Clinical laboratory reference values from North American and European populations are currently used in most Africans countries due to the absence of locally derived reference ranges, despite previous studies reporting significant differences between populations. Our aim was to define reference ranges for both genders in 18 to 24 year-old Mozambicans in preparation for clinical vaccine trials.
A cross-sectional study including 257 volunteers (102 males and 155 females) between 18 and 24 years was performedat a youth clinic in Maputo, Mozambique. All volunteers were clinically healthy and human immunodeficiency virus, Hepatitis B virus and syphilis negative.Median and 95% reference ranges were calculated for immunological, hematological and chemistry parameters. Ranges were compared with those reported based on populations in other African countries and the US. The impact of applying US NIH Division of AIDS (DAIDS) toxicity tables was assessed.
The immunology ranges were comparable to those reported for the US and western Kenya.There were significant gender differences in CD4+ T cell values 713 cells/µL in males versus 824 cells/µL in females (p<0.0001). Hematologic values differed from the US values but were similar to reports of populations in western Kenya and Uganda. The lower and upper limits of the ranges for hemoglobin, hematocrit, red blood cells, white blood cells and lymphocytes were somewhat lower than those from these African countries. The chemistry values were comparable to US values, with few exceptions. The upper limits for ALT, AST, bilirubin, cholesterol and triglycerides were higher than those from the US. DAIDStables for adverse events predicted 297 adverse events and 159 (62%) of the volunteers would have been excluded.
Citation: Tembe N, Joaquim O, Alfai E, Sitoe N, Viegas E, Macovela E, et al. (2014) Reference Values for Clinical Laboratory Parameters in Young Adults in Maputo, Mozambique. PLoS ONE 9(5): e97391. https://doi.org/10.1371/journal.pone.0097391
Editor: Robert Lane Schmidt, University of Utah School of Medicine, United States of America
Received: February 1, 2014; Accepted: April 18, 2014; Published: May 14, 2014
Copyright: © 2014 Tembe et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by funding from The Regional HIV/AIDS Team for Africa, Embassy of Sweden, Lusaka jointly funded by Sweden and Norway (Sida contribution number 2150012801). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The number of clinical trials related to HIV/AIDS, tuberculosis and malaria conducted in Africa is increasing sharply and is expected to increase further in the coming years , . Routine health assessment and management of clinical trials relies on accurate laboratory references. Of primary importance in vaccine clinical trials is the evaluation of safety and tolerability in a clinically “normal” population. Furthermore, volunteers may be assessed for disease progression and evaluation of possible clinical trial-associated toxicity and adverse events.
Laboratory reference intervals for healthy populations have not been established in most African countries. Common practice in these countries, including Mozambique, is to use reference ranges derived from populations living in Europe or the United States (US). Studies have shown differences between clinical reference ranges in African populations compared to those established in industrialized countries –. Several studies have also reported that laboratory parameters vary geographically by ethnic origin, genetics, gender, altitude and environmental factors , , –.
The use of improper clinical reference ranges to assess participant eligibility and safety for clinical trials may lead to unnecessary exclusion of eligible participants, contribute to over-reporting of adverse events (AEs)  and increase the number of referrals for clinical investigations. Laboratory abnormalities based on non-indigenous laboratory parameters and medical abnormalities were reported to be the main reasons volunteers were excluded from two Kenyan HIV vaccine clinical trials . Moreover, studies have suggested that use of the US NIH Division of AIDS (DAIDS) toxicity tables may not be appropriate for African populations , .
Prior to execution of a phase I/II HIV vaccine trial (TaMoVac 01) in Mozambique, we performed a study to define the prevalence and incidence of HIV and other sexually transmitted viruses in Maputo in a population of young adults. This was also an opportunity to establish clinical reference values in 18 to 24 year-old Mozambicans.
This study establishes reference ranges for immunological, hematological and chemistry parameters in healthy young adults in Mozambique. We determined gender differences and compared values established for Mozambican young adults with those previously reported for the same age group in other African countries and with established intervals from the US (Massachusetts General Hospital, MGH-USA) . Additionally, we applied the division of AIDS (DAIDS) toxicity tables for grading of AEs  to evaluate their potential implications for vaccine trials.
Materials and Methods
This study was approved by the National Bioethics Committee for Health of Mozambique. Written informed consent was obtained from each participant prior to conducting any study procedures.
Study site and subjects
The study took place at the SAAJ clinic (SAAJ: ‘Serviço Amigo de Adolescentes e Jovens’ or Adolescent and Youth Friendly Service) at the Department of Obstetrics and Gynecology, Maputo Central Hospital. The SAAJ clinic provides services (free of charge) for young people seeking counseling and treatment for any health problem, but with particular attention to reproductive health and control of sexually transmitted infections (STIs). A low prevalence of STIs and HIV and a high level of awareness has previously been reported in a study performed at the SAAJ clinic .
Maputo is the capital of Mozambique and is situated in a coastal area adjacent to the Indian Ocean, between the coordinates 25° 50′ and 26° 10′ S and 32° 30′ and 32° 40′ E . The city is situated at an average altitude of 47 meters, has an area of 346.77 km2 and had a population of 1,094,315 in 2007 .
A total of 257 healthy individuals between 18 and 24 years old were recruited from a cohort of youths participating in a study of the prevalence and incidence of sexually transmitted viruses at the SAAJ clinic, Maputo Central Hospital. Medical staff performed physical examinations and collected clinical histories. Volunteers who were febrile, pregnant, or seropositive for HIV, syphilis or hepatitis B surface antigens were excluded from the study.
HIV, syphilis and hepatitis B screening
The national algorithm for HIV testing was used to diagnose HIV. HIV testing was performed using two immunochromatographic assays, the Determine HIV-1/2 (Inverness Medical, Bedford, United Kingdom) followed by the UniGold HIV-1/2 (Trinity Biotech, Bray, Ireland). Syphilis testing was performed using SD Bioline Syphilis 3.0(Standard Diagnostics, Suwon City, South Korea). Serum samples were tested for Hepatitis B virus (HBV) using the Hepatitis B Surface Antigen (HBsAg) ELISA Kit (Human, Wiesbaden, Germany).
A urine pregnancy test was administered to all females prior to collection of blood samples, using the QuickVue One-Step HCG Urine Test (Quidel Corporation, San Diego, USA).
Sample collection took place between August 2009 and September 2012. Most samples (70%) were collected between November 2009 and August 2010, but the inclusion period was extended to recruit additional males to the study. Blood was collected in 4 ml EDTA Vacutainer tubes (Becton-Dickinson, Franklin Lakes, New Jersey, USA)in preparation for lymphocyte and hematological testing. Whole blood was collected in 10 ml serum Vacutainer tubes (Becton-Dickinson, USA)in preparation for testing chemical parameters and HBV status. Samples were collected in the morning between 8.00 AM and 12.00 noon, kept at room temperature and transferred to the laboratory of the National Institute of Heath in Maputo for analysis.
Flow cytometry analysis for Immunophenotyping
Immunophenotyping was performed using a FACS Calibur flow cytometer (Becton-Dickinson, Franklin Lakes, New Jersey, USA). Samples were analyzed within 24 h of specimen collection. In brief, 20 µl of CD3FITC/CD8PE/CD45perCP/CD4APC or CD3 FITC/CD16+CD56 PE/CD45PerCP/CD19 APC MultiTest reagents (Becton-Dickinson, USA) was mixed with 50 µl of whole blood and incubated in the dark at room temperature for 15 min. Red blood cells were then lysed by adding 450 µl of fluorescence-activated cell sorter lysing solution (Becton Dickinson, USA).The tubes were then incubated at room temperature for another 15 min. MultiSET software (Becton-Dickinson, USA) was used to perform the analysis.
A complete blood count and differential was performed using the Sysmex KX-21N Hematology Analyzer (Sysmex Corporation, Kobe, Japan) as recommended by the manufacturer. The samples were analyzed within 6 h of specimen collection. The machine automatically dilutes a whole-blood sample, lyses and counts the cells, and then gives a printout result. Seventeen parameters were analyzed; leukocytes (WBC), erythrocytes (RBC), platelets (PLT), lymphocytes (LYM), neutrophils (NEUT), hemoglobin concentration (Hb), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red blood cell distribution width measured by standard deviation (RDW-SD), red blood cell distribution width measured by coefficient of variation (RDW-CV), platelet distribution width (PDW), mean platelet volume (MPV), platelet larger cell ratio (P-LCR) and the percentages of lymphocytes (LYM), neutrophils (NEUT), and the mixed population of monocytes, basophiles and eosinophils (MXD). The absolute cell counts were expressed as number of cells ×  per liter.
Serum chemistry was performed using a Vitalab Selectra Junior (Vital Scientific, Dieren, Netherlands) per the manufactureŕs instructions. Serum was separated within 4 h of collection and analyzed within 7 h of blood draw. Each sample was analyzed for creatinine, total bilirubin (T-Bil), albumin, aspartate aminotransferase (AST), alanine aminotransferase (ALT), glucose, urea, uric acid, amylase, HDL cholesterol, triglycerides and alkaline phosphatase (ALP).
In addition to commercial controls run daily with each instrument, the laboratory routinely participates (3 times per year) in external quality assurance testing programs distributed by Thistle (Thistle QA, Bryanston, South Africa) for hematology and clinical chemistry, and the International Quality Assurance Program-QASI (Public Health Agency of Canada, Ottawa, Canada) for lymphocyte typing. In the case of internal control failure, testing was suspended.
Clinical and laboratory personnel trained in GCP and GCLP, respectively, conducted the study according to ICH/GCP guidelines, employing standard operating procedures.
Data was entered into Microsoft Office Excel 2007 and analyzed using R version 3.0.0 . Prior to calculation of the reference ranges each parameter was check to be normally distributed using Shapiro-Wilks test; and if a deviation from normality was found the best Box-Cox power transformation was performed. Grubbs tests were performed to check for outliers and box plots and qqnorm plots were used to assess outliers visually. In general outliers were removed but their influence was checked by calculating the reference range with the outlier and comparing the reference range without the outlier.
Median and 95% reference ranges (2.5th- 97.5th percentiles) were established for immunology, hematology and chemistry values. The non-parametric Wilcoxon rank-sum test (Mann-Whitney U test) was used to test for differences by gender. For all analysis the significance was set at 0.05.
For clinical laboratory parameters that did not meet the recommended minimum sample size of 120 individuals, which is recommended by the Clinical and Laboratory Standards Institute (CLSI) , percentiles were obtained from a boostrap procedure where 1000 resamples were performed per parameter and gender. On each resample, a percentile was calculated and the final percentile was the averages of the resampled percentiles , . The p-values were obtained using a 10000 boostrap resamples per parameter to obtain the distribution of the Wilcoxon rank-sum test under the assumption of no difference between genders and then the original was compared to this distribution and area under curve on the right side of the histogram was considered as p-value.
Further outlier presence was checked using the plots of the resample of the mean minus the trimmed mean as explained by Chernick and Labudde  and performing the Silverman for multimodality on that plot. If multimodality was found the extreme value was removed from the analysis.
We then compared values established for Mozambican young adults with those previously reported for the same age group in other African countries ,  and with established intervals from the US (the Massachusetts General Hospital, MGH-USA) . Additionally, we applied the division of AIDS (DAIDS) toxicity tables for grading of adverse events  to evaluate their potential implication in vaccine trials.
A total of 257 healthy volunteers aged 18 to 24 years were enrolled in this study. Of these, 102 (40.3%) were males and 155 (59.7%) were females. The target of >120 males was not reached because of the comparatively low attendance of males at the Youth Clinic from which volunteers were recruited. The full range of lymphocyte subsets, hematology and chemistry analytes was not run in all subjects, primarily due to insufficient access to reagents required to test some parameters.
Table 1 shows the median values and reference ranges for lymphocyte subsets generated from participants in the study. Analyses of lymphocyte subsets revealed statistically significant differences between genders with respect to the absolute number of CD4+ cells, percentage of CD4+ cells, percentage of CD8+ cells, and CD4:CD8 ratios. Reference ranges for the absolute number of CD4+ cells (p<0.0001) and percentage of CD4+ cells (p<0.0001), CD4:CD8 ratios (p = 0.0001) as well asCD3+ cells (p = 0.0420) were significantly higher in females than males; however, the percentage of CD8+ cells was significantly higher in males (p = 0.0038). There were no significant gender differences in CD45+cell counts (total lymphocytes), percentage of CD3+ and absolute CD8+ cell count. Gender comparisons were also performed for CD16+56+ (natural killer cells) and CD19+ cells (B-cells) despite the comparatively low number of male samples analyzed (n = 29). A statistically significant gender difference was seen in percentage of CD16+56+ cells (p = 0.0390). No significant gender difference was found with respect to absolute number of CD16+56+ cells or absolute number and percentage of CD19+ cells.
Our immune phenotype data were generally comparable to those derived from the North American populations provided by Becton-Dickinson or MGH-USA, although the lower and upper limits of the US reference ranges for all T cells were higher than those derived from the present study. The reference ranges for CD4+ and CD8+ T cells derived in the present study were more comparable to values reported from the population in western Kenya than in Uganda (Table 2).
Table 3 shows the medians and reference ranges for hematological parameters, grouped according to gender. We observed statistically significant differences between males and females in most hematological parameters, with the exception of PDW, MPV, absolute LYM, absolute MXD and percentage MXD. The males had higher values of RBC, Hb, HCT, MCV, MCH, MCHC and percentage LYM than females. The females had higher values of WBC, PLT, absolute NEUT and percentage NEUT, RDW-SD and RDW-CV than males.
Our reference ranges showed significant differences from the US population (Table 3). Close to 50% of our study participants had Hb values below the lower limits of the range determined in the US population; this was particularly true for females, 104 (69.3)% of whom had Hb values below the lower limit. We also found a considerable number of participants with platelet and WBC values outside the US reference ranges (47 (18%) and 64 (25%), respectively) (Table 4). Our reference ranges for young adults in Maputo were comparable to those derived from the western Kenyan and Ugandan populations. However, the lower and upper limits for some hematological parameters (Hb, HCT and lymphocytes, RBC and WBC)were lower than those derived from these African countries. The lower and upper limits of neutrophil and platelet reference ranges in our study were higher than those derived from populations in western Kenya and Uganda (Table 4).
Table 5 shows the median values and reference ranges for clinical chemistry parameters. There was a statistically significant difference between genders in all clinical chemistry analytes, with the exception of triglycerides and amylase. Males had significantly higher levels of T-Bil, glucose, uric acid, AST, ALT, albumin, urea, creatinine, and ALP than females. Females had significantly higher levels of cholesterol (p = 0.0196) and HDL cholesterol (p = 0.0066) than males.
The study reference ranges of clinical chemistry parameters were comparable to those derived from the North American population (MGH-USA), with a few exceptions. The upper limits of the reference ranges for ALT, AST, bilirubin and cholesterol derived in the present study were somewhat higher than those from the US. The Maputo values were lower compared to western Kenya in the same age group (Table 6). We also found a high proportion of study participants with glucose and T-Bil values outside of the US reference ranges, in 104 (41.1%) of 199 individuals and in 38 (15.0%) of 253 individuals, respectively (Table 7).
Implications for clinical trials
Table 7 shows the frequency of potential adverse events by applying the DAIDS toxicity grading, which is commonly used in clinical trials, to the values obtained from the cohort of young adults in Maputo. Among the hematologic parameters, Hb and neutrophil counts accounted for the majority of the abnormal classifications. The low Hb levels among our study participants would have resulted in 32 reported AEs; 21 (8.3%) as grade 1, 7 (4.6%) as grade 2 and 4 (2.6%) as grade 3. The low neutrophil counts would have resulted in 16 AEs: 14 (5.5%) as grade 1 and 2 (0.8%) as grade 2.With respect to the clinical chemistry parameters, ALT would have resulted in 27 AEs, 22 (8.7%) grade 1 and 5 (1.9%) grade 2 and AST would have resulted in 197 AEs, 145 (57.3%) grade 1 and 52 (20.6%) grade 2 (Table 7). Additionally, T-Bil values would have resulted in 6 reported AEs, 5 (1.9%) grade 1 and 1 (0.4%) grade 2. In total, based on all hematological and clinical chemistry parameters, 292 AEs are predicted to have been reported. Notably, 159 (62.8%) of the healthy volunteers included into the present study would have been excluded due to abnormal laboratory values should they have been evaluated for a clinical trial using US based criteria.
Clinical laboratory reference ranges have not previously been established in Mozambique. In this study we established reference ranges using samples derived from healthy young adults aged 18–24 years who attended a youth clinic at Maputo Central Hospital.
The number of males was disproportionate to the number of females (40.3% vs 59.7%), mainly because a lower proportion of males attended the SAAJ clinic. This affected the recruitment of males into the study, and the target of 120 subjects tested per analyte as recommended by CLSI  was not reached despite extending the enrollment period.However, a robust bootstrap analysis was used to eliminate bias due to the small sample size, as recommended by the Canadian laboratory initiative on pediatric reference intervals (CALIPER) .
Our immunophenotyping data showed significant gender differences between total lymphocyte counts, absolute T cell counts, absolute count of and percentage of CD4+ T cells and CD8+ T cell percentage, with females having higher values than males in each case. These findings are consistent with those reported for young adults in Kericho, Kenya . Several other studies have also reported that females have higher CD4+ T cell counts than males , –. Overall, the reference ranges reported here for lymphocyte subsets were comparable to ranges reported for young adults in the US (Becton-Dickinson, USA) and western Kenya .
Due to insufficient access to reagents for natural killer cells and B cells determinations, only 34 males and 104 females were tested for these parameters. Although the number of participants was fewer than the 120 recommended by CLSI, a significant gender difference was seen for percentage of CD16+56+ natural killer cells and absolute count of CD19+ B cells. Lower percentages of CD16+56+ natural killer cells were seen in females compared to males, while absolute CD19+ B cells were lower in males than females. These findings were in general comparable to those reported in two studies performed in Tanzanian adults, which reported significant gender differences in percentage of natural killer cells and both the absolute count and percentage of CD19+ B cells , . In both Tanzanian studies, females had a significantly lower percentage of natural killer cells than males, and males had a significantly lower absolute count and percentage of CD19+ B cells than females. These findings are consistent with a study of US adolescents (13–19 y) in which CD19+ B cells were reported to significantly decrease in males as age increased .
In the present study, we found significant gender differences in red blood parameters (RBC, HB, HCT and MCV), with males having higher values than females; these results are consistent with previous reports , , , , , , , , . The reason for this gender difference may be due to the influence of androgen on erythropoiesis and menstrual blood loss in females , , . Consistent with previous reports of African populations, a gender difference in platelet counts was seen, with females having higher values than males , , . We also noted significant gender differences in total WBC count, again with females having higher values than males. This result is consistent with reports from Uganda , Ghana  and Kericho, Kenya  but differs from reports from Ethiopia , Central African Republic , and western Kenya , where total white blood cell counts did not vary between genders. In a UK study that included women of different ethnic origins (Caucasian, Afro-Caribbean and African) gender differences in total WBC and platelet counts were reported for all ethnic groups; women had higher values than men .
Most of the hematological values derived from this study were lower than those derived from the US population, which is consistent with studies of similar age groups conducted in western Kenya and Uganda , . The lower limits for many parameters in the present study were lower than those derived from those two African countries. Notably, close to 50% of our study participants had Hb values that were outside the lower limits of the range derived from a US population, with 69.3% of females showing Hb values outside of the lower limit. Factors such as poor nutritional status, genetic red blood cell disorders or parasitic infections have been suggested to account for low Hb values. While a fairly low frequency of sickle cell trait (5.6%) was reported in an early study of pregnant women in Mozambique , a recent national survey of schistosomiasis and soil-transmitted helminthes showed that parasitic infections are common .
Clinical chemistry reference ranges derived in the present study were comparable to those derived from the North American population, with a few exceptions. The upper limits of the reference ranges for ALT, AST, Bilirubin and cholesterol derived in this study were slightly higher than those from the US, which is consistent with results from Uganda . The Maputo reference values were lower than those reported in western Kenya for the same age group .
The clinical reference ranges derived from US MGH have been used in most clinical research studies due to the absence of local ranges. A comparison between values derived from the present study and US MGH values  reveals higher variations in most values, especially in hematology. If US MGH ranges were used as inclusion and/or exclusion criteria in potential clinical studies, 159 (62%) of the participants of the present study would be excluded due to abnormal laboratory parameters. However, if the local reference ranges were used, only 40 (16%) of the participants would be excluded. The use of improper reference ranges has been reported to affect the time period of trial enrollment due to the large number of participants that must be screened to reach the requested target sample size. The long enrollment period impacts both workload and study cost .
The Division of AIDS (DAIDS) toxicity table for grading adverse events is used in many clinical trials but may not be appropriate for the young adults included in the present study. For example, the lower limit ranges of neutrophil (1200 cells/µl), Hb (7.7 d/L) and AST (13.7 U/L) would qualify as grade 1, 3 and 1 AEs, respectively. The upper limits for ALT (48.2 U/L) and T-Bilirubin (27.9 µmol/L or 1.63 mg/dl) would be qualified as a grade 1 AEs. This is consistent with results found in Uganda  and western Kenya .
Several limitations are apparent in the design of the present study. Importantly, we did not screen for all medical conditions that might have influenced the laboratory findings. Participants with parasitic infections or with other subclinical conditions may have been included, which may have influenced the results. Secondly, the participants were recruited at a youth health center in Maputo and the findings may therefore not be generalizable to young adults in the rest of Mozambique or other African countries. Despite these limitations, the reference values established here are being used in ongoing HIV-vaccine trials in Maputo that recruit healthy volunteers from the cohort of young adults attending the Maputo Central Hospital Youth Clinic. Additionally, given that indigenous reference values have not been previously defined for any Mozambican population, the ranges defined in this study are suitable for use in assessment of young adults in a more generalized setting.
In conclusion, this study is the first to assess clinical laboratory reference ranges in Mozambique. The hematological and biochemistry reference values from this African population differ from those derived from a North American population. This study also highlights the need for region-specific clinical reference ranges for patient management and clinical research.
Conceived and designed the experiments: NT IJ CN SA NO EM NS. Performed the experiments: NT EA NS. Analyzed the data: OJ NT. Wrote the paper: NT CN IJ SA. Served as study physicians: EV EG EM. Revised and approved the final version of the manuscript: NT CN IJ SA OJ EV NO EA EM EG NS.
- 1. Mgone CS, Makanga M (2010) Fighting HIV/AIDS, tuberculosis and malaria: one world, one partnership. Tropical Medicine and International Health 15: 973–974.
- 2. Mgone CS, Salami W (2009) EDCTP: a genuine north–south partnership.Tropical Medicine and International Health. 14: 1327–1328.
- 3. Tsegaye A, Messele T, Tilahun T, Hailu E, Sahlu T, et al. (1999) Immunohematological reference ranges for adult Ethiopians. ClinDiagn Lab Immunol 6: 410–414.
- 4. Lugada ES, Mermin J, Kaharuza F, Ulvestad E, Were W, et al. (2004) Population-based hematologic and immunologic reference values for a healthy Ugandan population. ClinDiagn Lab Immunol 11: 29–34.
- 5. Menard D, Mandeng MJ, Tothy MB, Kelembho EK, Gresenguet G, et al. (2003) Immunohematological reference ranges for adults from the Central African Republic. ClinDiagn Lab Immunol 10: 443–445.
- 6. Saathoff E, Schneider P, Kleinfeldt V, Geis S, Haule D, et al. (2008) Laboratory reference values for healthy adults from southern Tanzania. Trop Med Int Health 13: 612–625.
- 7. Omosa-Manyonyi G, Jaoko W, Anzala O, Ogutu H, Wakasiaka S, et al.. (2011) Reasons for ineligibility in Phase 1 and 2A HIV vaccine clinical trials at Kenya Aids Vaccine Initiative (KAVI), Kenya. PLoS ONE 6: : e 14580.
- 8. Clerici M, Butto S, Lukwiya M, Saresella M, Declich S, et al. (2000) Immune activation in Africa is environmentally-driven and is associated with upregulation of CCR5. Italian-Ugandan AIDS Project. Aids 14: 2083–2092.
- 9. Stevens W, Kamali A, Karita E, Anzala O, Sanders EJ, et al. (2008) Baseline morbidity in 2,990 adult African volunteers recruited to characterize laboratory reference intervals for future HIV vaccine clinical trials. PLoS ONE 3: e2043.
- 10. Eller LA, Eller MA, Ouma B, Kataaha P, Kyabaggu D, et al. (2008) Reference intervals in healthy adult Ugandan blood donors and their impact on conducting international vaccine trials. PLoS ONE 3: e3919.
- 11. Zeh C, Amornkul PN, Inzaule S, Ondoa P, Oyaro B, at al (2011) Population-based biochemistry, immunologic and hematological reference values for adolescents and young adults in a rural population in western Kenya. PLoS ONE 6: e21040.
- 12. Kratz A, Ferarro M, Sluss PM, Lewandrowski KB (2004) Case records of the Massachusetts General Hospital. Laboratory reference values. N Engl J Med 351: 1548–1563.
- 13. DAIDS (2004) Division of AIDS table for grading the severity of adult and pediatric adverse events. Bethseda, MD, USA: DAIDS.
- 14. Melo J, Folgosa E, Manjate D, Osman N, Francois I, et al. (2008) Low prevalence of HIV and other sexually transmitted infections in young women attending a youth counselling service in Maputo, Mozambique. Tropical Medicine and International Health 13: 17–20.
- 15. Vicente EM, Jermy CA, Schreiner HD (2009) Urban geology of Maputo, Mocambique. Engineering geology of tomorrow's cities. Geological Society, London, Engineering Geology Special Publication, 22..
- 16. CMM (2010) Perfil Estatístico do Município de Maputo. In: Sócio Económicas. Conselho Municipal de Maputo.
- 17. R Core Team (2013) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/
- 18. Clinical and Laboratory Standards Institute (2008) Defining, establishing, and verifying reference intervals in the clinical laboratory; approved guideline-Third Edition.National Committee for Clinical Laboratory Standards, Wayne, PA, USA C28-A3, Vol.28 No 30.
- 19. Schnabl K, Chan MK, Gong Y, Adeli K (2008) Closing the Gaps in Paediatric Reference Intervals: The CALIPER Initiative. ClinBiochem Rev 29: 89–96.
- 20. Coskun A, Ceyhan E, Inal TC, Serteser M, Unsal I (2013) The comparison of parametric and nonparametric bootstrap methods for reference interval computation in small sample size groups. Accreditation and Quality Assurance, 18(1): , 51–60.
- 21. Chernick MR, LaBudde R A (2011) An Introduction to Bootstrap Methods with Applications to R. John Wiley & Sons
- 22. Kibaya RS, Bautista CT, Sawe FK, Shaffer DN, Sateren WB, et al. (2008) Reference ranges for the clinical laboratory derived from a rural population in Kericho, Kenya. PLoS ONE 3: e3327.
- 23. Maini MK, Gilson RJ, Chavda N, Gill S, Fakoya A, et al. (1996) Reference ranges and sources of variability of CD4 counts in HIV-seronegative women and men. Genitourin Med 72: 27–31.
- 24. Rudy BJ, Wilson CM, Durako S, Moscicki AB, Muenz L, et al. (2002) Peripheral blood lymphocyte subsets in adolescents: a longitudinal analysis from the REACH project. ClinDiagn Lab Immunol 9: 959–965.
- 25. Urassa WK, Mbena EM, Swai AB, Gaines H, Mhalu FS, et al. (2003) Lymphocyte subset enumeration in HIV seronegative and HIV-1 seropositive adults in Dar es Salaam, Tanzania: determination of reference values in males and females and comparison of two flow cytometric methods. J Immunol Methods 277: 65–74.
- 26. Oladepo DK, Idigbe EO, Audu RA, Inyang US, Imade GE, et al. (2009) Establishment of reference values of CD4 and CD8 lymphocyte subsets in healthy Nigerian adults, Clin Vaccine Immunol. 16: 1374–1377.
- 27. Dosoo DK, Kayan K, Adu-Gyasi D, Kwara E, Ocran J, et al. (2012) Haematological and biochemical reference values for healthy adults in middle belt Ghana. PLoS ONE 7: e36308.
- 28. Subhashree AR, Parameaswari PJ, Shanti B, Revathy C, Parijatham BO (2012) The reference intervals for the haematological parameters in healthy adult population of Chennai, southern India. J ClinDiag Research 6: 1675–1680.
- 29. Wakeman L, Al-Ismail S, Benton A, Beddall A, Gibbs A, et al. (2007) Robust, routine haematology reference ranges for healthy adults. IntJnl Lab Hem 29: 227–283.
- 30. Bain B, Seed M, Godsland I (1984) Normal values for peripheral blood white cell counts in women of four different ethnic origins. J ClinPathol 37: 188–193.
- 31. Bain BJ (1996) Ethnic and sex differences in the total and differential white cell count and platelet count. J ClinPathol 49: 664–666.
- 32. Willcox MC, Liljestrand J, Bergström S (1986) Abnormal heamoglobin among pregnant women from Mozambique. J Med Genet 23: 151–152.
- 33. Augusto G, Nalá R, Casmo V, Sabonete A, Mapaco L, et al. (2009) Geographic distribution and prevalence of Schistosomiasis and soil-transmitted helminthes among schoolchildren in Mozambique. Am J Trop Med Hyg 81: 799–803.