Hematological and Biochemical Data Obtained in Rural Northern Uganda

Reference intervals for common hematological and clinical chemistry parameters constitute an important basis for health care. Moreover, with increasing priority in drug and vaccine development for infectious diseases in Africa, the first priority is the safety evaluation and tolerability of the candidate interventions in healthy populations. To accurately assess health status and address adverse events, clinical reference intervals in the target population are necessary. We report on hematological and biochemical indices from healthy volunteers who participated in a clinical trial in Lira, northern Uganda. Median and nonparametric 95% percentiles on five hematology and 15 biochemistry analytes are shown. Although most hematological analytes conformed to reported reference intervals and trends in Africa, literature review from different African countries highlight the need for a region-specific children reference interval that can be appropriate for the population.

Uganda population that can typically participate in intervention trials and vaccine studies. The results also underscore the need for region-specific children reference intervals that can be more appropriate for the population.

Study Population
In this study, only the screening visit data of volunteers who participated in the BK-SE36 Phase Ib clinical trial [2] was included in the analysis. The study took place at Lira Medical Center in Lira district, Northern Uganda. Lira district is approximately 1063 m above sea level and lies 347 km north of Kampala, the capital of Uganda. The district is predominantly agricultural, with the population comprising largely small landholder subsistence farmers: growing beans and cassava; owning chicken and goats [21]. The staple foods are mainly maize, cassava, sorghum and sweet potatoes. The area has a tropical climate with a temperature range of 17.4-30.3 °C and an average rainfall of 1478 mm/year [21].
Briefly, there were two stages of screening. Stage 1 was in [21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] year-olds (n = 56) and Stage 2 was in 3 age cohorts (6-10, 11-15 and 16-20 years old). Screening for Stage 1 was conducted in April-May 2010 and for Stage 2 in September-October 2010. The trial was conducted in compliance with the study protocol, the International Conference on Harmonization's Good Clinical Practice standards, the Declaration of Helsinki and Uganda regulatory requirements. Approvals for the protocol, subject information and informed consent forms were obtained from the ethical institutional review committees (IRC) of Osaka University (RIMD-IRC) (Japan), Research Foundation for Microbial Diseases of Osaka University (BIKEN-IRC) (Japan), Med Biotech Laboratories (MBL-IRC) (Uganda), and regulatory bodies Uganda National Council for Science and Technology (UNCST) and the Uganda National Drug Authority (NDA). Informed consent was obtained in English, Swahili or Luo (the language commonly spoken locally). Written consent was obtained from all enrolled subjects, in addition to individual assent from children aged eight and above. All participants were screened through a medical history questionnaire and complete physical examination. Females of child bearing age were tested for pregnancy. Selected participants were healthy, with no obvious symptoms/signs of either acute or chronic respiratory, cardiovascular, gastrointestinal, hepatic or renal disease; no history of blood donation/transfusion within one month and, for females, non-lactating. A detailed report of the study population recruitment, enrollment, inclusion/exclusion criteria, and blood sampling has been published [2].
Whole blood samples were collected from fasting subjects via venipuncture with a Vacutainer system (Becton Dickinson Biosciences, Franklin Lakes, NJ, USA), with all blood drawn in the morning (7:00-12:00 am) and hematological and biochemistry tests performed within the day. The samples were also checked for malaria parasites. Blood sampling and processing were done according to standard operating procedures by trained laboratory staff (EDTA treated tubes for hematology and untreated tubes for biochemistry tests). Hematological indices were measured by Sysmex KX-21N (Sysmex Corporation, Kobe, Japan); and biochemical indices by Cobas C111 (Roche Diagnostics, Mannheim, Germany) per manufacturer's instructions. Daily quality control was performed on the equipment and all laboratory procedures adhered to Good Clinical Laboratory Practices. Calibrators and controls were obtained from the instrument manufacturer. For hematology tests, daily running of the 3-part Differential Whole Blood Para ® Check (Para ® Check Low, Para ® Check Normal and Para ® Check High) controls ensured calibration and quality control limits for each analyte. For Cobas C111, we used normal (PreciNorm U) and abnormal (PreciPath U) control sera from Roche Diagnostics. In addition, the laboratory was under an external quality assurance program with the South African National Accreditation System. Assessments included white blood cell (WBC) count, red blood cell (RBC) count, hemoglobin (Hb), hematocrit (Hct), platelet count, alkaline phosphatase (AL-P), alanine aminotransferase (ALT), aspartate aminotransferase (AST), γ-glutamyl-transferase (-GTP), amylase, albumin, total protein, total bilirubin, cholesterol, glucose, creatinine, urea nitrogen (BUN), uric acid, potassium and sodium.

Statistical Methods
Data were pooled, analyzed untransformed and are shown as mean, median, maximum-minimum values, and 95% reference intervals according to CLSI C28-A3 guideline [12]. Of the measured analyte, 70% did not conform to a Gaussian distribution according to the D'Agostino and Pearson and Shapiro-Wilk tests for normality. Outlier exclusion was the recommended Dixon-Reed's outlier method. Non-parametric percentiles and 90% CIs were determined. Possible partitioning to two broad age groups: 6-15 years-old and 16-32 years-old were done using the robust method with possible outliers detected by visual inspection of distributions and Tukey's criterion. Analytes that have significant age differences based on the non-parametric Mann-Whitney U test are shown in Table 3. Statistical analyses were performed using MedCalc for Windows, version 13.1.1.0 (MedCalc Software, Ostend, Belgium) and GraphPad Prism 6 (GraphPad Software; La Jolla, CA, USA).

Baseline Characteristics of the Study Population
From April-May and September-October 2010 (two sampling periods), we enrolled a total of 140 healthy individuals [2]. We report on the health status of these individuals based on hematological and biochemical data obtained. The final analysis cohort pooled participants from age 6-32 years-old. Table 1 shows the general characteristics of the study population. In adults (21-32 years), mean age and body mass index (BMI) were similar across gender. Lower mean diastolic and systolic blood pressure (BP) but higher mean pulse rate was obtained from female than male volunteers. In younger cohorts, as expected weight, height and BMI increased with age. Diastolic and systolic BP were similar across age groups, although pulse rate of 6-10 years and 11-15 year-old cohorts were significantly higher than 16-20 years cohort. In the 16-20 year-old cohort, male-to-female ratio was 1:1. However, there was a preponderance of females in the 6-10 year-old cohort (representing 54% against 46%) and in [11][12][13][14][15] year-old cohort (82% vs.18% males).

Hematological Analysis
The mean, standard deviation, median, minimum/maximum values and 95% reference intervals for hematology analytes are given in Table 2. Volunteers had low WBC and platelet counts. Ranges for RBC count, hematocrit and hemoglobin values were generally broad when compared to Western reference intervals obtained from Pathology Associates Medical Laboratories [22]. The low hemoglobin; hematocrit; white blood cell, erythrocyte and platelet counts in Africans compared to northern European population or whites have been shown to be related to ethnicity and/or genetic factors [23][24][25]. Significant age differences for red cell indices were obtained when reference intervals were partitioned to two broad age groups (6-15 and 16-32 years-old). When computed by the robust method as recommended by CLSI C28-A3 guidelines, age differences were evident in the upper values for RBC count (p = 0.009), hemoglobin (p < 0.0001) and hematocrit (p < 0.0001) in 16-32 year-old compared to 6-15 year-old ( Table 3). The increase in hemoglobin, hematocrit and erythrocyte counts; as well as the decrease in platelet counts according to age has been reported [4,19,20]. Values were within the 2.5 and 97.5 percentiles obtained from Uganda and other African countries with an altitude below 2000 m ( Table 4). Higher hemoglobin and hematocrit values were reported for Ethiopia (at >2300 m above sea level) and lower hemoglobin and hematocrit were noted for Ghana (at 60-150 m above sea level).

Conclusions
The opportunity to re-assess currently used reference intervals in northern Uganda presented itself when we conducted a phase 1b trial in Lira, Uganda. Reference intervals serve as a valuable guide for interpretation of laboratory results. Hematology and biochemistry analytes are influenced by a number of factors such as sex, age, ethnic origin, environment (including geographical location, access to health care), constitutional abnormalities (thalassemia, sickle cell disease), pathologic conditions (malaria, HIV and HBV viral infections); and to some extent demographic changes and technological advancement in analyte measurements [4][5][6][7][13][14][15][16][17][18][19][20]23,27,28].
The revised international guidelines from CLSI recommend that if it is not possible to establish detailed reference studies from at least 120 individuals, validation using as few as 20 samples could be helpful together with comparison using published reference intervals and performing posteriori methods [29,30]. We, thus, examined and compared the variability of hematological and biochemical analytes obtained during participant screening for trial enrollment. The analytes were commonly used for clinical assessments to guide inclusion/exclusion into the trial and for adverse event evaluation.
The studied population is heterogeneous with regards to age and gender. However, overall, the upper and lower limits of the reference intervals obtained in this study were almost the same as the limits reported in previous adult hematologic reference intervals ( Table 4). Trends of age-specific differences in some parameters were also evident although this needs further verification in larger cohorts (Table 6). At present there are few reports available for adolescents and younger age groups, especially on biochemistry analytes in Africa [11,19,31,32].
Whether gender or age would alter the reference intervals obtained in the present study is for further studies. The screening activity did not screen out all medical conditions such as hepatitis B, HIV or other parasitic infections except for malaria. We have no information on social habits (tobacco use, diet, alcohol consumption, exercise, etc.) as these were not obtained during enrollment of the volunteers (e.g., factors such as smoking has been reported to affect hemoglobin values [14]). We did, however, have a well-defined trial inclusion and exclusion criteria, and all subject were evaluated for both hematological and biochemical indices. Volunteers were encouraged to come for screening to the study clinic after fasting for at least 8 h. Breakfast was given to all participants afterwards and this assisted in a high level of compliance with this request, although in some cases fasting was interpreted as an instruction not to eat or drink and, thus, some volunteers were mildly dehydrated during blood sample collection. Nevertheless, the study provides preliminary data needed to facilitate health care and research trials in a population in which reference interval generation has not yet been done. Participants in the study also represent volunteers that would participate in clinical trials. Importantly, we also highlight the need for local data especially in younger age groups as their participation in clinical trial is solicited.
necessarily represent the views of BIKEN. Authors also acknowledge the support of 'Funds for integrated promotion of social system reform and research and development', MEXT under the project: Development and Sustainability of Malaria Vaccine Clinical Research Center (38201103-01).

Author Contributions
Nirianne M. Q. Palacpac conceived and conducted the study, analyzed the data and prepared the draft of the manuscript; Edward Ntege and Betty Balikagala conducted the study, analyzed the data and prepared the draft of the manuscript; Adoke Yeka supervised and conducted the study; Hiroki Shirai conceived and supervised the study; Nahoko Suzuki conceived and conducted the study, and analyzed the data; Christopher Nsereko and Bernard N. Kanoi conducted the study; Takuya Okada conducted the study and analyzed the data; Thomas G. Egwang conceived and supervised the study; Toshihiro Horii conceived and supervised the study, and gave final approval of the manuscript to be submitted for publication.