Erythrocyte indices, anaemia levels and types in Kenyan injection and non-injection substance users

The impact of injection and non-injection substance use in HIV infections is an area of great public importance especially with respect to hematologic and immune profiles. Evaluations of hematologic and immune status are critical for better disease classification and clinical management especially of HIV positive substance users. However, not much information is known about the hematologic and immune derangements in HIV infected injection and non-injection substance users. This study, therefore, aimed at determining the laboratory markers of hematologic and immune derangements in HIV infected substance users. Hematologic and immune profiles were evaluated on venous blood specimens obtained from injection substance users, ISU (HIV-infected, n=62 and -uninfected, n=213) and non-injection substance users (HIV-infected, n=33 and -uninfected, n=186); and non-substance using controls (n=56) from Mombasa, coastal town of Kenya. The prevalence of anemia was higher in HIV infected ISU (48.4%) and non-ISU (63.6%) (p<0.0001); and HIV uninfected ISUs (56.3%) compared to HIV-uninfected non-ISUs (39.2%) and non-substance using controls (28.6%; p=0.0028). Hypochromic anaemia was more prevalent in the HIV-infected ISU (50.0%) and non-ISU (61.9%), and HIV-negative ISU (63.3%) relative to the HIV-negative non-ISU (39.7%) and non-substance using controls (56.3%; p=0.0007). Mild immunodeficiency dominated in the HIV infected individuals (HIV-infected ISU, 32.3% and non-ISU, 21.2%) versus HIV-uninfected ISU (16.9%); non-ISU (12.9%); and non-substance users (14.3%) while severe immunosuppression prevailed in HIV infected substance users (ISU, 14.5% and non-ISU, 15.2%) against HIV uninfected substance users (ISU, 5.2% and non-ISU, 3.8%); thus immunosuppression in substance users is aggravated with HIV infection. Moreover, drug-induced immunosuppression is associated with a higher likelihood of anaemia in HIV-uninfected substance users; ISU (OR=3.95, CI=1.934-8.077, p<0.0001) and non-ISU (OR=3.63, CI=1.571-8.39, p=0.003). Altogether, hypochromic anaemia, normochromic anaemia and CD4+ T-helper cytopenia are the most prevalent hemocytopenias in HIV infected and uninfected injection and non-injection substance users.

141 between 8.00am and 10.00am prior to the participants having breakfast to control for the 142 haematological changes due to the circadian rhythm and nutritional status hence obtaining 143 strictly comparable values. All laboratory tests were performed within two hours of sample 144 collection to maintain sample integrity. EDTA blood was used for haematological analysis 145 while SST was used for serum extraction in HIV-1 viral load quantification.
146 Hematologic measurements 147 Complete Blood Counts were done within the first hour of blood collection using the 148 quantitative BC-3200 Mindray auto-haematology analyser (Mindray TM Inc., Mahwah, USA).
149 Anaemia levels and types were classified based on haemoglobin concentration prescribed by 150 the World Health Organization (41) while anaemia aetiology was classified based on blood-151 markers, cellular morphology and staining characteristics (54-56).

Preparation of blood slides
153 Thin blood films were made on new microscope slides (labelled with participant ID) to 154 prevent cell aggregation and stain precipitation. Back up smears were also made. The thin 155 smears were thoroughly air-dried followed by methanol fixation for 10 minutes. The blood 156 smear was then completely covered with undiluted Leishman Stain which was added 157 dropwise using a bulb-pipette. Twice the volume of buffered water (pH. 6.8) was gently 158 added and thoroughly mixed. Staining was done for 10 minutes after which the slide was 159 washed off under running tap water. The back of the slide was wiped and the slide placed 160 standing on a draining rack for the smear to dry. 161 Microscopic analysis 162 Examination of the stained blood films was done by two independent and blinded hemato-163 technologists who assessed erythrocyte morphology. Slides with differences of more than 164 5% in the results of the two hemato-technologists were re-read by a third independent 165 hemato-technologist. Ten per cent (n=55) of the read slides were randomly selected and the 166 results confirmed by a haemato-pathologist.
167 CD4+ T-cell enumeration 168 Fifty microlitres (50μl) of EDTA anticoagulated blood was stained with anti-CD3  354 Chronic inflammation was the second most common mechanism associated with anaemia 355 prevailing in injection and non-injection substance users. Therefore, substance use is likely to 356 be associated with inflammation. Khat and alcohol use has been shown to cause intestinal 357 lesions leading to gastritis (85-90). This intestinal inflammation is likely to cause the liver to 358 secrete more of the hormone hepcidin which acts by preventing the body from utilizing stored 359 iron (ferritin) and subduing iron absorption in the duodenum. As a matter of fact, anaemia 360 due to nutritional deficiency was the third most common cause across all the study 361 participants. Nutritional deficiency anaemia is probably due to the low dietary intake of iron, 362 folate and vitamin B12 in the general population and substance-induced damage of the 363 gastrointestinal mucosa within the substance using groups (91). Mal-absorption states in these 364 groups need to be investigated including the production and inhibition of the intrinsic factor, 365 which is important in differentiating the types of nutritional anaemias.
366 Anaemia due to mixed aetiology was the most frequent mechanism across our study 367 participants. However, due to the limited resources and time constraints, we could not 368 perform further investigations to specifically determine the kinetics underlying the mixed 369 aetiology of anaemia. Despite this challenge, reports from our analysis indicated a 370 coexistence of the above mechanisms with other aetiologies whose haematological 371 "blueprints" were suggestive of underlying hemoglobinopathies and thalassemias. However, 372 this claim needs to be substantiated by further investigations. In addition, there were wispy 373 signs indicative of intravascular haemolysis and suppression of erythropoiesis. We speculate 374 that intravascular haemolysis could be attributable to the damping effect where the drug 375 metabolites are adsorbed onto the RBCs which become antigenic resulting in their untimely 376 destruction by the immune and the reticuloendothelial system. 377 Anaemia observed was also classified based on the RBC chromasia as hyperchromic, 378 hypochromic and normochromic. Hypochromic anaemia was the most prevalent type of 379 anaemia accounting for more than 50% of the anaemia. Hypochromic anaemia was common 380 across all the study groups. Some of the mechanisms driving the existence of hypochromic