risk of drug induced hemolysis, there are few reports of drug induced hemolytic reactions from local medical staff (B. Ley, personal communication) and the literature”. This requires more explanations. How are these deficient individuals challenged? Do they receive routinely PQ treatment or other triggers for haemolysis? In the absence of information on exposure to triggers for haemolysis it is difficult to assess the outcome.

I am not convinced by the conclusion that testing could/should be conducted according to ethnicity as even in tribes with low G6PD def. prevalence deficient individuals were detected. It may be wiser to recommend routine use of rapid tests/biosensors for wider use of 8-aminoquinoline regimens?

Reviewer #2: Partly please see comments below)

Reviewer #3: (No Response)

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: not applicable...

Reviewer #2: (No Response)

Reviewer #3: (No Response)

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: A very nicely designed study which provides new information on G6PD deficiency essential for the rollout of the radical cure and hence the elimination of vivax malaria.

Reviewer #2: The study by Ley et al. is a well-conducted study about G6PDd prevalence in Bangladesh population. Both genotype and phenotype were screened for subjects from 11 different ethnicities. The novelty is that Ley´s study brings a more comprehensive picture regarding G6PD activity in that population. This is important in the context of radical cure of P. vivax malaria. Whilst some of the data are interesting, the authors need to clarify many points before publication.

Major comments

1. The authors show the mean Hb values for each population in Table 1; however, this data should be presented and analyzed stratified by gender since it differs between men and women.

2. It is not clear why authors included the two individuals of Lushai ethnic group in the analysis since only two subjects are not representative. I suggest removing them and consider analyzing 10 ethnic groups.

3. From the main text or tables I could not find the results for qPCR concerning parasite species. Please add this information to the manuscript.

4. It is clear from previous studies that heterozygous women cannot be accurately identified through G6PD enzyme activity assays. I suggest presenting G6PD activity data in Table 2 and Fig. 2 stratified by gender. Did the authors find significant differences in G6PD activity among ethnic groups stratified by gender? This could be due to different populations presenting different variants of G6PD gene (for variants assayed and for those that could not be assayed in the present study).

5. In the Discussion, the authors claim that malaria status was not found to be a significant predictor of G6PD activity irrespective of diagnostic assay applied. However, it should be made clear that the study was not powered to assess any association between G6PDd and positivity for malaria.

6. In Lines 339-342, it was raised the possibility of a recent haemolytic event to explain inconsistent results for genotype-phenotype correlations. But other factors as concomitant haemoglobinopathies and high leukocyte counts could also lead to a false normal G6PD result (Malaria Journal, 2013; 12:391). Did the authors take these other issues into account? Please, discuss these potential factors.

7. The authors should evaluate the possibility (due to sample size) of estimating the Hardy-Weinberg equilibrium for each ethnic group.

Minor comments

Line 161. Please add the reference of qPCR assays used for molecular diagnosis of malaria.

Lines 166-167. Specify the final concentration of each reagent (or the volume used and the stock concentration) used in the PCR reaction.

Line 219. This data was not shown in Table 1. This information needs to be removed.

Line 305. Multivariate is more appropriate than “multivariable” (please see in Am J Public Health., 2013; 103(1): 39–40). While multivariable analysis refers to statistical models in which there are multiple response variables, multivariate analysis refers to statistical models that have ≥ 2 outcome variables.

Table 2. I suggest including the number of individuals with no variant determined.

Table 2. Correct “Number with with”.

Reviewer #3: This is an interesting manuscript that describes G6PD genotypes and phenotypes in a malaria endemic area of Bangladesh where potentially hemolytic radical cure for Plasmodium vivax will be used. The aim of the study is reasonable but the laboratory analyses have major issues and some of the results interpretation are questionable. The description of the analyzed population and the sample handling is also not very detailed. Probably my biggest question is: why the authors have used the Hb value assessed in the field (up to 7 days earlier) instead of repeating the Hb measurement the same day as spectrophotometric test? This would have been really easy (with hemocue or CBC) and would have given much more reliable results.

Major issues

Line 123 Can the authors show a map with villages and ethnicities? Was the delay from sample collection to analysis in central lab correlated to village of sampling and/or ethnicity?

Line 140 participants <7, how many? Which age range? It is known that G6PD activity is higher in the first 6 months of life, so if small babies were included in the study it should be clearly stated.

Line 155 procedures described earlier need a reference and more info on number of replicates and CV among them, temperature conditions, Quality Controls, accuracy, etc. As far as I have seen in literature, the Randox system does not seem extremely accurate for G6PD activity so the authors might want to give more details to show how it performs in their setting.

Line 156 Hb from the field could be potentially highly unreliable if sample have been shipped without appropriate control of temperature, they could have hemolysed and Hb estimation changes when samples get older (usually after 2-3 days).

Line 192 “The dependent variable was non-normalized G6PD activity (in U/dL), independent variables were backwards selected, initially considering age, gender, body weight, height and temperature, Hb, the delay between sample collection and processing, spectrophotometry result, and G6PD genotype”. Please explain why you used the non-normalized G6PD activity, how gender is independent from G6PD and why body weight and height should be considered variables in this analysis. Please explain why this analysis is even needed.

Line 203 how were the 10 sample per ethnic group chosen? Where they both females and males? Where they from different villages?

Line 239 The delay between sample collection and analysis is rather large; what is written here does not correspond to what written in the discussion in Line 349.

Line 240 “100% G6PD activity was defined as 7.03 U/gHb (interquartile range (IQR): 5.38 – 8.69, total range: 0.73 – 17.17) and median G6PD activity varied significantly with ethnicity (p<0.001).“ Is this the AMM or the Male Median (MM)? The median activity between groups should be compared within the same gender (especially since gender were not represented equally in all groups).

Line 249 “when females were included, the Chak had the lowest median G6PD activity (4.92 U/gHb,

IQR: 1.83 – 7.12)”; unclear what this means. The AMM in the Chak (Table 2) has a very low first quartile, is this really the AMM or the MM?

Line 257-262 and Table 3: these comparisons can only be made if the proportion of females and males in the different groups is the same.

Line 274: “A total of 133 (13.3%) individuals had intermediate G6PD activity, which was present in 12.2% (49/401) of males and 14.0% (84/598) of females; p=0.465.” The authors need to explain how this is possible. The expected number of males with intermediate activity in a population should be very little. Are these data calculated using the overall AMM or the ethnic specific AMM?

Line 281-285 “Of the 223 participants diagnosed with G6PD activities below 70%, a G6PD variant was identified in 94 (42.2%) cases, comprising 68/90 (75.6%) G6PD deficient individuals […]. A total of 12 (12.0%) of the randomly selected G6PD normal participants had a detectable G6PD variant.” I see two problems with results here: 1) 75.6% of deficient individuals had none of the analysed mutations. While this is possible, the authors should bring some comparative data on prevalence of other known mutations in the area (not included in this analysis) to give us an idea of how likely it is that they make up >25% of deficient samples. 2) 12% of G6PD normal subjects had a variant, if they are females there is no problem at all, if they are males this needs to be explained.

Line 288 and Table 4 “In total, 45 males were hemizygous and 12 females homozygous for the Mahidol variant with a median G6PD activity of 0.49 U/gHb (7.0% of the AMM, range: 0.0% to 185.8%)”; same here, based on published data, a hemi or homozygous Mahidol is not expected to have >50% activity, is there any plausible explanation for this result?

Line 305-312 I do not see the point of using a regression model and in particular to use the non-normalized G6PD activity.

Line 339 Although this is a possible explanation, I would like the authors to also bring other more likely explanations for this large proportion of genetically deficient but phenotypically normal subjects. As mentioned before, analysing the Hb in the field few days prior the assessment of G6PD activity might have brought a large variation in the G6PD activities. A test such as Hemocue (or even CBC) would have been pretty easy to run in the equipped central laboratory at the same time as G6PD. The delay in G6PD analysis, the way samples were transported (was the temperature checked during transportation? Did any sample show signs of hemolysis?), whether the sample came from small children or adults, accuracy of laboratory test used, village of origin or sampling day of samples with unexpected G6PD phenotypes…all these are factors that need to be clearly stated in the manuscript and taken into consideration when analysing the results.

Line 355 “Testing should be considered at least for ethnic groups with high proportions of G6PDd and G6PD intermediate individuals, albeit ethnicity specific testing may be difficult to implement in practice.” Since the data show that all ethnic groups analysed, with the exception of “Tripura”, have a prevalence of G6PD deficiency >5%, it seems that a universal G6PD testing strategy would be more appropriate than an ethnic specific one.

Minor issues/comments

Line 36 by reading the methods, it is clear that participants where not randomly enrolled. This should be corrected in the whole manuscript.

Line 146 50% discrepancy seems pretty high

Line 164 how was the subset chosen?

Line 187 why you test for normality of G6PD activity distribution which is expected to be bimodal?

Line 214 suggest eliminate the 2 samples from the Lushai ethnic group

Line 224 “While Hb and body temperature differed significantly among ethnic groups (all p<0.001)”; was sex and age taken into account?

Line 325-334 The authors should give some info on how many subjects get actually treated with long course primaquine in the region; low access to primaquine could completely justify the lack of haemolytic reactions.

Line 331 if it is not significant, it is not higher! I also do not understand this sentence “malaria status was not found to be a significant predictor of G6PD activity irrespective of diagnostic assay applied.”

Line 335 how many papers instead do report a good correlation between G6PD genotype and phenotypes?

Figure1 shows that the distribution in males, although bimodal, has a large number of subjects in the intermediate range

Figure 2 I would show only males

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, PLOS recommends that you deposit laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see https://journals.plos.org/plosntds/s/submission-guidelines#loc-methods

23 Jun 2020

Submitted filename: Ley et al_PLoSNTD_Response to Reviewers_v0.3.docx

Click here for additional data file.

8 Jul 2020

Dear Dr. Ley,

Thank you very much for submitting your manuscript "Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh" for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. The reviewers appreciated the attention to an important topic. Based on the reviews, we are likely to accept this manuscript for publication, providing that you modify the manuscript according to the review recommendations.

Please prepare and submit your revised manuscript within 30 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to all review comments, and a description of the changes you have made in the manuscript.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Thank you again for your submission to our journal. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Wuelton Marcelo Monteiro, Ph.D.

Associate Editor

PLOS Neglected Tropical Diseases

Mary Lopez-Perez

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: (No Response)

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: (No Response)

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: (No Response)

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: (No Response)

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: (No Response)

Reviewer #2: The authors have satisfactorily responded to all my questions and made the necessary changes to the manuscript.

Reviewer #3: The authors have adequately addressed some of my questions but there are a number of issues that have not been resolved and some have arisen during revision. In general, I feel the authors have not explored all the aspects of the questions and comments received from the reviewers.

They state their main aim in line 34 and again line 102: “Aim of this study was to determine the need for routine testing within a multi-ethnic population by assessing the prevalence of G6PDd among the local population”. If the aim was to assess the prevalence of G6PD deficiency, a qualitative test (in the field or in the lab) would have been sufficient. The authors took a brave decision to perform a rather more complicated work in order to bring more details and possibly useful data but then fail to perform the reference test in the correct way (using Hb from previous assessment) and in their main statistical analysis they do not use the right values of G6PD activity (see my comment below). After showing that all ethnic groups have a moderate to high prevalence of G6PD deficiency they reach the conclusion that “practically” G6PD screening might still be done based on ethnicity (even after 3 reviewers suggested that was not the right interpretation).

And while it is useful to know that the laboratory performing the analysis collaborates with WHO and PATH, I believe the comments and suggestions from reviewers where certainly not about the quality or reputation of the laboratory but about the quality of data presented, possibly the study design and quite clearly the conclusions.

Below there are few more comments to specific issues:

1) In my first review I had asked about the accuracy of Randox G6PD assessment and to specify whether testing was carried out in single replicate or duplicate. Accuracy is not reported and there is no mention of duplicates in the methods but there is a mention in line 359-361 of discussion. This should be written in the methods. It is also unclear to me why the haemoglobin was not re-assessed at the laboratory when the samples were analysed for G6PD, which is in my opinion one of the biggest technical limitations of this study. The authors replied that in previous studies they have tested samples kept in laboratory for up to 1 week but I presume they had retested both G6PD and Hb over time.

2) In the multivariate analysis, the variable listed in the methods (line 199-202) are not the same as those listed in the results (315-317). For example, “ethnicity” is missing in the variable listed in methods. And on a more substantial aspect of this analysis:

a) I think the authors are wrong in using the “non-normalized” values of absorbance, because G6PD activity is defined as normalized by either Hb or number of RBCs. The “non-normalized” absorbance is simply not the enzymatic activity.

b) While they make a point of not using Hb-normalized G6PD in order to then analyse Hb as independent variable, they use a lot of variables that are actually collinear (sex and heterozygote genotype, sex and weight, sex and height) and others that might be collinear (for example if a mono-ethnic village was very far away and samples analysis was delayed, then “ethnicity” and “delay to test” variables would be collinear). I believe this should be checked and made clear in the analysis.

3) Lushai ethnic group: a sentence in the text would be enough to report that G6PD deficiency was found in 1 person of this ethnicity, without needing to add this ethnic group in the table.

4) The authors confirm that “Although G6PD genotype is a key determinant of enzyme activity, we observed poor correlation between G6PD genotype and phenotype, a phenomenon that has been described previously (44-47).” The first paper used retrospective hospital data collected during a decade (1993-2003) and phenotype were defined using four different ranges over time; hardly comparable to the current study. The second paper shows similar poor genotype-phenotype correlation and indeed used the same reagents as the current one. I think there are many reasons to believe that the poor correlation observed here and rarely in other settings is due primarily to technical issues rather than reflecting a true biological phenomenon.

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, PLOS recommends that you deposit laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see http://journals.plos.org/plosntds/s/submission-guidelines#loc-materials-and-methods

6 Aug 2020

Submitted filename: Ley et al_PLoSNTD_Response to Reviewers_2_v4.docx

Click here for additional data file.

11 Aug 2020

Dear Dr. Ley,

We are pleased to inform you that your manuscript 'Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh' has been provisionally accepted for publication in PLOS Neglected Tropical Diseases.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Wuelton Marcelo Monteiro, Ph.D.

Associate Editor

PLOS Neglected Tropical Diseases

Mary Lopez-Perez

Deputy Editor

PLOS Neglected Tropical Diseases

***********************************************************

After the return of the modified manuscript, I believe that all requests were met properly. I suggest accepting the manuscript as it is in its latest version.

2 Sep 2020

Dear Dr Ley,

We are delighted to inform you that your manuscript, "Wide range of G6PD activities found among ethnic groups of the Chittagong Hill Tracts, Bangladesh," has been formally accepted for publication in PLOS Neglected Tropical Diseases.

We have now passed your article onto the PLOS Production Department who will complete the rest of the publication process. All authors will receive a confirmation email upon publication.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any scientific or type-setting errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript. Note: Proofs for Front Matter articles (Editorial, Viewpoint, Symposium, Review, etc...) are generated on a different schedule and may not be made available as quickly.

Soon after your final files are uploaded, the early version of your manuscript will be published online unless you opted out of this process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Shaden Kamhawi

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Paul Brindley

co-Editor-in-Chief

PLOS Neglected Tropical Diseases