IntroductionrnChronic Kidney Disease CKD is a progressive reduction in renal function 1. It is arncondition where the kidneys lose their normal function especially excretory and regulatoryrnfunctions which can be due to infections autoimmune diseases diabetes hypertensionrncancer and toxic chemicals 2. CKD is heading towards becoming a major health problem 3rnand is rapidly assuming epidemic proportions globally 4. India has highest number ofrndiabetics in the world having a prevalence of 3.8 in rural and 11.8 in urban adults 3. It isrnassociated with adverse outcomes in all stages of CKD 5. It has been estimated thatrnapproximately 25-40 of diabetic and hypertensive patients usually develop CKDrnNephropathy 3. Studies conducted on renal patients revealed that up to 90 were found tornhave oral symptoms of uremia like ammonia like taste and smell stomatitis gingivitisrndecreased salivary flow xerostomia and parotitis 2. The objectives of early diagnosis isrnidentification of asymptomatic disease at that time when intervention has a reasonablernpotential of a positive impact on outcome 3 Elevated fasting lipid profiles including total cholesterol TChol non-high density lipoprotein cholesterol non-HDLChol low density lipoprotein cholesterol LDL-Chol andrntriacylglycerol TG and low levels of high densityrnlipoprotein cholesterol HDL-Chol are associated with thernrisk of cardiovascular diseases including heart attack strokernand atherosclerosis Slhessarenko et al. 2015 13rn. Highrndensity lipoprotein cholesterol HDL-Chol goodrncholesterol controls the level of low density lipoproteinrncholesterol LDL-Chol bad cholesterol by transporting fatrnmolecules out of the arterial walls and decreasingrnmacrophage accumulation resulting in preventing orrnregressing atherosclerosis. Cardiovascular diseases are thernmajor cause of death in many countries including KenyarnSlhessarenko et al. 2015 13rn. Estimation of the normalrnstatus or otherwise of a medical report of lipid profilernanalytes for individual patients is based on comparing thernanalytes value with the reference interval value. A referencerninterval is defined as a value of a parameter or analyternobtained by observation or measurement from a referentrnindividual selected on the basis of fulfilling a specificrninclusion criteria. Reference intervals are estimated using arnminimum sample size of 120 referent individuals and spansrn95 of the sample referent population. Where stratificationrnis required on the basis of sex and age the minimum samplernsize for each age and sex group is 120 referent individualsrnCLSI EP28 3c guideline 2010 5rn. Reference intervalsrnfor fasting lipid profiles are used by clinicians to assess thernrisk of cardiovascular diseases in patients using relativernratios such as total cholesterol: high density lipoproteinrncholesterol low density lipoprotein cholesterol: high densityrnlipoprotein cholesterol triacylglycerols: high densityrnlipoprotein cholesterol of different fractions of plasma lipidsrnto diagnose cardiovascular diseases assess the stage ofrncardiovascular disease assess toxicity to xenobiotics and tornmonitor response to therapy used in managingrncardiovascular diseases patients Achila et al. 2017 1rn.rnReference intervals of fasting lipid profiles for adults andrngeriatrics from different parts of the world are affected byrnfactors such as age sex geographical location dietaryrnhabits foods taken life-style sedentary or active smokerrnor non-smoker alcoholic or non-alcoholic health statusrnenvironment climate comprehensiveness of the inclusionrncriteria used in recruiting the referent individual socioeconomic status of the referent individual ethnicityrngenetics and the method and reagents used to estimate thernlipid profile analyte of interest of the referent individualrnSlhessarenko et al. 2015 13rn. For this reason thernInternational Federation of Clinical Chemistry andrnLaboratory Medicine IFCC requires that each clinicalrnlaboratory develops its own local reference intervals forrnfasting lipid profiles analytes for adults and geriatrics usingrnits own local population for accurate early disease detectionrnmonitoring response to therapy and predicting clinicalrnoutcomes. Despite the fact that there cannot be universalrnreference interval limits for biological parameters mostrnAfrican countries including Kenya use fasting lipid profilernreference intervals for adults and geriatrics developed usingrnthe Caucasian/Western population Europeans and NorthrnAmericans whose diets lifestyle and genetics differ fromrnthat of Africans reported in medical literature medicalrnbooks refereed research articles in medical journals andrndiagnostic test manufacturers insert reagent kits Liu et al.rn2019 11 which is inappropriate. For the Kenyan populationrnincluding that of the Taita-Taveta County there are nornreported fasting lipid profile reference intervals for adultsrnand geriatrics. There is therefore a dare need to developrnfasting lipid profile reference interval for the adults andrngeriatrics of Taita-Taveta County Kenya. The aim of thisrnstudy therefore was to establish age and sex specific 95rnreference interval limits for fasting lipid profiles for adultsrnand geriatrics of Taita-Taveta County Kenya and comparernthem with those previously reported in medical literature.
Materials and MethodsrnStudy ArearnThis study was conducted in four subcounties MwataternTaveta Wundanyi and Voi of Taita-Taveta County Kenyarnbetween May 2015 and December 2017.rnStudy PopulationrnThe study population involved a total of 271 referentrnindividuals who had fasted for between eight and twelvernhours of age 18-93 years. They comprised 123 males andrn148 females recruited from the four subcounties of TaitaTaveta County Kenya who fulfilled the inclusion criteria.rnTo participate in this study the referent individual requiredrnto be above eighteen years fulfilled the inclusion criteriarnand consented to participate by signing the consent forms.rnStudy DesignrnThis was a cross-section study design involving 271rnrandomly recruited male and female adults and geriatrics ofrnTaita-Taveta County Kenya.rnEthical ConsiderationsrnThis study protocol was approved by Kenyatta UniversityrnEthical Review Committee KU-ERC Ref NumberrnI84/31987/15/ National Commission for SciencernTechnology and Innovation NACOSTI Ref numberrn16/22096/14531 and Taita-Taveta county Medical director.rnRecruitment of the Referent IndividualsrnThe 271 referent subjects involved in this study were drawnrnfrom the four sub-counties of Taita Taveta County Kenya.rnThe referent individuals were recruited by the researcherrnand his team mates by meeting the potential participants inrnthe churches chief barazas and door to door visits andrnexplaining to them the purpose of the study and itsrnimportance to them. Those interested in participating in thernstudy were then recruited and given instructions to fast afterrntaking their super. The exclusion criteria used to select thernreferent individuals based on a self-reporting questionnairernand physical clinical observation by a physician includedrnparticipants who consented but were not obese pregnant onrnstrenuous exercise suffering from diabetes mellitusrnabnormal lipid profiles and acanthosis nigerians consumersrnof alcohol and medications that modify fasting lipid profilesrnincluding taking coffee. In addition those with personalrnhistory of hypertension thyroid hormone renal or liverrndisease acute myocardial infarction angina arrhythmiarnand/or cerebrovascular disease were also excluded fromrnparticipating in the study. Also excluded were potentialrnparticipants of age below 18 years and those positive forrnHIV/AIDS syphilis and hepatitis B and C.rnBlood Sample CollectionrnAfter eight to twelve hours of fasting five milliiters syringernwas used to draw five milliters of blood from the auxiliary in from each referent individual into plain vacutainerrntubes labeled with the study number and the name of thernreferent individual. This blood sample was allowed to clotrnand then placed into an ice box and transported to thernanalysis laboratory where each sample was centrifuged atrn3000g for five minutes. Serum was separated using arnPasteur pipette and then transferred into duplicate vials withrnbar codes for identification purposes and stored at -20ornCrnuntil needed for analysis of fasting lipid profiles usingrnClinical Chemistry Autoanalyzer Integra 400 machinernwithin three months.rnLaboratory AnalysisrnThe fasting lipid profiles analyzed using the discreternClinical Chemistry Autoanalyzer Integra 400 machinernincluded total cholesterol T-Chol using an enzymaticrnmethod involving cholesterol esterase cholesterol oxidasernand peroxidase high density lipoprotein cholesterol HDLChol using an enzymatic method involving antihuman-lipoprotein cholesterol esterase cholesterol oxidase andrnperoxidase low density lipoprotein cholesterol LDL-Cholrnusing an enzymatic method involving cholesterol esteraserncholesterol oxidase sodium azide and peroxidase andrntriacylglycerols TG using an enzymatic method involvingrnlipase glycerol kinase glycerol-3-phosphate oxidase andrnperoxidase. Non-high density lipoprotein cholesterol wasrnobtained by subtracting high density lipoprotein cholesterolrnHDL-Chol value from total cholesterol T-Chol value forrneach referent individual. The ratio of total cholesterol tornhigh density lipoprotein cholesterol low density lipoproteinrncholesterol to high density lipoprotein cholesterol andrntriacylglycerols to high density lipoprotein cholesterol werernalso calculated for each referent individual.rnData Management and Statistical AnalysisrnLaboratory obtained values for all the measured andrncalculated fasting lipid profile parameters were recordedrninto the laboratory notebook entered into the Excelrnspreadsheet checked for errors and cleaned. This clean datarnwas exported into SPSS software and tested for normalityrnusing mean standard error of mean median modernvariance standard deviation Skewness standard error ofrnskewness SES Kurtosis standard error of kurtosis SEKrnrange minimum and maximum and 2.5 and 97.5rnpercentiles. Outliers were removed using Box plots usingrnthe formula Q31.5 IQR for the upper limit and Q1-1.5rnIQR for the lower limit where IQR Q3-Q1. IQR is therninterquartile range. Q1 and Q3 are the interquartile 1 andrninterquartile 3 respectively. Q2 is the median or interquartilern2 Bauman 2014 4rn. Since Clinical Laboratory StandardsrnInstitute CLSI 2010 5 recommends that referencernintervals be expressed as median and 95 range 2.5 andrn97.5 percentiles the difference between male and femalernvalues for each parameter for the referents were comparedrnusing Mann-Whitney U test. Further the difference betweenrnthe two age categories for this data for each gender was alsorncompared using Mann-Whitney U test. A 0.05 wasrnconsidered significant.rnResultsrnResults of the Quality Control lipid profile materialrnThe results of the daily analysis of the quality controlrnmaterial presented in Table i indicates that the study qualityrncontrol results were similar to the assigned quality controlrnvalues. This implies that the analytical process wasrnoperating optimally and therefore the results generated usingrnthis analytical process were accurate and reliable.
ResultsrnResults of the Quality Control lipid profile materialrnThe results of the daily analysis of the quality controlrnmaterial presented in Table i indicates that the study qualityrncontrol results were similar to the assigned quality controlrnvalues. This implies that the analytical process wasrnoperating optimally and therefore the results generated usingrnthis analytical process were accurate and reliable.rnTable 1: Results for quality control material for lipid profilesrnAnalyte unitrnAssigned QC report Study QCrnreportrnUpperrnlimit Target Lowerrnlimit Mean SD CVrnT-CHOL mmol/L 6.99 6.61 6.63 6.59 0.13 1.97rnHDL-CHOL mmol/L 2.01 1.61 1.43 1.62 0.03 2.85rnLDL-CHOL mmol/L 2.6 2.3 2.1 2.2 0.05 2.27rnTG mmol/L 2.18 2.10 1.95 2.10 0.03 1.43rnResults of the normality statistics for fasting lipidrnprofiles for adults and geriatrics of Taita-Taveta CountyrnKenyarnThe results of the normality statistics for fasting lipidrnprofiles for adults and geriatrics of Taita-Taveta CountyrnKenya are presented in Table 2. Results indicate that therndata for total cholesterol T-Chol and low densityrnlipoprotein cholesterol are normally distributed for thernwhole sample combined males and females separate malesrnand females for the two age categories 18-55 years and 56-rn95 years and whole sample 18-95 years based on thernvalue for skewness 1 through zero to -1 and kurtosis 2rnthrough zero to -2 in addition to the values for meanrnmedian and mode. For high density lipoprotein cholesterolrnHDL-Chol the data for males in age category 18-55 yearsrnskewness kurtosis males in age category 56-95 yearsrnkurtosis and males in whole sample 18-95 yearsrnkurtosis are not normally distributed based on the value ofrnmean median mode skewness and kurtosis. For non-highrndensity lipoprotein cholesterol non HDL-Chol the data forrnmales in age category 18-55 years kurtosis is not normallyrndistributed while that for age category 56-95 years andrnwhole sample 18-95 years are normally distributed basedrnon the value of mean median mode skewness and kurtosis.rnFor triacylglycerols TG the data for combined males andrnfemales and separate males in age category 18-55 years arernnot normally distributed skewness that for combinedrnmales and females and separate males and females in agerncategory 56-95 years and whole sample 18-95 years arernnot normally distributed mean median mode skewnessrnand kurtosis. The data for total cholesterol: high densityrnlipoprotein cholesterol ratio triacylglycerols: high densityrnlipoprotein cholesterol ratio and low density lipoproteinrncholesterol: high density lipoprotein cholesterol ratio in agerncategory 18-55 years is not normally distributed meanrnmedian mode skewness and kurtosis combined males andrnfemales and separate males for total cholesterol: highrndensity lipoprotein cholesterol ratio and low densityrnlipoprotein cholesterol: high density lipoprotein cholesterolrnratio mean median mode skewness and kurtosis andrncombined males and females and separate male and femalesrnin age category 56-95 years and whole sample 18-95rnyears are not normally distributed mean median modernskewness and kurtosis based on the value of mean medianrnmode skewness and kurtosis Table ii. Reference interval limits for fasting lipid profile forrnadults and geriatrics of Taita Taveta County KenyarnThe established median reference interval for low densityrnlipoprotein cholesterol LDL-Chol non-high densityrnlipoprotein cholesterol non-HDL-Chol triacylglycerolsrnTG and the ratio of low density lipoprotein cholesterolrnLDL-Chol: high density lipoprotein cholesterol HDLChol for male adult and geriatric population of TaitaTaveta County Kenya were statistically similar to those ofrnthe female population of similar age range 0.05.rnTherefore a combined reference interval of these parametersrnfor this population was established. The established medianrnreference interval limits for adults and geriatrics of TaitaTaveta County Kenya for low density lipoproteinrncholesterol LDL-Chol is 3 1-6 mmol/L non-high densityrnlipoprotein cholesterol non-HDL-Chol is 4 1-6 mmol/Lrntriacylglycerols TG is 2 1-4.2 mmol/L and low densityrnlipoprotein cholesterol LDL-Chol: high density lipoproteinrncholesterol HDL-Chol ratio is 3 1-6.rnHowever the established median reference interval for thernadult and geriatric male population of Taita-Taveta CountyrnKenya for total cholesterol T-Chol high densityrnlipoprotein cholesterol HDL-Chol total cholesterol TChol: high density lipoprotein cholesterol HDL-Cholrnratio and triacylglycerol TG: high density lipoproteinrncholesterol HDL-Chol ratio were statistically significantlyrndifferent from that of the female population of similar agernrange 0.05. The established median reference intervalrnlimits for the adult and geriatric population of Taita-TavetarnCounty Kenya for total cholesterol T-Chol for males 5rn2-7.9 mmol/L with mean rank of 125.57 are significantlyrnlower than those for females 5 2-8 mmol/L with meanrnrank of 144.67 U 7819 z -2.061 0.039 r rn0.1252 high density lipoprotein cholesterol HDL-Cholrnfor males 1 0-2 mmol/L with mean rank of 121.81 arernsignificantly lower than those for females 1 0-2.28rnmmol/L with mean rank of 147.79 U 7356.5 z -3.579rn 0.000 r 0.2174 total cholesterol T-Chol: highrndensity lipoprotein cholesterol HDL-Chol ratio for malesrn4 2-10.9 with mean rank of 146.76 are significantlyrnhigher than those for females 4 1.73-8 with mean rank ofrn127.06 U 7778.5 z -2.099 0.036 r 0.1275 andrntriacylglycerols TG: high density lipoprotein cholesterolrnHDL-Chol ratio for males 1 0-6.9 with mean rank ofrn145.96 are significantly higher than those for females 1 0-rn6.28 with mean rank of 127.72 U 7876.5 z -2.065 rn 0.039 r 0.1254 Table iii. Effect of age and gender on the reference intervals forrnfasting lipid profiles for adults and geriatrics of TaitarnTaveta County KenyarnThe effect of age and gender on the reference intervalsrnlimits for fasting lipid profiles for the adult and geriatricrnpopulation of Taita Taveta County Kenya was investigatedrnby categorizing the results into two age groups: a agerngroup 1 18-55 years and b age group 2 56-95 years.rnThe statistical difference between the reference intervals forrnfasting lipid profiles for adult and geriatric males andrnfemales were estimated using Mann-Whitney U test and values less than 0.05 was considered statistically significant.rnThe Median and 95 range for males and females for therntwo age groups are presented in Table iv. Results indicaternthat all the measured and calculated fasting lipid profilernanalytes including total cholesterol Chol high densityrnlipoprotein cholesterol HDL-Chol low density lipoproteinrncholesterol LDH-Chol non-high density lipoproteinrncholesterol non-HDL-Chol and the ratios of totalrncholesterol Chol: high density lipoprotein cholesterolrnHDL-Chol ratio low density lipoprotein cholesterolrnLDL-Chol: high density lipoprotein cholesterol HDLChol ratio and triacylglycerol TG: high densityrnlipoprotein cholesterol HDL-Chol ratio were statisticallyrnunaffected by advancement in age 0.05 Table iv.rnHowever in the age category 18-55 years males 1 0.45-2rnmmol/L with mean rank of 61.39 had a statisticallyrnsignificantly lower levels of high density lipoproteinrncholesterol HDL-Chol compared to females 1 0-2rnmmol/L with mean rank of 75.21 U 1846 z -2.666 rn 0.008 r 0.2269. Further in age category 56-95 yearsrnmales 1 0-2 mmol/L with mean rank of 60.75 had arnstatistically significantly lower levels of high densityrnlipoprotein cholesterol than females 1 0-3 mmol/L withrnmean rank of 73.16 U 1798.5 z -2.416 0.016 r rn0.2095 and males 5 2-16 with mean rank of 75.10 had arnstatistically significantly higher levels of total cholesterolrnT-Chol: high density lipoprotein cholesterol HDL-Cholrnratio than females 4 0.7-7.9 with mean rank of 59.02 Urn 1676.5 z -2.445 0.014 r 0.2120. Comparison of the developed reference interval limitsrnfor fasting lipid profile for adults and geriatrics of TaitaTaveta County Kenya with those reported in medicalrnliteraturernA comparison of the developed reference intervals forrnfasting lipid profile for adults and geriatrics of Taita-TavetarnCounty Kenya with those reported in medical literature arernpresented in Table v. For this study the lower limit of thernreference interval for fasting total cholesterol T-Cholrnlevels is similar to that of the Haryana population but thernupper limit is greater than that of the Haryana population.rnHowever the reference intervals for fasting total cholesterolrnwere similar for males and females of both populations. Thernlower limit for the reference interval of fasting totalrncholesterol for Punjab population is higher than thatrngenerated in this study for the Taita-Taveta countyrnpopulation while the upper limit are lower than thatrngenerated in this study. Further the reference interval forrnfasting total cholesterol for the Punjab population differ byrnsex while those generated in this study are similar. Thernlower reference interval limit for fasting total cholesterol forrnthe Assamese population of India were similar to thoserngenerated for the Taita-Taveta County population while thernupper reference interval limit is lower than that generated inrnthis study. However the Assamese population referencerninterval for fasting total cholesterol differs by sex. For thernWestern Maharashtra population of India the lower andrnupper reference interval limit for fasting total cholesterol arernlower than those developed for the Taita-Taveta Countyrnpopulation of this study. Further the Western Maharashtrarnpopulation had different reference interval for fasting totalrncholesterol for males and females. For the VenezuelanrnMaracaibo population the lower reference interval forrnfasting total cholesterol is higher than that established forrnTaita-Taveta County population in this study while thernupper limit for the same population is lower than that of thernTaita-Taveta County population. However the VenezuelanrnMaracaibo population has different reference intervals forrnfasting total cholesterol for males and females Table v.rnFor the Punjab Western Maharashtra Assamese and TaitaTaveta populations all had different reference interval limitsrnfor fasting lipid profiles for males and females. For thernPunjab population the lower reference interval limit forrnfasting high density lipoprotein cholesterol HDL-Chol isrnhigher than that established for the Taita-Taveta Countyrnpopulation while the upper reference interval limit is lower.rnFor the Western Maharashtra and Assamese populations thernlower reference interval limit for fasting high densityrnlipoprotein cholesterol are higher than those established forrnthe Taita-Taveta County population while the upperrnreference interval limits were lower. For the Haryana andrnVenezuelan Maracaibo populations the combined male andrnfemale lower reference interval limits for fasting highrndensity lipoprotein cholesterol are higher than thernestablished separate for males and females for the TaitaTaveta County population while the upper referencerninterval limit is lower Table v.rnPunjab Western Maharashtra and Haryana populationrnreported combined reference intervals for TG for adult andrngeriatric males and females just like that established for thernadult and geriatric male and female population of TaitaTaveta County Kenya. However the lower combined established reference interval limits for triacylglycerol forrnadult and geriatric male and female population of TaitaTaveta County Kenya is lower than that reported for Punjabrnpopulation and the upper reference interval limit is higherrnthe lower combined established reference interval limits forrntriacylglycerol for adult and geriatric male and femalernpopulation of Taita-Taveta County Kenya is higher thanrnthat reported for Western Maharashtra population and upperrnreference interval limit is lower the lower combinedrnestablished reference intervals for TG for adult and geriatricrnmale and female population of Taita-Taveta County Kenyarnis similar to that reported for the Haryana population andrnthe upper reference interval limit is lower. Further the lowerrncombined established reference interval limits forrntriacylglycerol for adult and geriatric male and femalernpopulation of Taita-Taveta County Kenya is higher than thernreported separate male and female reference interval limitrnfor Venezuelan Maracaibo and Assamese population andrnthe upper separate reference intervals for the twornpopulations are lower. In addition the established lowerrnseparate male and female reference interval limit for totalrncholesterol: high density lipoprotein cholesterol ratio for thernadult and geriatric population of Taita-Taveta CountyrnKenya is lower than that reported for the combinedrnreference interval for the Punjab population and the upperrnseparate reference interval limit is lower Table v.
In conclusion this study generated sex and age specificrnfasting lipid profile reference interval limits for adults andrngeriatrics of Taita-Taveta County Kenya which wasrndifferent from those previously reported in medicalrnliterature from other parts of the world. These developedrnfasting lipid profiles reference intervals for adult andrngeriatric population of Taita-Taveta County Kenya can bernadopted and used to make appropriate clinical decisionsrnleading to improved diagnosis of cardiovascular diseases.