Asymptomatic Urinary Tract infection in Pregnant and Non-pregnant Women in the Limbe Health District of Cameroon: A Phenotypic and Biochemical analytic study

Background Asymptomatic bacteriuria in pregnancy is associated with signicant maternal and fetal risks. However, there seem to be a dearth of information of its prevalence in our community. This study investigated the prevalence of the etiology and antibiotic susceptibility prole of bacterial agents incriminated in urinary tract infection (UTI) in pregnant and non- pregnant women. The association between leukocyte esterase, nitrite, serum creatinine and asymptomatic urinary tract infection (UTI) was assessed. and using microbiological and for detection of isolates and their antibiotic sensitivity and level, data the Chi-square (X 2 ) test was used to examine differences at 5% level. Citrobacter diversus and Shigella species were most susceptible to ciprooxacin (100%), while Enterococci, Lactobacillus and Citrobacter diversus were most susceptible to ceftriaxone (100%). Asymptomatic UTI had no signicant effect on serum creatinine level in the study population. Leukocyte esterase was a signicant indicator of asymptomatic UTI in non-pregnant women, while nitrite was not a signicant indicator of asymptomatic UTI in both study groups. The use of biochemical indicators in the diagnosis of UTI in our setting should be used with caution.


Background
Bacteriuria is the presence and growth of bacteria in the urinary tract. This may lead to urinary tract infection (UTI) [1]. UTI is a predisposing factor to adverse pregnancy outcomes and presents with some effects on renal function [2]. Hormonal and immunological changes in pregnancy, including high level of progesterone secretion, which could lead to stasis and decrease immunity are contributing factors to UTI [3]. Asymptomatic bacteriuria occurs in 2 to 7% of pregnant women. However, this prevalence can be up to 30% in developing countries [4]. A study on bacteriuria in pregnant women in three hospitals in Douala, Cameroon reported a signi cant prevalence of symptomatic bacteriuria of 9.9% and asymptomatic bacteriuria of 5.7% [1]. In another study in Buea, Cameroon, the prevalence of bacteriuria in pregnant women was 23.5% and 7.8% [5], for symptomatic and asymptomatic cases, respectively.
Escherichia coli and Staphylococcus saprophyticus account for about 80% of community-acquired UTI [6]. Other Gram-negative pathogens, including Pseudomonas spp, Enterobacter spp, Citrobacter spp, and Klebsiella spp, are also incriminated in urinary tract infection [6]. Gram-positive bacteria, including Enterococcus spp and Staphylococcus spp [7] can cause nosocomial UTI due to selective pressure from the antimicrobial agents used in hospitalized patients. The relatively high prevalence of UTIs during pregnancy, and the signi cant consequences for women and the pregnancy outcome justify screening of pregnant women for UTIs [8].
In a urinalysis, the presence of leukocyte esterase suggests the presence of white blood cells while nitrite test detects the products of nitrate reductase, an enzyme produced by many gram negative bacteria. Serum creatinine concentration remains the standard, single-point assessment for kidney function in pregnant populations, yet a normal range in pregnancy has not been established [9]. However, routine analysis of urine in pregnant women is generally limited to just two parameters; glucose and albumin; whereas UTI and kidney malfunction are certainly dangers to pregnancy as it may affect the general health of the mother and fetus in the future.
Biochemical indicators of UTI such as the presence of leucocyte esterase and nitrite are obviously rapid laboratory procedures, which can be incorporated into the testing panel of pregnant women in resource limited settings such as Cameroon, where these tests are available and affordable. Uncomplicated UTIs are generally self-limiting, but the social stigma associated to UTI is of concern to patients and, as such, they engage in unorthodox treatment or self-medication using antibiotics [10].
Without treatment, up to 20 to 30% of pregnant women with asymptomatic bacteriuria will develop a symptomatic urinary tract infection [11]. In a previous study [12], E. coli, S. aureus, Beta haemolytic Streptococci, coagulase negative Staphylococci and Group B Streptococci were reported to be 100% susceptible to nitrofurantoin. In another study, E. coli reported 91.7% and 88.3% susceptibility to cipro oxacin and ceftriaxone, respectively [8]. Chloramphenicol was also found to be successful in the treatment of UTI, being 100% effective against Klebsiella pneumoniae, Enterobacter aerogenes, and E. coli [13].
Bacterial resistance to antibiotics is one of the most challenging global health threats faced in modern medicine, and resistance of urinary tract pathogens has been reported to increase worldwide with a pattern which may vary over short periods, and depends on site of isolation and environmental conditions [14].
Clinicians are challenged in selecting appropriate treatment due to increase antibiotic resistance by these pathogens [15,16]. Studies have demonstrated geographic variation in etiologic characteristics of UTI and their resistance patterns to antibiotics [10].
In our environment, urine culture is not routinely done for antenatal patients. The dip stick test is the most widely used method for assessing urine in pregnant women. Consequently, the microbiological picture of such urine specimen cannot be fully ascertained, as the strip cannot quantify the extent of infection as well as provide indication for antimicrobial chemotherapy. To ensure appropriate therapy for these infections, current knowledge of the etiologic agents, their biochemical indicators and antibiotic susceptibility pro le of isolates are necessary for improved intervention. The aim of this study was to isolate and identify the bacterial agents causing asymptomatic UTI in pregnant and non-pregnant women in Limbe Health District, assess their association with certain biochemical indicators, serum creatinine level and their antibiotic susceptibility pro le.

Study Design
This was a hospital based cross sectional study, in which asymptomatic, 75 pregnant and 75 non-pregnant women, aged 13-44 years consulting in some health care facilities in the Limbe Health District, were investigated for asymptomatic UTI and kidney function. It was based on laboratory investigations with urine and blood samples obtained from the women. Data on sociodemographic characteristics of participants was obtained using a questionnaire.

Study site and participants
This study was conducted in health facilities representing the key health areas in the Limbe Health District, along the coast line of South West Region, Cameroon. Limbe is a cosmopolitan town, made of people from different tribes, cultures and background. This district has a population of about 118,210 inhabitants with a total of 60,228 women and 8,712 women of child bearing age [17]. Most of these women are sexually active, under educated, and make use of public toilets. Those who live around the beaches mostly use seawater and springs for personal hygiene, which can predispose them to UTIs.

Ethical Considerations
All methods were carried out in accordance with relevant local and national regulations. An administrative authorization was obtained from the South West Regional Delegation of Public Health; ethical approval obtained from the Institutional Review Board (IRB), Faculty of Heath Sciences, University of Buea with reference number 2020/1038-12/UB/SG/IRB/FHS. A verbal authorization was obtained from the different participating health facilities. An informed consent was signed by all subjects and an informed consent by parent or guardian of subjects less than 21 years.

Sample Collection
Samples were collected from all pregnant and non-pregnant women who gave their consent, without obvious clinical signs of UTI and/or kidney malfunction, and were not on antibiotics for the last 72 hours. Urine and blood samples were collected for culture, urine dipstick, and creatinine tests respectively. About fteen milliliters (15 mL) of clean-catch midstream urine was collected in sterile, dry, wide-mouth, leak-proof container. The women were instructed to clean the area around the urethral opening with absorbent paper, and collect urine while holding the labia apart. Five milliliters of blood were collected aseptically from each study participant by venipuncture and transferred to a sterile vacutainer tube without anticoagulant. Samples were transported on ice to the Microbiology Laboratory (urine) and to the Clinical Chemistry Laboratory (blood), of the District Hospital, Limbe, for analysis. In situations where delay was anticipated before analysis, samples were preserved in a refrigerator at 2-8 °C for at most 4 hours (urine samples) and 24 hours (blood).

Determination of Creatinine Level
Blood samples were allowed to clot by leaving them undisturbed at room temperature for 45 min, dislodged and centrifuged between 2200-2500 rpm (revolution per minute) for 10 minutes and approximately 2 mL of serum harvested. Creatinine test was performed, using the creatinine test kit (Cypress Diagnostic, Belgium) as described by Cheesbrough (2000) [18], and according to manufacturer's instruction. Brie y, reagent 1 (R1) and reagent 2 (R2) were reconstituted in equal proportion for use within seven days, stored at 2-8 o C. Quality control (QC) runs of normal and pathological commercial control sera (CYPRESS DIAGNOSTICS Belgium) were performed daily before analysis. Brie y, 1000 µ/L of the reagent was pipetted into each of ve sterile glass test tubes and labeled as follows: test tube 1(blank), test tube 2 (standard), test tube 3 (pathological control), test tube 4 (normal control) and test tube 5 (subject code). Exactly 100 µ/L of the standard, pathological control, normal control and 100 µ/L of participants' serum were added sequentially into respective reagent tubes. The content of each tube was mixed, its absorbance and concentration measured and recorded by spectrophotometry (Sinothinker China, at wavelength 630 nm) before reconstituting the next tube. Results for normal and pathological controls were recorded and veri ed to fall within +/-2SD before samples were tested, results accepted and recorded.

Urine Dipstick test
Urine samples were tested with the rapid urine dipstick (ACON Laboratories, Inc., China) by dipping the strip into urine and results for leucocyte esterase and nitrite obtained within 30 seconds to 1 minute [19]. Strips were dipped in the well-mixed urine for no longer than four seconds and excess urine, along the rim of strips, wiped. After 1 minute, the colour change in the strips was compared with the color scales provided with the kit. A violet colour produced was considered as leucocyte esterase positive, and pink colour as nitrate positive. Colour changes appearing only along the edges of the test patches or developing after more than 2 minutes were considered negative. Most gram negative uropathogens produce nitrate reducing enzymes that reduce nitrate to nitrites. Nitrite reacts with an amine impregnated on the dipstick pad to form a diazonium compound resulting in a pink colour reaction within 60 seconds. An enzyme produced by in ammatory cells, leukocyte esterase, reacts with the chloroacetate stain impregnated in a dipstick pad resulting in an iodoxyl moiety that is oxidized by room air and produces a colour change in 1-2 minutes.
Urine culture test Following standard procedure, exactly 0.01 mL of urine samples were inoculated aseptically on CLED agar (HiMedia Laboratories Pvt. Ltd.,India) by streaking technique. The culture plates were incubated aerobically at 37 °C for 24-48 hours, after which they were observed for growth. Colonies were counted and the counts multiplied by 100 to have the Colony Forming Units per milliliter (CFU/mL) of urine. Plates with growth or colonies less than 1000 CFU/mL were incubated for another 24 hours. Only plates with counts >10 5 CFU/mL were considered as signi cant bacteriuria [20]. The isolates were further identi ed on the basis of colonial morphology, cultural characteristics, Gram's staining and a battery of biochemical reactions [1,5].

Antimicrobial Susceptibility Testing
The susceptibility pattern of isolates to antibiotics employed in the empiric treatment of UTI was studied. Antimicrobial susceptibility tests were performed on bacterial isolates from midstream urine using antimicrobial discs, according to Clinical and Laboratory Standards Institute (CLSI) on Muller Hinton agar [21].
Brie y, 2-3 colonies from the pure culture were suspended in a vial of 7 mL physiological saline to obtain a turbidity equivalent to 0.5 McFarland. A sterile swab was dipped into the suspension and excess moisture removed by pressing the swab against the side of the tube. The surface of Muellar-Hinton agar was swabbed completely and allowed to dry for about 5 minutes. The following antibiotics from Fluka Industry, United Kingdom, were used: Amoxiclav (30 μg), Flucloxacillin (5 μg), Ampicillin (10μg), Ceftriaxone (30 μg), Ce xime (5 μg), Levo oxacin (5 μg), Cipro oxacin (30μg), O oxacin (5 μg) and Gentamicin (10 μg). Antibiotic disks were placed on the agar keeping at least a distance of 24 mm between them. About 15 minutes after disks were applied, the plates were inverted and incubated at 37 o C for 18-24 hours. Results were reported as sensitive or resistant based on the speci c diameter of zone of inhibition for the different antibiotics [22].

Statistical analysis
Data obtained was entered into excel spread sheet and analysed using SPSS statistical package. Descriptive statistics was used to determine the prevalence of asymptomatic UTI. The frequencies, percentages and measures of central tendencies were obtained. The Pearson's Chi-square was used to determine if there was correlation between the different variables. To compare the means, the student t-test was used. Statistical signi cance was taken at the level p < 0.05.

Socio-demographic Characteristics
A total of 150 women were enrolled in this study with age range of 13-44 years and a mean (standard deviation) of 26.97 (+6.17) years. Seventy-ve pregnant women with age range 16-38 years and a mean age (standard deviation) of 26.53 (+5.16) years, and 75 non-pregnant women with age range 13-44 years and a mean age (standard deviation) of 27.41 (+7.05) years participated in this study. Among the pregnant subjects, 28 (37.3 %) were between the ages 13-24 years, 44(58.7 %) between 25-34 years, and 3 (4.0 %) between 35-38 years (

Prevalence of asymptomatic UTI
The overall asymptomatic UTI prevalence in the study population was 30%. The prevalence in non-pregnant and pregnant women was 29.3% and 30.7%, respectively (Table 2). However, the difference was not statistically signi cant (P=0.859).  (7) and in non-pregnant 31.9 % (7) women. Shigella species (2.3 %) and Citrobacter diversus (2.3 %) were detected only in non-pregnant women. There was no signi cant difference of bacterial isolates in pregnant and non-pregnant women (P=0.482) ( Table 3). Association between leukocyte esterase/nitrite and Asymptomatic UTI There was no signi cant association between leukocyte esterase and nitrite with asymptomatic UTI in pregnant women. In non-pregnant women, leukocyte esterase was signi cantly associated to asymptomatic UTI (P= 0.009) ( Table 4). Furthermore, women who presented with positive leukocyte esterase were 25.3 % (19) and 9.3 % (7) pregnant and non-pregnant, respectively.

Relationship between Asymptomatic UTI and serum creatinine
Abnormal creatinine levels (≥1.4 mg/dL) were found more in pregnant women 10(13.3 %) than in non-pregnant women 9(12.0 %). However, there was no statistically signi cant association between serum creatinine level and asymptomatic UTI in pregnant (P=0.154) and non-pregnant women (P=0.617) ( Table   6).   Results of this study indicate that there is a relatively high prevalence of asymptomatic UTI (30.0 %) in pregnant and non-pregnant women in the Limbe Health District, with a higher occurrence of gram positive bacteria. The prevalence of 30.7 % in pregnant women is higher than 5.7 %, 7.8 %, 8.4 % and 9.0 %, respectively reported by other investigators [1,5,8,13], close to 28.7 % [23] and lower than 38.0 % reported by another [24]. The high prevalence in our study may be due to differences in sample size and methods of isolation, and to the fact that UTIs are one of the most frequent and recurrent bacterial infections in women. Furthermore, the prevalence in non-pregnant women (29.3 %) was lower than 75.0 % [24]. However, they studied a very small symptomatic nonpregnant population (n=24), a possible reason for their very high prevalence. Generally, women suffer from urinary tract infections more often because bacteria can reach the bladder easily because of the close proximity of the vagina and anus to the urethra [6], and di culty with hygiene due to a distended pregnant belly. During sexual activity, bacteria can be pushed into the urethra. The use of diaphragms could also be associated with the increase prevalence of asymptomatic UTI, as urine that stays in the bladder is more likely to grow bacteria and cause infection. Though we did not investigate this, it has been reported as a predisposing factor for UTI [6].
Gram-positive organisms were the most prevalent pathogens in both pregnant 87.0 % (n=23) and non-pregnant 68.3 % (n=22) women. This corroborates the ndings in another study [24], which reported a gram positive bacterial prevalence of 60.0 % in pregnant and 55.5 % in non-pregnant women; another study [8], also documented a prevalence of 80.95 % in pregnant women. Other researchers reported gram positive bacteria as frequent etiologic agents of UTI [7]. However, contrary to these ndings, Alemu et al (67.5 %) [13] and Ngong et al (85.4 %) [23], identi ed gram negative bacteria as the most prevalent urinary tract pathogen in pregnant women.
From our ndings, there was no signi cant difference in the etiologic pro le of asymptomatic UTI in pregnant and non-pregnant women. This may imply that UTI in women of child-bearing age, can be caused by same organisms in pregnant and non-pregnant women. Though we did not investigate the role of immunosuppression in relation to infection, we are constrained to think that, immunosuppression in pregnancy, as has been previously reported [1], may not have signi cantly in uenced occurrence of UTI in this study. S. saprophyticus occurred more in pregnant (26.2 %) than in non-pregnant women (4.5 %), while S. aureus occurred more in non-pregnant (22.8 %) than in pregnant women (17.4 %). E coli occurred more in non-pregnant (13.6 %) than in the pregnant (8.7 %) population. The dominant isolate in this study population was S. aureus which is similar to a previous report [24]. They reported over 45.0 % and 38.9 % prevalence in pregnant and non-pregnant population. S. saprophyticus was the most common bacteria isolated in pregnant women; however, this does not agree with other studies [1,5,8,13,23], who reported E. coli as the leading cause of UTI in pregnant women.
The occurrence of N. gonorrhoeae was noticed in pregnant (4.3 %, 1/23) and non-pregnant (9.1 %, 2/22) women. This could have been acquired through multiple mucosal sites in the lower genital tract, including the urethra, cervix, Bartholin's and Skene's glands [25]. Entrococci, Shigella and Citrobacter were among the least isolated microbes (2.2 % each) in the study population. The low prevalence of these isolates could be a result of the sample size and the detection method used in this study.

Association between Leucocyte esterase/Nitrite and asymptomatic UTI
The occurrence of Leucocyte esterase in pregnant women (25.3 %) was higher than in non-pregnant women (9.3 %). Within the pregnant population, the association between Leucocyte esterase and asymptomatic UTI was not statistically signi cant (p= 0.749), while a signi cant association was noticed in nonpregnant women (p=0.009). Being an indicator of immune response, its occurrence is remarkable. However, sterile pyuria is common in pregnancy, usually due to contamination by physiological vaginal discharge [26]. Sterile pyuria is a condition where a urinary dipstick test is leukocyte esterase positive with no associated positive urinary culture [27]. Pyuria could also occur as a result of the presence of non-detected bacterial and viral infections [28]. Moreover, about 58.7 % (44/75) of the pregnant population was within the most sexually active group (25-35years) which could be a predisposing factor to UTI. Over 49.3 % of the non-pregnant population also fell within this high sexually active group; a possible predisposing factor that could have accounted for signi cance in association between Leucocyte esterase and asymptomatic UTI, though Ngong et al [23] reported no statistically signi cant association between UTI and sexual activity.
Nitrite was virtually absent as an indicator of asymptomatic UTI in both study groups, probably because it is produced by speci c gram negative microbes, notably E .coli [29], whose occurrence was low in this study. The absence of nitrite may be related to the low frequency of such speci c isolates. This study reported a dominance of gram positive microbes, which mostly do not metabolize nitrate to nitrite. More so, microbial population density is also a factor related to the concentration of this waste product; and such population density may hardly occur in asymptomatic UTI cases [27].
Association between asymptomatic UTI and serum creatinine level Serum creatinine levels in asymptomatic bacteriuria in this study population ranged from 0.2 to 1.1mg/dL and 0.6 to 1.5mg/dL in pregnant and non-pregnant women, respectively. The physiologic increase in glomerular ltration rate during pregnancy results in a decrease in concentration of serum creatinine, which falls by an average of 0.4 mg/dl. [30]. Low serum creatinine level could also be due to low muscular activity [31]. More (13.3 %) of the pregnant women had abnormal values of serum creatinine according to biological reference range compared to non-pregnant women (12.0 %) with abnormal results. These abnormal serum creatinine values in pregnant women could be due to renal physiologic changes that occur during pregnancy, and not necessarily a UTI [32].
This nding maybe indicative of the effect of non-bacterial factors associated with renal malfunction such as immune mediated mechanisms [33].

Antimicrobial Sensitivity
Antimicrobial resistance has increased worldwide and susceptibility patterns show substantial geographic variation as well as differences in population and environment [34]. Treatment decreases the risk of pyelonephritis, low birth weight infants, and preterm births. In our study, most isolates were sensitive to ceftriazone and cipro oxacin. E. coli was highly sensitive (80.0 %) to cipro oxacin, in line with another study [35]. All three isolates of N. gonorrhoeae (1 pregnant and 2 non-pregnant women), were 100 % sensitive to cipro oxacin, and 66.7 % to ceftriazone.
S. aureus was 88.9 % susceptible to ceftriazone and 77.8 % to cipro oxacin. Susceptibility to cipro oxacin is similar to another nding [36] in Nigeria, where uropathogenic S. aureus isolates were reported to have a 60-70 % susceptibility to this antimicrobial agent.
Most isolates were completely resistant (100 %) to ce xime. This high resistance to ce xime contradicts a previous study [5], in which ce xime was reported as the most effective antimicrobial agent against UTI pathogens in pregnant women. However, their most predominant isolate was E. coli, a gram negative bacilli, unlike in this study, where gram positive organisms were more prevalent. This may indicate that ce xime could be more effective against gram negative bacteria than gram positive, though the few gram negative isolates in this study were still resistant to ce xime. N. gonorrhoeae has progressively developed resistance to antimicrobials used for the treatment of gonorrhea. Declining susceptibility to ce xime resulted in a change to the CDC treatment guidelines in 2012, to implement dual therapy with ceftriaxone and azithromycin used for treatment of gonorrhea [37]. In our study, N. gonorrhoeae was completely resistant to ce xime, ampicillin, ucloxacillin, o oxacin and gentamicin. In another study [38] in Vientiane-Laos, this isolate was reported to be highly resistant (84.8 %) to cipro oxacin.

Conclusions
The prevalence of asymptomatic UTI was 30.7% and 29.3% in pregnant and non-pregnant women, respectively, with gram positive bacteria as the most prevalent bacteria identi ed. Neisseria gonorrhoeae, Citrobacter diversus and Shigella species were most susceptible to cipro oxacin (100%), while Enterococci, Lactobacillus and Citrobacter diversus were most susceptible to ceftriaxone (100% All methods were carried out in accordance with relevant local and national regulations. An administrative authorization was obtained from the South West Regional Delegation of Public Health; ethical approval obtained from the Institutional Review Board (IRB), Faculty of Heath Sciences, University of Buea with reference number 2020/1038-12/UB/SG/IRB/FHS. A verbal authorization was obtained from the different participating health facilities. An informed consent was signed by all subjects and an informed consent by parent or guardian of subjects less than 21 years.

Consent for publication
Not applicable Availability of data and material All data generated and analysed during this study are included in this article Competing interest