Association of aflatoxin B1 levels with mean CD4 cell count and uptake of ART among HIV infected patients: A prospective study.

BACKGROUND: Aflatoxin suppresses cellular immunity and accentuates HIV-associated changes in T- cell phenotypes and B- cells.
OBJECTIVE: This prospective study was conducted to examine the association of aflatoxin levels with CD4 T-cell count and antiretroviral therapy uptake over time.
METHODS: Sociodemographic and food data were collected from antiretroviral therapy naïve HIV-infected patients. CD4+ counts were collected from participants' medical records. Plasma samples were tested for aflatoxin B1 albumin adducts, hepatitis B surface antigen, and HIV viral load. Participants were separated into high and low aflatoxin groups based on the median aflatoxin B1 albumin adduct level of 10.4 pg/ml for data analysis.
RESULTS: Participants with high aflatoxin B1 albumin adduct levels had lower mean CD4 at baseline and at each follow-up period. Adjusted multivariable logistic regression analysis showed that higher baseline aflatoxin B1 adduct levels were associated with statistically significant lower CD4 counts (est = -66.5, p = 0.043). Not starting ART and low/middle socioeconomic status were associated with higher CD4 counts (est = 152.2, p<0.001) and (est = 86.3, p = 0.027), respectively.
CONCLUSION: Consistent correlations of higher aflatoxin B1 adduct levels with lower CD4 over time indicate that there is an independent early and prolonged effect of aflatoxin on CD4 even with the initiation of antiretroviral therapy. The prospective study design, evaluation of baseline and follow-up measures, extensive control for potential confounders, and utilization of objective measures of aflatoxin exposure and CD4 count provide compelling evidence for a strong epidemiologic association that deserves careful attention in HIV care and treatment programs.

Introduction

Aflatoxins are carcinogenic metabolites produced in food crops primarily by two species of Aspergillus fungi, namely A. flavus and A. parasiticus [1]. These fungi are ubiquitous in soil and on vegetation and produce toxins in a variety of staple food crops such as cereals (e.g. maize, millet, rice, and wheat), legumes and oilseeds (soybean and groundnuts), and a number of other crops such as tree nuts, root and tuber crops and spices [2]. In tropical and subtropical regions of the world, the high temperatures and high humidity favor aflatoxin contamination of crops. In addition, poor post-harvest practices and storage and marketing conditions in many countries of sub-Saharan Africa, Latin America, and South and Southeastern Asia result in fungal proliferation and accumulation of aflatoxin levels in crops that exceed the 20 μg/kg limit for total aflatoxins in foods set by the United States [3]. Consequently, outbreaks of acute aflatoxicosis have occurred with the death rate as high as 39.4% [4-6].

Of the four main aflatoxin chemotypes (AFB1, AFG1, AFB2, AFG2), AFB1 is generally the predominant and most toxic form [7]. AFM1 is a major harmful metabolite of AFB1 that is excreted in urine and milk [8, 9]. Although chronic exposure to aflatoxin is known predominantly for its role in the development of liver cancer in humans, especially in those with hepatitis B and hepatitis C infections [10-13], several animal and human studies show that aflatoxins modulate the immune system, mainly causing immune suppression [14-16]. Studies conducted in humans with chronic exposure to dietary aflatoxin show that blood levels of aflatoxin B1 albumin adducts (AF-ALB) are associated with antibody and cellular immunity [17-19].

Human immunodeficiency virus (HIV) infection suppresses the immune system and results in the development of acquired immunodeficiency syndrome (AIDS) unless those infected are treated with antiretroviral therapy (ART). Investigation of immune status of HIV-positive people chronically exposed to aflatoxin in their diets, showed significantly lower percentages of CD4+ T regulatory cells, naïve CD4+ T-cells, perforin-expressing CD8+ T-cells, and B-cells in those with high AF-ALB compared to those with lower AF-ALB levels [19]. These findings indicate that changes in T-cell phenotypes and B-cells that occur in HIV are amplified by aflatoxin exposure. In cross-sectional studies, we also found consistent strong associations between high AF-ALB levels and high HIV viral loads, indicating that aflatoxin exposure may contribute to high viral loads and faster progression to AIDS [20, 21]. Therefore, we conducted this prospective study among ART-naïve HIV-positive asymptomatic Ghanaians with mean baseline CD4 counts of 631±281 cells/mm3 of blood to examine the association of AF-ALB levels in blood with changes in CD4 cell count and uptake of ART over a five-year period.

Methods

Study site, participants and data collection

This study was conducted in the Kumasi South Regional Hospital (KSRH) and Bomso Hospital (BH) in Kumasi in the Ashanti Region of Ghana. KSRH is a large hospital that serves 56 communities of approximately 400,000 people. BH is a 163 bed private hospital located in close proximity to KSRH that provides comprehensive HIV care, treatment, and support and works closely with KSRH. Potential study participants were HIV-positive patients with CD4 count ≥300 cells /mm3 blood who were asymptomatic had received no antiretroviral therapy in accordance with the World Health Organization 2006 treatment guideline [22]. Clinic staff told patients attending the hospitals of the study and asked if they would be interested in participating. Patients who expressed interest in participating were introduced to the study team who explained the study and asked them to read the informed consent and ask questions. After all patient questions were answered to their satisfaction those willing to participate gave written informed consent. An interviewer-administered questionnaire that was developed for the study has been included as S1 File. It contained questions related to sociodemographic, health, food acquisition, storage and consumption practices, awareness of aflatoxin and information on HIV/AIDS and sexually transmitted diseases. Eight staff members (doctors, nurses, and administrative personnel) from both hospitals reviewed the questionnaire for understanding, clarity and cultural appropriateness after which it was revised. It was then pilot tested among six clinic patients similar to the ones recruited for the study and revised before use. To ensure confidentiality, the interviews were conducted in private rooms at the hospitals. Data on HIV diagnosis date and CD4+ T cell count were collected from medical records of the patients. The mean, standard deviation (SD) and range of CD4 cells at recruitment were mean ± SD = 618.10 ± 284.32; range 301–1616. A 20 mL blood sample was collected from each patient and plasma was prepared and tested for AF-ALB, HIV viral load, and HBV surface antigen (HBsAg). CD4 and ART initiation data were collected for up for five years post-recruitment.

Ethical approval

Ethical approval for the study was obtained from the Institutional Review Board at the University of Alabama at Birmingham (UAB) and the Committee on Human Research, Publications and Ethics at the School of Medical Sciences, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana. Written informed consent was obtained from each participant.

Quantification of AFB1-lysine adducts

Aflatoxin B1-lysine adducts in plasma of participants was measured using a modified High Performance Liquid Chromatography (HPLC)-fluorescence method developed by Qian, et al. (2013) and outlined in detailed in Jolly, et al. (2015) [23, 24]. Aflatoxin B1-lysine adducts indicate exposure to aflatoxin in the previous 2–3 months [25].

Quantitation of HIV viral load using the Roche HIV-1 RNA Assay

HIV-1 RNA in the plasma of study participants was measured at the UAB Hospital Laboratory using the Roche COBAS Ampliprep/COBAS TaqMan HIV-1 Test, version 2.0 according to the manufacturer instructions (Roche Molecular Systems, Inc, Pleasanton, CA). The test has been approved by the United States Food and Drug Administration, and quantifies HIV-1 RNA based on the co-amplification of the HIV LTR (Long Terminal Repeat) and gag. This method has been outlined in detail in Jolly et al. (2013) [21].

Test for antibodies to HBV surface antigen

Antibodies to HBsAg in plasma samples were determined using the Bio-Rad Enzyme Immunoassay according to the manufacturer’s directions (Bio-Rad, Redmont, WA, USA) and previously outlined by Jolly et al. (2011) [20].

Statistical analyses

Two hundred and ninety-five participants who had complete data on baseline AF-ALB levels are included in the current analysis. Participants were divided into high and low AF-ALB levels based on the median AF-ALB level of 10.4pg/mg. We describe sociodemographic characteristics, food consumption patterns, and clinical variables of participants by high (≥10.4 pg/mg) and low (<10.4 pg/mg) AF-ALB levels at baseline using chi-square tests for categorical variables, fisher’s exact test for cell counts <5, and t-tests to compare group means for continuous variables. Principal components analysis (PCA) was conducted to assess lasting household indicators such as housing type, plumbing, water, and electricity to determine the socioeconomic status (SES). An SES score was attained from the PCA analysis by weighting each indicator by the coefficient of the first principal component, with each member of the household being assigned the same SES. Further, SES was categorized into tertiles ranging from lowest to highest. Mixed methods analysis was employed in Statistical Analysis System (SAS; SAS Institute Inc., Cary, North Carolina, USA) to analyze this longitudinal data to assess statistically significant predictors of CD4 levels over time from baseline to 5 years. Initial adjustments included sociodemographic variables (age, gender, and SES score), season (dry and rainy), baseline AF-ALB levels and knowledge of HIV-positive status. A second model was additionally adjusted for alcohol intake and food consumption patterns. A final model included further adjustment for health status, viral load, HBV status, and ART. For all analyses, p-values ≤0.05 were considered as statistically significant. All statistical analyses were performed with SAS 9.4.

Results

The mean AF-ALB level for the participants in this study was 14.75 pg/mg (standard deviation ±15.61); median = 10.37, range = 0.20–109.87, and inter-quartile range = 4.67–19.56 pg/mg. The median AF-ALB was used to separate participants into high and low AF-ALB groups (Table 1). A majority of participants were 30–39 years of age and married/cohabitating. A significantly higher proportion of males had high baseline AF-ALB levels as compared to females (73.0% vs. 44.0%; p<0.001). Significantly higher mean viral load (176,264 vs. 73,900; p = 0.010) was observed among participants with high AF-ALB as compared to those with low AF-ALB (Table 1). Among participants with high AF-ALB, 61.6% reported buying 20% or more of their food as compared to 49.6% of participants with lower AF-ALB (p = 0.045). A higher proportion of participants with high AF-ALB (44.2%) reported storing 25% to >50% of maize as compared to 25.9% of participants with low AF-ALB (p = 0.005). A higher percent of participants with high AF-ALB also reported storing maize for 3 to >6 months (41.0% vs. 23.2%; p = 0.006).

Table 1

Descriptive statistics of the study participants by median baseline aflatoxin albumin adduct (AF-ALB) levels.

	High AF-ALB (> = 10.4pg/mg) N = 147	Low AF-ALB (<10.4pg/mg) N = 148	P value
Age			0.972
18–29	41 (27.9)	43 (29.0)
30–39	67 (45.6)	67 (45.3)
40 and above	39 (26.5)	38 (25.7)
Gender			<0.001
Female	102 (69.4)	132 (89.2)
Male	45 (30.6)	16 (10.8)
Marital status (missing = 6)			0.735
Married/ Cohabitating	100 (69.9)	100 (68.5)
Separated/Divorced/Widowed	23 (16.1)	21 (14.4)
Single	20 (14.0)	25 (17.1)
Socioeconomic status (missing = 13)			0.380
Low	48 (33.8)	40 (28.6)
Middle	52 (36.6)	48 (34.3)
High	42 (29.6)	52 (37.1)
Religion (missing = 6)			0.981
Christian	124 (86.1)	125 (86.2)
Muslim/Other	20 (13.9)	20 (13.8)
Proportion of food grown (missing = 21)			0.243
None	77 (56.2)	85 (62.0)
<20%	28 (20.4)	31 (22.6)
> = 20%	32 (23.4)	21 (15.4)
Proportion of food bought (missing = 18)			0.045
<20%	53 (38.4)	70 (50.4)
> = 20%	85 (61.6)	69 (49.6)
Proportion of maize stored (missing = 18)			0.005
<25%	77 (55.8)	103 (74.1)
25%-49%	55 (39.9)	31 (22.3)
> = 50%	6 (4.3)	5 (3.6)
Months maize stored (missing = 18)			0.006
0–2	82 (59.0)	106 (76.8)
3–5	52 (37.4)	29 (21.0)
> = 6	5 (3.6)	3 (2.2)
Groundnut consumption (missing = 20)			0.278
Never	9 (6.5)	7 (5.1)
Once or less a week	63 (45.3)	68 (49.6)
2–3 times a week	48 (34.5)	35 (25.5)
Everyday	19 (13.7)	27 (19.7)
Maize consumption (missing = 21)			0.050
Never	1 (0.7)	2 (1.5)
Once or less a week	36 (25.9)	51 (37.5)
2–3 times a week	46 (33.1)	27 (19.9)
Everyday	56 (40.3)	56 (41.1)
Drink alcohol (missing = 2)			0.487
No	134 (91.8)	138 (93.9)
Yes	12 (8.2)	9 (6.1)
Health status (missing = 5)			0.052
Poor	4 (2.8)	0 (0.0)
Average	58 (40.3)	72 (49.3)
Good	82 (56.9)	74 (50.7)
Hepatitis B virus status (missing = 8)			0.345
Negative	128 (87.1)	134 (90.5)
Positive	19 (12.9)	14 (9.5)
HIV viral load# (missing = 2)	176264 (453675)	73900 (129910)	0.010
CD4 T cell count #	613.6 (281.6)	644.7 (277.4)	0.340
On antiretroviral therapy (ART)			0.147
No	105 (71.4)	94 (63.5)
Yes	42 (28.6)	54 (36.5)
Season			0.865
Dry	79 (53.7)	81 (54.7)
Rainy	68 (46.3)	67 (45.3)

N = 306 with Missing baseline aflatoxin levels for 11

*SES rank includes: education, employment, housing type, plumbing, water, electricity, house material

#Mean and SD

Table 2 shows the unadjusted mean CD4 estimates during baseline, and at 1 to 5 years of follow-up visits by sociodemographic variables and season. A lower mean CD4 count was observed among those with high levels of baseline AF-ALB and at each collection period as compared to those with low AF-ALB levels, with the lowest levels during the 5th year (533.1±41.5 vs. 755.0±38.8). Fig 1 shows the mean CD4 estimates over time by baseline AF-ALB levels. Participants who started ART during the study period had a lower baseline mean CD4 count as compared to those who did not start ART (483.9±24.8 vs. 702.4±21.7); although, by year 5 of follow-up, the mean CD4 count was higher among those on ART as compared to those who did not start ART (666.6±34.5 vs. 526.5±41.3). Additionally, males and those of high SES were observed to have lower mean CD4 counts. Older participants (≥40 years) had lower mean CD4 counts at baseline and during years 1–3. Lower mean CD4 counts were also observed during the dry seasons at baseline and years 1–4 as compared to rainy seasons (Table 2). Fig 2 shows the mean CD4 estimates over time by baseline AF-ALB levels and ART initiation. Fig 3a and 3b show the CD4 distribution with time of ART initiation for the low and high AF-ALB groups, respectively. At baseline, the mean CD4 level was high for the study group and no participant was on ART. At year one the mean CD4 had dropped and 39 participants (15 in the high and 24 in the low AF-ALB groups) started ART, while at year two the mean CD4 was at its lowest and an additional 33 participants (12 in the high and 21 in the low AF-ALB groups) started ART. At year three, the mean CD4 was high but about 16 new participants (3 in the high and 13 in the low AF-ALB groups) started ART. At year four the mean CD4 was at its highest for the group with 8 new participants (3 in the high and 5 in the low AF-ALB groups) starting ART, and at year five no new participant started ART. At year 5, the mean CD4 was much higher than the baseline level for the low AF-ALB group but lower than baseline for the high AF-ALB group.

Table 2

Unadjusted mean estimates for CD4 count over the study period.

	Baseline	1 Year	2 Years	3 Years	4 Years	5 Years
	Mean±Std Err	Mean±Std Err	Mean±Std Err	Mean±Std Err	Mean±Std Err	Mean±Std Err
AF-ALB pg/mg (baseline)
Low	648.0±25.1	618.4±34.8	610.7±32.1	680.8±30.2	741.0±30.0	755.0±38.9
High	608.5±23.5	611.9±33.1	554.7±30.9	626.0±29.1	629.1±33.9	533.1±41.5
On ART
No	702.4±21.7	695.8±31.8	668.3±30.4	704.5±28.3	627.3±34.7	526.5±41.3
Yes	483.9±24.8	480.5±32.0	468.8±28.2	553.5±26.7	677.4±30.4	666.6±34.5
Sex
Female	651.2±19.0	631.4±27.2	594.5±25.1	654.8±23.6	693.4±27.2	651.3±31.5
Male	548.9±39.8	538.1±52.3	569.5±49.5	620.2±46.7	631.7±59.3	579.8±67.4
Age (years)
18–29	703.3±32.9	677.0±49.9	610.1±49.8	644.1±44.2	637.0±52.6	581.7±57.2
30–39	628.4±26.1	622.7±35.8	630.3±33.7	660.2±31.9	691.0±38.3	629.1±47.8
≥40	552.4±31.1	528.1±41.8	511.3±36.4	629.8±35.3	699.0±39.3	684.6±43.2
Socioeconomic status
Low	638.0±33.3	640.4±47.0	595.2±43.7	709.4±41.8	711.8±47.4	716.0±56.1
Middle	654.0±30.6	664.3±42.6	642.4±40.9	662.6±38.6	675.4±43.2	682.7±50.0
High	609.2±27.5	561.7±37.1	552.1±34.0	608.5±31.3	639.5±39.7	549.5±44.1
Season
Dry	584.8±20.8	575.6±29.9	546.2±28.4	611.7±25.7	632.5±29.6	676.1±36.8
Rainy	683.4±27.9	657.6±38.3	637.1±34.9	692.7±33.9	741.9±40.8	615.8±43.6

ART = antiretroviral therapy

Fig 1

Mean CD4 estimates over time by baseline AF-ALB levels.

Fig 2

Mean CD4 estimates over time by baseline AF-ALB levels and ART status.

Fig 3

(a) CD4 count distribution with start time for ART for the low AF-ALB group and (b) CD4 Count Distribution with Start Time for ARTfor the high AF-ALB group.

Table 3 presents the results of the adjusted multivariable models. Model 1 shows that higher baseline AF-ALB was associated with statistically significant lower CD4 counts (est = -66.5, p = 0.043); not starting ART was associated with higher CD4 count (est = 152.2, p<0.001). The ART results remained significant after adjustment for other confounders and clinical variables in Models 2 and 3. Low and middle SES was associated with higher CD4 counts in all models. When adjusted multivariable logistic regression models were run with AF-ALB as a continuous variable, in addition to the results reported above female gender was associated with higher CD4 count.

Table 3

Multivariable adjusted models of mean CD4 estimates over the study period.

	Model 1		Model 2		Model 3
	Beta estimates	P value	Beta estimates	P value	Beta estimates	P value
AF-ALB pg/mg (baseline; high and low based on median of 10.4 pg/mg)		0.043		0.117		0.085
High	-66.5		-57.4		-61.4
Low	Ref		Ref		Ref
On ART		<0.001		<0.001		<0.001
No	152.2		130.9		125.2
Yes	Ref		Ref		Ref
Gender		0.090		0.053		0.060
Female	71.2		92.8		93.0
Male	Ref		Ref		Ref
Age		0.349		0.878		0.812
18–29	17.6		-26.4		-26.8
30–39	6.9		-31.9		-29.7
40 and above	Ref		Ref		Ref
Socioeconomic status		0.027		0.028		0.014
Low	86.3		79.9		95.1
Middle	63.3		23.7		36.1
High	Ref		Ref		Ref
Season		0.137		0.248		0.260
Dry	-45.2		-31.8		-31.2
Rainy	Ref		Ref		Ref

1Adjusted for age, gender, SES rank, art status, season, aflatoxin level at baseline, knowledge of HIV-positive status

2Adjust for model 1 variables + alcohol consumption + food consumption patterns

3Adjust for model 2 variables + health status + viral load + Hepatitis B status

ART = antiretroviral therapy

Discussion

This study was conducted in a major metropolitan area of Kumasi in the Ashanti Region of Ghana where people are at high risk for exposure to aflatoxin in food. Maize and groundnuts are staple crops in Kumasi, with maize being the principal crop and these crops having the highest aflatoxin contamination [26-28]. Most of the food that participants purchased was grown in rural areas and traded in urban markets. The purchased maize may vary in quality and level of aflatoxin contamination depending on the time reaped and the post-harvest processing and storage methods prior to selling [29]. Additional storage over a longer period under hot and humid conditions by study participants would likely result in fungal proliferation, buildup of aflatoxin levels, and greater intake in food [30].

AF-ALB adduct levels in our study participants ranged from 0.20–109.87 pg/mg. AF-ALB is a reliable biomarker that has been used as a standard for assessment of population exposures. It represents accumulation of adducts from repeated (chronic) exposure to aflatoxin over a 2–3 month period and has been used in epidemiological and clinical intervention studies on aflatoxin exposure in humans in different countries. Studies conducted in regions of the world at high risk of aflatoxin exposure show that more than 95% of people are positive for AF-ALB adducts with concentrations ranging from 3–5 pg/mg albumin to >1000 pg/mg [31].

The most significant finding from this prospective study is that adjusted multivariable logistic regression analysis showed that higher baseline AF-ALB was associated with significantly lower CD4 counts. The unadjusted data showed that higher AF-ALB was associated with lower mean CD4 at baseline and at each follow-up time point over the 5-year period, with the lowest mean level during the fifth year. This indicates that the effect of aflatoxin occurs early in HIV infection and remains consistent over time even with the initiation of ART. In the third and fourth years of the study as more participants initiated ART, the mean CD4 levels increased but the increases were less among those with high AF-ALB compared to those with low AF-ALB and the difference was much greater at year 5.

Participants reporting higher SES had lower mean CD4 counts at all time-points. This was a significant finding in all of the multivariable logistic regression models. One possible explanation could be that these higher SES participants were diagnosed with HIV after a longer period of HIV infection and/or took charge of their own health care for a longer period before attending the public clinic for HIV care. We observed that in the early to mid-2000s, HIV infected people with better economic means would attend private clinics/hospitals to avoid disclosure of their HIV-positive status. However, there was an agreement at the United Nations General Assembly High-Level Meeting on AIDS in 2006, to scale up HIV prevention, treatment, care, and support services. In response, the Ghana AIDS Commission developed the National Strategic Framework outlining targets to increase ART coverage to 60% and increase the number of persons receiving HIV care by 200% by 2013 [32]. This resulted in the establishment of programs for the provision of ART in public hospitals and health centers in districts in all ten regions of Ghana and removal of the availability of ART from private clinics/hospitals. HIV infected people of higher SES may have taken a longer time to attend the public facilities for HIV care. ART was provided free of cost to all HIV-positive patients in this study who met the WHO recommended guideline for ART in 2009 and updated in 2010 [22].

When adjusted multivariable logistic regression models were run with AF-ALB as a continuous variable, male gender was found to be significantly associated with lower CD4 count. Lower CD4 counts could be due to a longer time of HIV infection before diagnosis among males who access healthcare facilities less frequently than females [33]. However, the models were adjusted for knowledge of when participants knew of their HIV-positive status. Interestingly, a significantly higher percent of males also had high AF-ALB levels at baseline and at each of the 5-year follow-up time points when compared to females, and in previous studies, males were found to have higher AF-ALB levels than females [24, 34]. Animal studies have shown greater effects of mycotoxins on feed intake and weight gain and in clinical and immunological parameters among males [35-37]. It is possible that there is greater susceptibility of males to aflatoxin related to differences in metabolism of mycotoxins by the liver as suggested by previous authors [38, 39] and this may be an indication of the greater immunological effect of aflatoxin among HIV infected males.

It is understandable that participants who initiated ART during the study had significantly lower baseline mean CD4 counts than those who did not initiate ART and that by year 5 of follow-up, the mean CD4 count was higher among those on ART when compared to those who did not initiate ART. Although ART is available in most countries of the developing world where people are chronically exposed to aflatoxin, the number of people who need ART globally has increased as the WHO guidelines change and have become more difficult to meet. In 2018, 62% of HIV-positive people globally and 64% in the WHO Africa Region who needed ART were able to receive it [40]. Therefore, more than one-third of people who need ART are not receiving it. In addition, the pool of people who will need ART will continue to increase as new HIV infections continue to occur; 1.7 million new infections occurred in 2018, of which 800,000 occurred in sub-Saharan Africa) [41]. Furthermore, although deaths from HIV-related illnesses have decreased from the peak of 1.7 million in 2004, 770,000 people died of AIDS-related illnesses mostly in low- and middle-income countries in 2018 [41]. Since our studies indicate that aflatoxin contributes to an increase in HIV viral load, changes in immune status, and CD4 decline, the full impact of ART is not being, and will not be attained in countries and among those with chronic high exposure to aflatoxin.

Further, HIV-positive people in these high aflatoxin exposure countries who are not diagnosed early, as well as those who do not receive ART, will die faster. It has been shown that HIV-positive people in high aflatoxin-exposure countries progress to AIDS and die more rapidly than those living in high income countries with strict regulation of aflatoxin levels [42]. The median time from HIV seroconversion to clinical AIDS was 11 years for HIV-positive people from Europe, North America, and Australia compared to only 7.4 years in a group of Thai soldiers [42]. A West African study showed that the mean time from HIV seroconversion to clinical AIDS ranged from 5 to 7.2 years depending on the HIV-1 subtype [43]. Since the effect of aflatoxin on increasing HIV viral load and decreasing CD4 is more pronounced in men, and men are more likely to be diagnosed with HIV at a later stage of infection than women, the adverse effects of HIV and disease progression among men can be expected to be greater in aflatoxin exposed areas.

The WHO has set global targets for the elimination of HIV as a public health threat by 2030 [44]. However, the decline in the number of people newly infected with HIV is too slow to reach the goal of 500,000 new infections set for 2020. Combination prevention targets such as increase in voluntary male medical circumcision, condom use, pre-exposure prophylaxis, and harm-reduction services continue to be insufficient in stemming the tide of new infections and financial resources available for the AIDS response have declined. While greater efforts are needed to tackle difficult aspects of HIV that prevent its elimination as a major public health problem, aflatoxin contamination of crops, and dietary exposure to the toxin is a specific problem associated with HIV infection and progression that is overlooked. Several agricultural, dietary and clinical strategies to reduce aflatoxin exposure and its adverse effects on health have been proposed [45]. The cost-effectiveness of these methods has also been elaborated in detail; however, these methods have not been adopted to any extent in countries most affected by aflatoxin exposure [46]. Policy makers, health officials, and researchers should apply this information if they are to be successful in current efforts in combating HIV and other major health problems.

Limitations

This study has certain limitations that should be considered in interpreting the results. First, a convenience sampling method was used that is prone to inherent bias in representation. As such, the sample may not be representative of the study population and the results may not be generalizable to the population being studied. However, convenience sampling enabled us to study the population in a relatively expedient way and to obtain novel results that would not have been possible otherwise. Secondly, we were unable to evaluate aflatoxin levels in food consumed by study participants, examine their nutritional status, or assess their exposure to other mycotoxins or environmental toxins. In addition, the survey data on food grown, maize storage, and food consumption is likely to have recall bias. However, there are several strengths of this research as well. Foremost, the prospective design allowed us to observe changes in CD4 levels over time and to examine CD4 levels by sociodemographic (age, sex, and SES) and seasonal factors that could not be done using a cross-sectional design. Additionally, the use of the AF-ALB biomarker is a direct and valid measure of aflatoxin exposure in participants over the previous two to three months. Finally, the relatively homogenous dietary pattern among study participants allows for greater generalization of our findings to others in the population.

This study advances previous work by revealing significant consistent correlations between higher AF-ALB and lower CD4 counts every year over a 5-year period, in a unique cohort of ART-naïve HIV infected asymptomatic individuals, from across all SES. This strongly indicates that the effect of aflatoxin occurs early in HIV infection and remains consistent over time even with the initiation of ART. Our prospective study design, evaluation of baseline and follow-up measures, extensive control for potential confounders, and utilization of objective measures of AF-ALB and CD4 counts provide compelling evidence for a strong epidemiologic association that deserves careful attention. Guidelines and assessments to minimize chronic exposure to aflatoxin among people living with HIV should be incorporated into HIV care and treatment programs for optimum effect of ART and the healthiest survival of those affected.

Conclusions

The finding of the association of significantly higher AF-ALB levels with lower CD4 counts at baseline and over time is a distinctive contribution to the literature on the effect of aflatoxin on CD4 levels in HIV-infected people. The results indicate that aflatoxin has an immunological effect that contributes to decrease in CD4 and that the effect of aflatoxin occurs early in HIV infection and remains consistent over time even with the initiation of ART. Although the mean CD4 levels increased when patients initiated ART, after year four the CD4 dropped among those with high AF-ALB compared to those with low AF-ALB.

Multivariable adjusted models of mean CD4 estimates over the study period (with aflatoxin at baseline as continuous variable).

(DOCX)

Click here for additional data file.

Baseline questionnaire aflatoxin and health status in HIV disease.

(PDF)

Click here for additional data file.

Study dataset.

(XLSX)

Click here for additional data file.

6 May 2021

PONE-D-21-00637

Association of Aflatoxin B1 Levels with Mean CD4 Cell Count and Uptake of ART among HIV Infected Patients: A Prospective Study

PLOS ONE

Dear Dr. Jolly,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. In particular, patient's description and statistical analysis as well as a number or other points that may be of importance.

Please submit your revised manuscript by 3 months. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see:  http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at  https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Isabelle Chemin, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

4. Thank you for submitting the above manuscript to PLOS ONE. During our internal evaluation of the manuscript, we found significant text overlap between your submission and the following previously published works.

- https://moam.info/mycotoxins_5a3c12051723dd42662aa122.html

- https://doi.org/10.3390/toxins7124868

- https://doi.org/10.2217/fmb.13.166

We would like to make you aware that copying extracts from previous publications, especially outside the methods section, word-for-word is unacceptable, even for works which you authored. In addition, the reproduction of text from published reports has implications for the copyright that may apply to the publications.

Please revise the manuscript to rephrase the duplicated text, cite your sources, and provide details as to how the current manuscript advances on previous work. Please note that further consideration is dependent on the submission of a manuscript that addresses these concerns about the overlap in text with published work.

We will carefully review your manuscript upon resubmission, so please ensure that your revision is thorough.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This clinical prospective study was among ART naïve HIV+ patients in Ghana over 5 years between 2009-2013 with relatively high baseline CD4 count with mean in the 600s (wide range of 300-1616). Participants were grouped into two groups, high AF-ALB vs. low AF-ALB levels. The authors observed that higher AF-ALB group at each time point was associated with lower CD4 count and even with multivariable analysis, this association was significant. Data on diet and sociodemographics were based on questionnaires and HIV disease and diagnosis (including CD4 and ART initiation) were based on medical record. Blood was collected for measurement of AF-ALB level (quantifies level in the past 2-3 months) through HPLC, as well as HIV-1 RNA viral load.

The study placed participants into high (63%) and low (37%) AF-ALB groups. The results showed that higher level AF-ALB group, there was a greater proportion of males compared to females; higher mean VL and lower CD4 count observed compared to lower group. Higher level also was correlated with storing maize for longer period of time, buying food.

Overall, the hypothesis and data supporting it is intriguing; reporting a significant observation that higher AF-ALB group associated with lower CD4 count at baseline and continued through year 1-4. If validated, this could be clinically important in guiding food consumption/storage guidelines to minimize exposure to high levels of AF-ALB exposure among PWH.

The study had several important observations:

• The cohort was unique, in that all were treatment naïve and across socieoconomic status with good spread in age groups. There was some uneven distribution in low/middle SES group between the two groups, with more in the high AF-ALB group (67% vs. 60%) which may effect the endpoint measured, like CD4, access to ART and baseline VL.

• Lower CD4 count significantly observed in those with baseline high AF-ALB levels which persisted at each collection period gives a consistent correlation between the two factors.

• Interesting finding of lower CD4 count observed at baseline and throughout years during dry season, which previous study had showed higher AF-ALB levels in dry season. This is intriguing, but could be confounded by many factors not accounted for in this study.

Minor criticisms:

• Survey data about food grown, maize storage and food consumption likely has significant recall bias and unclear if these questionnaire had not been previously validated to be accurate measures.

• More lower + middle SES in the high AF-ALB group may bias results although comparison of SES between groups was not statistically different based on authors analysis. However, if the group was divided into low+middle vs. high there is likely difference between the two groups based on AF-ALB levels. For example, baseline VL was 200K in high vs 83K in low AF-ALB group. This could be biased by SES and related assess to care and ART which is suggestive of greater proportion of high socioeconomic status in the low AF-ALB group (28 vs. 34%). Similarly, lower SES is related to high AF-ALB exposure, but low SES likely means less access to ART or care and so allows, therefore, augmenting the CD4 difference.

• Data needs to be presented on proportion in each group on ART vs. not on ART at each year of follow up as this ART frequency in each group is unclear and could certainly be biasing the CD4 observations. Figure 3 shows ART frequency but not broken down by frequency in each AF-ALB group, which is the basis of all the comparisons.

• Figure 3: Unclear why CD4 dropping in this cohort at year 5 when each year more people were started on ART, this should continue to go up, thus suggesting groups of patients are not being started on ART and therefore will dropping CD4 counts.

• Discussion mentioned many participants were females in “antenatal clinics”; female hormone and pregnancy is known to affected CD4 and VL count and therefore may again bias data. It would be important to know how many of participants were pregnant and recent post-natal because of effects of pregnancy hormones on CD4 count.

Major criticism:

Overall, the study reports a significant association between high AF-ALB level and lower CD4 baseline and recovery. It is purely a correlative study and does not establish causality even though it has the samples and cohort to do so. A major issue with the study is the heterogeneity of the two groups in terms of HIV status (nadir CD4 count/ years of infection, baseline CD4 count, ART status and years on ART while in study) and the unclear rationale for dividing into two groups based on cutoff of 15pg/mg of AF-ALB. I am concern of the strengths of the associations found in this study as these are all important confounders which could have biased results to show an effect when in fact, biologically they may not be.

For example, conclusion that males had lower CD4 count at baseline and then at each follow up timepoint. If they started at lower T cell, they will continue to be lower regardless of AF-ALB level. Although this difference can be due to AF-ALB effect, this observation could simply be due to the fact that males in this cohort were infected for longer periods. Again, showing that duration of HIV infection of each participant (or marker such as nadir CD4 count) would need to be considered in the analysis. Similarly, the observation that age was associated with lower mean CD4 count in the study could again be simply due to difference in duration of HIV infection (which the author acknowledges but if this data was collected then could be adjusted for in analysis). The two groups would have to matched or controlled for differences in baseline CD4 count.

The second major criticism is that the study treated AF-ALB levels as a categorical or dichotomous variable (high vs. low) instead of continuous variable which would have been more informative as I would argue that seems more clinically relevant as we expect in real-life that AF-ALB levels would run a wide range and there will be a concentration effect if there is indeed an immunosuppressive effect on CD4 cells. It is unclear why the 15pg/mg was used as the cut-off to determine the high vs. low categories as the author never described the significance of this chosen level. It would be interesting to run the analysis with AF-ALB levels as a continuous variable and assess for correlation with CD4 count, but need to control duration of HIV infection or nadir CD4 count.

The third major criticism is that the study does not control or separate out analysis of those on ART, many started during study period. They observed that HIV VL was higher and CD4 cell count lower in the high AF-ALB group at baseline but this could be accounted for by the greater proportion of those NOT on ART (69% vs. 66.3%). Looking at CD4 level change over time is important, but not as a group (AF-ALB high vs. low). The difference proportion in each on ART and/or ART initiation in the middle of study is not accounted for and still treated as two groups. It would be more accurate to look at trajectory of CD4 count each group (ie, see if there is difference in rise of CD4 count after 1 years, 2 year, 3 years, etc after ART initiation between AF-ALB high vs. low groups). Making this comparison with a heterogenous and unevenly distributed groups (ART frequency in Figure 3 should be divided into the AF-ALB groups to compare proportion started in each group) is not accurate.

Fourth major criticism of the study is that this is a correlative study and does not establish or even suggest causality. To make a more casual association, authors could assess change in level of AF-ALB in each patient to change in T cell frequency or even test for function (activation/exhaustion markers, etc) over time. This could have been done with this prospective longitudinally followed cohort of 5 years.

**********

6. 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

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment 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. Registration is free. 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 PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

15 Jul 2021

Isabelle Chemin, PhD

Academic Editor

PLOS ONE

Re: PONE-D-21-00637

Association of Aflatoxin B1 Levels with Mean CD4 Cell Count and Uptake of ART among HIV Infected Patients: A Prospective Study

PLOS ONE

Dear Dr. Chemin,

Thank you for sending the comments from the reviewers of our paper submitted to PLOS ONE. We have made the corrections requested and have attached the revised paper with highlights for further consideration. We thank the reviewers for their careful review and believe that the changes have significantly improved the quality of the manuscript. This is a point-by-point response detailing the revisions that have been highlighted in the manuscript.

COMMENTS FOR THE AUTHOR:

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

Response: The laboratory protocols used in this paper have been published previously; the references for the publications are given in the paper.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response: We read the PLOS ONE style requirements at the websites above and ensure that our manuscript meets the style requirements.

2. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

Response: We have specified that written informed consent was obtained from each participant in the ethics statement in the Methods and in the online submission information. This study did not include minors.

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

Response: We have explained that we developed a questionnaire as part of this study. A copy has been included as Supporting Information file S1_File.pdf.

4. Thank you for submitting the above manuscript to PLOS ONE. During our internal evaluation of the manuscript, we found significant text overlap between your submission and the following previously published works.

- https://moam.info/mycotoxins_5a3c12051723dd42662aa122.html

- https://doi.org/10.3390/toxins7124868

- https://doi.org/10.2217/fmb.13.166

We would like to make you aware that copying extracts from previous publications, especially outside the methods section, word-for-word is unacceptable, even for works which you authored. In addition, the reproduction of text from published reports has implications for the copyright that may apply to the publications.

Please revise the manuscript to rephrase the duplicated text, cite your sources, and provide details as to how the current manuscript advances on previous work. Please note that further consideration is dependent on the submission of a manuscript that addresses these concerns about the overlap in text with published work.

Response: We have revised the manuscript to completely rewrite the duplicated text, cite the sources, and provide details as to how the current manuscript advances previous work (pages 18-19).

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Response: We have revised the data analysis and the manuscript in response to the reviewer’s critiques and suggestions below to ensure that it is technically sound and that the data support the conclusions.

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Response: We have re-run and conducted additional statistical analyses in response to the reviewer’s comments below. Tables 1 and 3 have been re-done completely using the median aflatoxin value to divide the study participants into high and low aflatoxin groups. Based on this change, the first two rows on Table 2 have been revised and highlighted. Figure 3 has been changed to into Figures 3a and 3b to show CD4 counts and ART initiation for the low and high AF-ALB groups separately over time. A Supplemental Table has been added for the multivariable logistic models with aflatoxin as a continuous variable. ________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Response: The data for this study involves human research participant information (including private medical record information of participants) and therefore will be made available upon reasonable request.

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Response: We thank the reviewer.

________________________________________

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Reviewer’s description of the study and findings: This clinical prospective study was among ART naïve HIV+ patients in Ghana over 5 years between 2009-2013 with relatively high baseline CD4 count with mean in the 600s (wide range of 300-1616). Participants were grouped into two groups, high AF-ALB vs. low AF-ALB levels. The authors observed that higher AF-ALB group at each time point was associated with lower CD4 count and even with multivariable analysis, this association was significant. Data on diet and sociodemographics were based on questionnaires and HIV disease and diagnosis (including CD4 and ART initiation) were based on medical record. Blood was collected for measurement of AF-ALB level (quantifies level in the past 2-3 months) through HPLC, as well as HIV-1 RNA viral load.

The study placed participants into high (63%) and low (37%) AF-ALB groups. The results showed that higher level AF-ALB group, there was a greater proportion of males compared to females; higher mean VL and lower CD4 count observed compared to lower group. Higher level also was correlated with storing maize for longer period of time, buying food.

Overall, the hypothesis and data supporting it is intriguing; reporting a significant observation that higher AF-ALB group associated with lower CD4 count at baseline and continued through year 1-4. If validated, this could be clinically important in guiding food consumption/storage guidelines to minimize exposure to high levels of AF-ALB exposure among PWH.

The study had several important observations:

• The cohort was unique, in that all were treatment naïve and across socioeconomic status with good spread in age groups. There was some uneven distribution in low/middle SES group between the two groups, with more in the high AF-ALB group (67% vs. 60%) which may effect the endpoint measured, like CD4, access to ART and baseline VL.

Response: There was some uneven distribution in AF-ALB in the low/middle vs high SES group (70% vs 63% using the median), therefore our analytical strategy included covariate adjustment for SES to account for potential confounding. We created an SES score using Principal Components Analysis based on household indicators such as housing type, plumbing and electricity, similar to approaches used in studies of LMIC populations. We observed that low/middle SES status was associated with higher CD4 counts. A possible explanation for this is that higher SES participants were diagnosed with HIV after a longer time of HIV infection and/or took charge of their own health care for a longer period before attending the public clinic for HIV care. We observed that in the early to mid-2000s, HIV infected people with better economic means would attend private clinics/hospitals to avoid disclosure of their HIV-positive status. However, in accordance to scale up HIV prevention, treatment, and support services (United Nations General Assembly Meeting on AIDS in 2006), the Ghana AIDS Commission established programs to provide ART in public hospitals and health centers in districts in all ten regions of Ghana and removed of the availability of ART from private clinics/hospitals [34]. HIV infected people of higher SES may have taken a longer time to attend the public facilities for HIV care. Access to ART was not a factor in this study as ART was available free of cost to ALL participants during the time of the study regardless of SES. Baseline viral load was significantly higher in the high AF-ALB group. It is established that high viral load is associated with low CD4. This explanation is included in the discussion on pages 15-16.

• Lower CD4 count significantly observed in those with baseline high AF-ALB levels which persisted at each collection period gives a consistent correlation between the two factors.

• Interesting finding of lower CD4 count observed at baseline and throughout years during dry season, which previous study had showed higher AF-ALB levels in dry season. This is intriguing, but could be confounded by many factors not accounted for in this study.

Response: We have removed the discussion on season from the paper since season was not significant in any of the multivariable models.

Minor criticisms:

• Survey data about food grown, maize storage and food consumption likely has significant recall bias and unclear if these questionnaire had not been previously validated to be accurate measures.

Response: We agree with the reviewer that there could be recall bias in the survey data on food grown, maize storage and food consumption, and have included this as a limitation in the paper (page 18). The questionnaire was developed for the study and has been included as supporting information. Eight staff members (doctors, nurses, administrative personnel) from the hospitals reviewed it for understanding, clarity and cultural appropriateness after which it was revised. It was then pilot tested among six clinic patients similar to the ones recruited for the study and revised before use in the study.

• More lower + middle SES in the high AF-ALB group may bias results although comparison of SES between groups was not statistically different based on authors analysis. However, if the group was divided into low+middle vs. high there is likely difference between the two groups based on AF-ALB levels. For example, baseline VL was 200K in high vs 83K in low AF-ALB group. This could be biased by SES and related assess to care and ART which is suggestive of greater proportion of high socioeconomic status in the low AF-ALB group (28 vs. 34%). Similarly, lower SES is related to high AF-ALB exposure, but low SES likely means less access to ART or care and so allows, therefore, augmenting the CD4 difference.

Response: In re-running multivariable model 1 using the median AF-ALB to separate participants into high and low groups, SES is significant. However, contrary to the reviewer’s comments that the findings “could be biased by SES and related access to care and ART which is suggestive of greater proportion of high SES status in the low AF-ALB group (now 30 vs. 37%). Similarly, lower SES is related to high AF-ALB exposure, but low SES likely means less access to ART or care and so allows, therefore, augmenting the CD4 difference”, we found that low/middle SES was associated with higher CD4 counts. We explained on page 4 above in our response to the reviewer that the likely reason for this finding is that HIV infected people of higher SES may have taken a longer time to attend the public hospitals and clinics for HIV care. Access to ART was not a factor in this study as ART was available free of cost to ALL participants during the time of the study regardless of SES. Ghana Government hospitals and clinics provided free HIV care and treatment to all HIV infected patients.

• Data needs to be presented on proportion in each group on ART vs. not on ART at each year of follow up as this ART frequency in each group is unclear and could certainly be biasing the CD4 observations. Figure 3 shows ART frequency but not broken down by frequency in each AF-ALB group, which is the basis of all the comparisons.

Response: We have added Figures 3a and 3b that present the proportions of participants in the low and high AF-ALB groups on ART vs. not on ART at each year of follow-up. A higher number of participants in the low AF-ALB group went on ART in the first two years.

• Figure 3: Unclear why CD4 dropping in this cohort at year 5 when each year more people were started on ART, this should continue to go up, thus suggesting groups of patients are not being started on ART and therefore will dropping CD4 counts.

Response: The mean drop in CD4 for the total study group (original Figure 3 submitted with the paper) was 40 cells (from 680 at year 3 to 640 at year 4). The new Figures 3a and 3b show that the drop in CD4 at year 5 occurred among the high AF-ALB group. There could be a number of reasons for this besides access to ART. As a result of their high AF-ALB status, the participants could have developed poorer health over time, developed less ability to tolerate ART regimen, or decreased adherence to ART. ART was available to ALL participants.

• Discussion mentioned many participants were females in “antenatal clinics”; female hormone and pregnancy is known to affected CD4 and VL count and therefore may again bias data. It would be important to know how many of participants were pregnant and recent post-natal because of effects of pregnancy hormones on CD4 count.

Response: It has been shown that “CD4 counts were an average of 56 cells/mm3 lower during pregnant compared to non-pregnant periods and 70 cells/mm3 lower during pregnant compared to postpartum periods” (Heffron et al. JAIDS 2014; 65(2): 231-236). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3898601/. This “drop” is only temporary and is not believed to be a real reduction in CD4 cells but the result of the same amount of cells in a larger amount of blood (the amount of blood increases in pregnancy). In our study, we did not collect pregnancy or recent postnatal data for the women. However, our results show that women had higher CD4 counts than men at ALL time points.

Major criticism:

Overall, the study reports a significant association between high AF-ALB level and lower CD4 baseline and recovery. It is purely a correlative study and does not establish causality even though it has the samples and cohort to do so.

Response: Typical epidemiologic studies outside of clinical trials are unable to establish causality in any context. However, our prospective study design, evaluation of baseline and follow-up measures, extensive control for potential confounders, and utilization of objective measures of AF-ALB and CD4 counts provide compelling evidence for a strong epidemiologic association that deserves careful attention.

A major issue with the study is the heterogeneity of the two groups in terms of HIV status (nadir CD4 count/ years of infection, baseline CD4 count, ART status and years on ART while in study) and the unclear rationale for dividing into two groups based on cutoff of 15pg/mg of AF-ALB. I am concern of the strengths of the associations found in this study as these are all important confounders which could have biased results to show an effect when in fact, biologically they may not be.

Response: In the original paper submitted the mean AF-ALB level of 15pg/mg was used to divide the participants into high and low AF-ALB groups. Since AF-ALB is skewed, we went back and ran all of the analyses using the median AF-ALB level of 10.4pg/mg and present entirely new Tables and Figures.

For example, conclusion that males had lower CD4 count at baseline and then at each follow up timepoint. If they started at lower T cell, they will continue to be lower regardless of AF-ALB level. Although this difference can be due to AF-ALB effect, this observation could simply be due to the fact that males in this cohort were infected for longer periods.

Response: We removed the discussion on gender from the paper since gender is not significant in the adjusted models using the median. However, when the models were run with AF-ALB as a continuous variable, female gender was significantly associated with higher CD4 count in all of the adjusted models.

Again, showing that duration of HIV infection of each participant (or marker such as nadir CD4 count) would need to be considered in the analysis.

Response: We included knowledge of HIV positive status in the new analysis.

Similarly, the observation that age was associated with lower mean CD4 count in the study could again be simply due to difference in duration of HIV infection (which the author acknowledges but if this data was collected then could be adjusted for in analysis). The two groups would have to matched or controlled for differences in baseline CD4 count.

Response: We agree with the reviewer, we have removed the discussion on age from the paper. The former discussion was based on unadjusted data. Age is not significant in the adjusted models.

The second major criticism is that the study treated AF-ALB levels as a categorical or dichotomous variable (high vs. low) instead of continuous variable which would have been more informative as I would argue that seems more clinically relevant as we expect in real-life that AF-ALB levels would run a wide range and there will be a concentration effect if there is indeed an immunosuppressive effect on CD4 cells. It is unclear why the 15pg/mg was used as the cut-off to determine the high vs. low categories as the author never described the significance of this chosen level. It would be interesting to run the analysis with AF-ALB levels as a continuous variable and assess for correlation with CD4 count, but need to control duration of HIV infection or nadir CD4 count.

Response: We thank the reviewer for this critique and based on the skewedness in AF-ALB levels, we used the median level of 10.4 pg/mg (instead of the mean level of 15pg/mg) to separate participants into low and high groups and re-analyzed the data. We also ran the adjusted multivariable models with AF-ALB as a continuous variable and obtained results similar to that obtained using the median AF-ALB. Additionally, we obtained significant results for gender with females having higher CD4 levels than males in all models (Supplemental Table). We controlled for the time when participants knew of their HIV positive status in the new analysis.

The third major criticism is that the study does not control or separate out analysis of those on ART, many started during study period. They observed that HIV VL was higher and CD4 cell count lower in the high AF-ALB group at baseline but this could be accounted for by the greater proportion of those NOT on ART (69% vs. 66.3%). Looking at CD4 level change over time is important, but not as a group (AF-ALB high vs. low). The difference proportion in each on ART and/or ART initiation in the middle of study is not accounted for and still treated as two groups. It would be more accurate to look at trajectory of CD4 count each group (ie, see if there is difference in rise of CD4 count after 1 years, 2 year, 3 years, etc after ART initiation between AF-ALB high vs. low groups). Making this comparison with a heterogenous and unevenly distributed groups (ART frequency in Figure 3 should be divided into the AF-ALB groups to compare proportion started in each group) is not accurate.

Response: We have separated CD4 counts and ART initiation for the high and low AF-ALB groups over time and presented the results in Figures 3a and 3b as the reviewer suggested.

Fourth major criticism of the study is that this is a correlative study and does not establish or even suggest causality. To make a more casual association, authors could assess change in level of AF-ALB in each patient to change in T cell frequency or even test for function (activation/exhaustion markers, etc) over time. This could have been done with this prospective longitudinally followed cohort of 5 years.

Response: Please see response on causality above. Observational studies are not methodologically able to establish causality, however our study design approach enabled rigorous control for potential confounders that taken together indicate strong epidemiologic associations. We thank the reviewer for the suggestion about evaluating AF-ALB in relation to T-cell frequency over time. Our goal with the current analysis and manuscript is to present initial findings that serves as the basis for future studies examining T cell function over time. This area of research is markedly understudied, and our findings will shed more light on the biology of AF-ALB exposure in this population. We will explore funding opportunities to conduct additional analysis of immune cell function as a next step.

Thank you for your kind consideration.

Respectfully,

Pauline Jolly, PhD, MPH

Professor Emerita,

Director, UAB Minority Health International Research Training Program

Recipient, 2014 Ellen Gregg Ingalls/UAB National Alumni Society Award for Lifetime Achievement in Teaching

2018 Fulbright Specialist Scholar, Institute of Public Health, Ho Chi Minh City, Vietnam

Submitted filename: Response to Reviewers.pdf

Click here for additional data file.

19 Nov 2021

Association of Aflatoxin B1 Levels with Mean CD4 Cell Count and Uptake of ART among HIV Infected Patients: A Prospective Study

PONE-D-21-00637R1

Dear Dr. Jolly,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Isabelle Chemin, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The answers provided and changes in the manuscript are sound and improved the quality of the manuscript.

Reviewers' comments:

14 Jan 2022

PONE-D-21-00637R1

Association of Aflatoxin B1 Levels with Mean CD4 Cell Count and Uptake of ART among HIV Infected Patients: A Prospective Study

Dear Dr. Jolly:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Mrs Isabelle Chemin

Academic Editor

PLOS ONE