Jomo Kenyatta University of Agriculture and Technology

Summary Sorghum is a food staple for millions of people in sub-Saharan Africa, but its production is greatly diminished by Striga, a parasitic weed. An efficient and cost-effective way of managing Striga in smallholder farms in Africa is to deploy resistant varieties of sorghum. Here, we leverage genomics and the vast genetic diversity of sorghum—evolutionarily adapted to cope with Striga parasitism in Africa—to identify new Striga-resistant sorghum genotypes by exploiting a resistance mechanism hinged on communication molecules called strigolactones (SLs), exuded by hosts to trigger parasite seed germination. We achieved this by mining for mutant alleles of the LOW GERMINATION STIMULANT 1 (LGS1) that are ineffective in stimulating Striga germination from the sorghum accession panel (SAP). Our analysis identified lgs1 sorghum genotypes, which we named SAP-lgs1. SAP-lgs1 had the SL exudation profile of known lgs1 sorghum, whose hallmark is the production of the low inducer of germination, orobanchol. Laboratory and field resistance screens showed that the SAP-lgs1 genotypes also exhibited remarkable resistance against Striga. Our findings have the potential to reduce crop losses because of Striga parasitism and therefore have far-reaching implications for improving food security in Africa.

Transfer learning involves using previously learnt knowledge of a model task in addressing another task. However, this process works well when the tasks are closely related. It is, therefore, important to select data points that are closely relevant to the previous task and fine-tune the suitable pre-trained model’s layers for effective transfer. This work utilises the least divergent textural features of the target datasets and pre-trained model’s layers, minimising the lost knowledge during the transfer learning process. This study extends previous works on selecting data points with good textural features and dynamically selected layers using divergence measures by combining them into one model pipeline. Five pre-trained models are used: ResNet50, DenseNet169, InceptionV3, VGG16 and MobileNetV2 on nine datasets: CIFAR-10, CIFAR-100, MNIST, Fashion-MNIST, Stanford Dogs, Caltech 256, ISIC 2016, ChestX-ray8 and MIT Indoor Scenes. Experimental results show that data points with lower textural feature divergence and layers with more positive weights give better accuracy than other data points and layers. The data points with lower divergence give an average improvement of 3.54% to 6.75%, while the layers improve by 2.42% to 13.04% for the CIFAR-100 dataset. Combining the two methods gives an extra accuracy improvement of 1.56%. This combined approach shows that data points with lower divergence from the source dataset samples can lead to a better adaptation for the target task. The results also demonstrate that selecting layers with more positive weights reduces instances of trial and error in selecting fine-tuning layers for pre-trained models.

Weighted Cox regression models were proposed as an alternative to the standard Cox proportional hazards models where consistent estimators can be obtained with more relative strength compared to unweighted cases. We proposed censoring balancing functions which can be built in a way that allows us to obtain the maximum possible significant treatment effects that may have gone undetected due to censoring. The harm caused by this is compensated and new weighted parameter estimates are found. These functions are constructed to be monotonic because even the hazard ratios are not exactly constant as in the ideal case, but are violated by monotonic deviations in time. For more than one covariate, even the interaction between covariates in addition to censoring can lead to the loss of significance for some covariates’ effects. Undetected significant effects of one covariate can be obtained, still keeping the significance and approximate size of the remaining one(s). This is performed by keeping the consistency of the parameter estimates. The results from both the simulated datasets and their application to real datasets supported the importance of the suggested censoring balancing functions in both one covariate and more than one covariate cases.

Background Kenyan adolescent girls and young women (AGYW) experience a dual burden of HIV and common mental disorders (CMD). HIV clinics are a key entry point for AGYW in need of integrated CMD and HIV care; however, rates of screening and referral for CMDs are low. Our objective was to test an evidence-based provider training strategy, simulated patient encounters (SPEs), on CMD service delivery for AGYW in a Kenyan HIV clinic.Methods This pilot study was conducted in a public HIV clinic in Thika, Kenya from January to November 2021. The simulated patient encounter (SPE) implementation strategy included case script development from prior qualitative work, patient actor training, and a three-day SPE training including four standardized mock clinical encounters followed by quantitative surveys assessing provider competencies for each encounter. We abstracted medical record data related to HIV and CMDs such as HIV status, reason for visit, CMD screening test performed, and counselling or referral information. We conducted an interrupted time series analysis using abstracted HIV and CMD screening rates from AGYW ages 16–25 years visiting the clinic 7 months before and 3 months after SPE training. We used generalized linear models to assess changes in CMD screening rates after training.ResultsA total of 10 providers participated in the training. Competency ratings improved across four mock encounters (mean score from 8.1 to 13.7) between first and fourth encounters. We abstracted all medical records (n = 1,154) including from 888 (76%) AGYW seeking HIV treatment, 243 (21%) seeking prevention services, and 34 (3%) seeking other services. CMD screening rates increased immediately following training from 8 to 21% [relative risk (RR) = 2.57, 95% confidence interval (CI) = 1.34–4.90, p < 0.01]. The 3 months following the SPE training resulted in an 11% relative increase in CMD screening proportion compared to the 7 months pre-SPE (RR: 1.11, 95% CI: 1.04–1.17, p < 0.01). Finally, 1% of all pre-SPE screens resulted in referral versus 5% of post-SPE screens (p = 0.07).Conclusion The SPE model is a promising implementation strategy for improving HIV provider competencies and CMD service delivery for adolescents in HIV clinics. Future research is needed to explore effects on adolescent clinical outcomes in larger trials.

Given the phenomenal growth of urban centres in developing countries, cities have become important sites for engaging with environmental issues. Among the many issues, the prevalence of informal settlements in cities and their implications on environmental sustainability have emerged as a growing concern. However, the significance of these spaces in the environmental sustainability function has not been adequately considered. Policy makers and urban planners have often failed to recognize informal settlement upgrading processes as one of the most important avenues for realizing environmentally sustainable cities. As a result, these projects have not only failed to improve the inherent poor environmental quality, but they have also aggravated consumption inefficiencies, waste management problems and disrupted the social structures through which environmental management can be propagated. This paper evaluates the environmental sustainability performance of government led informal settlement upgrading projects in Kenyan cities in an effort to determine the extent to which environmental sustainability outcomes are evident in upgraded settlements. The evaluation focuses on environmental performance of 11 upgraded informal settlements in three cities on resource consumption domain based on 5 indicators using comparative performance rating. Data was collected through multistage cluster and simple random sampling of 600 household; coupled with key informant interviews, observations and focus group discussions. The results indicate low resource consumption environmental sustainability performance in the upgraded settlements in Kenya. There is need for deliberate action to integrate environmental sustainability in informal settlement upgrading processes.

Micronutrient deficiencies, particularly of iron (Fe) and zinc (Zn), in the diet contribute to health issues and hidden hunger. Enhancing the Fe and Zn content in globally staple food crops like rice is necessary to address food malnutrition. A Genome-Wide Association Study (GWAS) was conducted using 85 diverse rice accessions from the Democratic Republic of Congo (DRC) to identify genomic regions associated with grain Fe and Zn content. The Fe content ranged from 0.95 to 8.68 mg/100 g on a dry weight basis (dwb) while Zn content ranged from 0.87 to 3.8 mg/100 g (dwb). Using MLM and FarmCPU models, we found 10 significant SNPs out of which one SNP on chromosome 11 was associated with a variation in Fe content and one SNP on chromosome 4 was associated with the Zn content, and both were commonly detected by the two models. Candidate genes belonging to transcription regulator activities, including the bZIP family genes and MYB family genes, as well as transporter activities involved in Fe and Zn homeostasis were identified in the vicinity of the SNP markers and selected. The identified SNP markers hold promise for marker-assisted selection in rice breeding programs aimed at enhancing Fe and Zn content in rice. This study provides valuable insights into the genetic factors controlling Fe and Zn uptake and their transport and accumulation in rice, offering opportunities for developing biofortified rice varieties to combat malnutrition among rice consumers.

Periodontal diseases are prevalent and have significant implications for oral health and overall well-being. Current diagnostic methods have limitations in accuracy and standardization. The recent Food and Drug Administration approval of Videa Perio Assist (VPA), an AI model for diagnosing periodontal diseases, presents a breakthrough in dental diagnostics. VPA is a cloud-based, AI-powered software that automatically measures and visualizes bone levels associated with each tooth from radiographic images. Clinical testing has demonstrated VPA's efficacy in accurately diagnosing periodontal diseases with high sensitivity and specificity. The integration of AI in dentistry has the potential to revolutionize periodontal disease diagnosis, improve patient care, and enhance decision-making. However, further research, education, cost-effectiveness, and collaboration are essential for maximizing the benefits of AI in dental settings. The approval and implementation of VPA mark a significant advancement in dental diagnostics, paving the way for more effective solutions and a healthier global population.

The wild African harlequin quails (Coturnix delegorguei delegorguei) of Western Kenya suffer from incessant hunting, habitat fragmentation, and the effects of climate change. These challenges, among others, have forced them to breed under intensive pressure, disrupting normal evolutionary processes. Here, we provide the first overview of the selection signatures in wild African harlequin quails using genotyping-by-sequencing information from 76 captured individuals. Additionally, 19 domestic Japanese quails (Coturnix coturnix japonica) were incorporated for comparative signatures of selection analysis between wild and domesticated quail species that undergo different selection pressures. Composite likelihood ratio test (CLR) and integrated haplotype score (iHS) methods were used to detect selection signatures. As a result, 252 and 424 candidate genes were detected in wild African harlequin and domestic Japanese quails, respectively, through the CLR test, whereas 150 and 457 candidate genes were identified through iHS. Some of the essential candidate genes identified in the wild African harlequin quail were associated with important traits such as immune response (MAPK13, MAPK14, CREB1, ITGB3, and PPP1CA) and morphological traits (WNT5A, GRIA1, CREB1, ADCY8, and ALK) whereas, in domestic Japanese quail, primarily production-related genes such as VIPR2, DYNLL2, COL6A3, MSX2, PRF1and GNA12 were identified. Our findings provide insights into the role of selection in shaping both wild and domestic quail genomes in terms of significant immune response, growth, reproduction, and morphological and behavioral traits.

Background: Multiplicity of infection (MOI) is an important measure of Plasmodium falciparum diversity, usually derived from the highly polymorphic genes, such as msp1, msp2 and glurp as well as microsatellites. Conventional methods of deriving MOI lack fine resolution needed to discriminate minor clones. This study used amplicon sequencing (AmpliSeq) of P. falciparum msp1 (Pfmsp1) to measure spatial and temporal genetic diversity of P. falciparum. Methods: 264 P. falciparum positive blood samples collected from areas of differing malaria endemicities between 2010 and 2019 were used. Pfmsp1 gene was amplified and amplicon libraries sequenced on Illumina MiSeq. Sequences were aligned against a reference sequence (NC_004330.2) and clustered to detect fragment length polymorphism and amino acid variations. Results: Children < 5 years had higher parasitaemia (median = 23.5 ± 5 SD, p = 0.03) than the > 5-14 (= 25.3 ± 5 SD), and those > 15 (= 25.1 ± 6 SD). Of the alleles detected, 553 (54.5%) were K1, 250 (24.7%) MAD20 and 211 (20.8%) RO33 that grouped into 19 K1 allelic families (108-270 bp), 14 MAD20 (108-216 bp) and one RO33 (153 bp). AmpliSeq revealed nucleotide polymorphisms in alleles that had similar sizes, thus increasing the K1 to 104, 58 for MAD20 and 14 for RO33. By AmpliSeq, the mean MOI was 4.8 (± 0.78, 95% CI) for the malaria endemic Lake Victoria region, 4.4 (± 1.03, 95% CI) for the epidemic prone Kisii Highland and 3.4 (± 0.62, 95% CI) for the seasonal malaria Semi-Arid region. MOI decreased with age: 4.5 (± 0.76, 95% CI) for children < 5 years, compared to 3.9 (± 0.70, 95% CI) for ages 5 to 14 and 2.7 (± 0.90, 95% CI) for those > 15. Females' MOI (4.2 ± 0.66, 95% CI) was not different from males 4.0 (± 0.61, 95% CI). In all regions, the number of alleles were high in the 2014-2015 period, more so in the Lake Victoria and the seasonal transmission arid regions. Conclusion: These findings highlight the added advantages of AmpliSeq in haplotype discrimination and the associated improvement in unravelling complexity of P. falciparum population structure.

Kenya has registered over 300,000 cases of COVID-19 and is a high-burden tuberculosis country. Tuberculosis diagnosis was significantly disrupted by the pandemic. Access to timely diagnosis, which is key to effective management of tuberculosis and COVID-19, can be expanded and made more efficient through integrated screening. Decentralized testing at community level further increases access, especially for underserved populations, and requires robust systems for data and process management. This study delivered integrated COVID-19 and tuberculosis testing to commercial motorbike (Bodaboda) riders, a population at increased risk of both diseases with limited access to services, in four counties: Nairobi, Kiambu, Machakos and Kajiado. Testing sheds were established where riders congregate, with demand creation carried out by the Bodaboda association. Integrated symptom screening for tuberculosis and COVID-19 was conducted through a digital questionnaire which automatically flagged participants who should be tested for either, or both, diseases. Rapid antigen-detecting tests (Ag-RDTs) for COVID-19 were conducted onsite, while sputum samples were collected and transported to laboratories for tuberculosis diagnosis. End-to-end patient data were captured using digital tools. 5663 participants enrolled in the study, 4946 of whom were tested for COVID-19. Ag-RDT positivity rate was 1% but fluctuated widely across counties in line with broader regional trends. Among a subset tested by PCR, positivity was greater in individuals flagged as high risk by the digital tool (8% compared with 4% overall). Of 355 participants tested for tuberculosis, 7 were positive, with the resulting prevalence rate higher than the national average. Over 40% of riders had elevated blood pressure or abnormal sugar levels. The digital tool successfully captured complete end-to-end data for 95% of all participants. This study revealed high rates of undetected disease among Bodaboda riders and demonstrated that integrated diagnosis can be delivered effectively in communities, with the support of digital tools, to maximize access.

The large scale and dynamic nature of cloud has added extra complexity when it comes to fault detection and management. Availability directly depends upon how fast the cloud infrastructure can detect any faults and take necessary steps to troubleshoot the problem. It is critical for service providers to provide stable service or else it may cause losses for clients. It is important to detect a failure in its embryonic stages to employ preventive measures to avoid a disastrous failure before it occurs. Researchers have focused on pure-bred failure characterization and analysis machine learning models to enhance cloud failure prediction accuracy for large cloud data centers but limited research has been done with ensemble models. In this paper we develop an enhanced cloud failure prediction model based on Adaboost ensemble Machine Learning algorithms that predicts hardware and software failures using Google Cluster 2019, Azure Clouds and Alibaba clouds datasets. Our model employs ensemble classification using Logistic Regression, Random Forest Classifier and Decision Tree Classifier. Results indicate our approach recorded a marginal improvement in accuracy prediction compared with results from previous researchers in the area. Decision Tree yielded best average model performance results recording 91.7% Precision, 88.8% Recall, 89.7% F1 Score, 94.0% Accuracy and 89.0% AUC.

Objective: Multipurpose prevention technologies (MPTs) are developmental dual-purpose options that would provide women with a contraceptive as well as a prevention modality aimed at sexually transmitted infections. The contraceptive vaginal ring (CVR) has many properties that makes it an ideal MPT candidate. The objective of this study is to understand women’s attitudes towards menstrual suppression, a potential side effect of using a CVR, and how to address these attitudes for MPT vaginal rings in development. Materials and methods: We analyzed data derived from a subset of cohort study participants (n=45) in Thika, Kenya between January 2016- December 2018. The primary study enrolled 121 women 18-40 years with bacterial vaginosis and randomized them to cyclic or continuous CVR use for eight months. During the 6-month follow-up, a questionnaire eliciting attitudes towards menstrual suppression was administered. Responses to the menstrual suppression questionnaire between participants in the cyclic and continuous CVR use groups were compared. Likert-scale responses were summed to create a menstrual suppression attitude summary score, and a hierarchical cluster analysis was conducted to identify similarities in response patterns among all participants. Results: Totally 81.8% of continuous CVR users believed that one was less likely to get pregnant after using hormonal medication to suppress menses, compared to 47.8% of cyclic CVR users (P=0.02), and were more worried it would cause long-term health effects (86.4% vs 60.9%, p = 0.05). The menstrual suppression attitude summary score ranged from 8 to 42, with lower scores indicating negative attitudes. The summary score identified three distinct clusters. When asked if menstrual suppression effects long-term health; 100% of Cluster 3 was worried compared to 80.8% of Cluster 2 and 46.2% of Cluster 1 (p = 0.03). The average summary score among Cluster 3 (Mean = 14.8, SD = 4.6) was lower (p < 0.001) and women were more worried about discomfort during sex (p=0.05) as well as their sexual partners feeling the ring (p=0.02). Conclusion: Women are more likely to have negative attitudes if they believe menstrual suppression hinders future reproductive health. Education on cycle control and fertility could mitigate negative attitudes and improve uptake of CVRs.

The mango cultivar 'Apple' is an important fruitcrop in Kenya, but it is highly susceptible to russeting. The objective was to establish whether lenticels predispose cv. 'Apple' mango to russeting. Fruit mass and surface area increased in a sigmoidal pattern with time. The frequency of lenticels per unit surface area decreased during development. The number of lenticels per fruit was constant. Lenticels were most frequent in the apex region and least common in the cheek and nak (ventral) regions. The cheek region also had lenticels with the largest core areas, whereas the lenticel core areas in the apex region were significantly smaller. Microscopy revealed stomata became covered over with wax deposits at 33 days after full bloom (DAFB). By 78 DAFB, periderm had formed beneath the pore. At 110 and 161 DAFB, cracks had developed and the periderm had extended tangentially and radially. The presence of lenticels increased the strain released upon excision of an epidermal segment, further strain releases occurred subsequently upon isolation of the cuticle and on extraction of the cuticular waxes. The number of lenticels per unit surface area was negatively correlated with the fruit surface area (r2 = 0.62 **), but not affected by fruit size. Mango cv. 'Apple' had fewer, larger lenticels and more russet, compared with 'Ngowe', 'Kitovu' or 'Tommy Atkins' mango. In cv. 'Apple', the lowest lenticel frequency, the largest lenticels and the most russeting occurred at a growing site at the highest altitude, with the highest rainfall and the lowest temperature. Moisture exposure of the fruit surface resulted in enlarged lenticels and more microcracking of the cuticle. Our results establish that russeting in 'Apple' mango is initiated at lenticels and is exacerbated if lenticels are exposed to moisture.

Genomic surveillance is vital for detecting outbreaks and understanding the epidemiology and transmission of bacterial strains, yet it is not integrated into many national antimicrobial resistance (AMR) surveillance programmes. Key factors are that few scientists in the public health sector are trained in bacterial genomics, and the diverse sequencing platforms and bioinformatic tools make it challenging to generate reproducible outputs. In Kenya, these gaps were addressed by training public health scientists to conduct genomic surveillance on isolates from the national AMR surveillance repository and produce harmonized reports. The 2-week training combined theory and laboratory and bioinformatic experiences with Klebsiella pneumoniae isolates from the surveillance repository. Whole-genome sequences generated on Illumina and Nanopore sequencers were analysed using publicly available bioinformatic tools, and a harmonized report was generated using the HAMRonization tool. Pre- and post-training tests and self-assessments were used to assess the effectiveness of the training. Thirteen scientists were trained and generated data on the K. pneumoniae isolates, summarizing the AMR genes present consistently with the reported phenotypes and identifying the plasmid replicons that could transmit antibiotic resistance. Ninety per cent of the participants demonstrated an overall improvement in their post-training test scores, with an average increase of 14 %. Critical challenges were experienced in delayed delivery of equipment and supplies, power fluctuations and internet connections that were inadequate for bioinformatic analysis. Despite this, the training built the knowledge and skills to implement bacterial genomic surveillance. More advanced and immersive training experiences and building supporting infrastructure would solidify these gains to produce tangible public health outcomes.

In recent ages, green nanotechnology has gained attraction in the synthesis of metallic nanoparticles due to their cost-effectiveness, simple preparation steps, and environmentally-friendly. In the present study, copper oxide nanoparticles (CuO NPs) were prepared using Parthenium hysterophorus whole plant aqueous extract as a reducing, stabilizing, and capping agent. The CuO NPs were characterized via UV–Vis Spectroscopy, Fourier Transform Infrared Spectroscopy (FTIR), powder X-Ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), and Dynamic Light Scattering (DLS). The UV–Vis spectra of CuO NPs showed a surface plasmonic resonance band to occur at 340 nm. FTIR analysis revealed the presence of secondary metabolites on the surface of CuO NPs, with a characteristic Cu–O stretching band being identified at 522 cm ⁻¹ . Scanning electron micrographs and transmission electron micrographs showed that CuO NPs were nearly spherical, with an average particle of 59.99 nm obtained from the SEM micrograph. The monoclinic crystalline structure of CuO NPs was confirmed using XRD, and crystallite size calculated using the Scherrer-Debye equation was found to be 31.58 nm. DLS showed the presence of nanoparticle agglomeration, which revealed uniformity of the CuO NPs. Furthermore, the degradation ability of biosynthesized nanoparticles was investigated against rifampicin antibiotic. The results showed that the optimum degradation efficiency of rifampicin at 98.43% was obtained at 65℃ temperature, 50 mg dosage of CuO NPs, 10 mg/L concentration of rifampicin solution, and rifampicin solution at pH 2 in 8 min. From this study, it can be concluded that CuO NPs synthesized from Parthenium hysterophorus aqueous extract are promising in the remediation of environmental pollution from antibiotics. In this light, the study reports that Parthenium hysterophorus -mediated green synthesis of CuO NPs can effectively address environmental pollution in cost-effective, eco-friendly, and sustainable ways.

Background Propolis is one of the hive products made by bees from wax and plant resins, which makes it rich in phytochemicals such as flavonoids and phenolic. The chemical composition of propolis from stingless bees is affected by several factors including vegetation type and species. In Kenya, we have diverse landscape vegetation and over 12 stingless bee species. Lifestyle diseases are affecting more people due to unhealthy diets and exposure to harmful chemicals, which lead to oxidation in cells. This led to the production of super radicals, which cannot be neutralized only by the endogenous antioxidants in the body. This has led to increased demand for exogenous antioxidants obtained from the diet. Propolis is a known source of natural antioxidants too due to the presence of polyphenols. Results In our study, we determined the total content of flavonoids, phenolic, and the antioxidant activity of propolis from six stingless bee species from Kenya. The total flavonoid content (TFC) in all species ranged from 651.90 to 3262.26 mg QE/100 g, while the total phenolic content (TPC) ranged from 586.36 to 2010.53 mg GAE/100 g. Among the species, Meliponula beccarii had the highest concentration of total flavonoids, Meliponula togoensis had high concentrations of total phenolic, while Meliponula ferruginea had the highest antioxidant activity. Conclusion The outcome of this study demonstrates that propolis from Kenyan stingless bees have varying amounts of phytochemicals, which is dependent on the species identity. Hence it is a good source of exogenous natural antioxidants.

Air quality is an important part of environmental health, having serious consequences for human health and well-being. The Air Quality Index (AQI) is a frequently used metric for assessing air quality in various areas and at different times. However, AQI data, like many other types of environmental data, can contain outliers - data points that deviate significantly from other observations, indicating exceptionally good or poor air quality, a critical step in identifying and understanding extreme pollution episodes that can have serious environmental and public health consequences. These outliers can be caused by a variety of variables, including measurement mistakes, odd meteorological circumstances, and pollution occurrences. While outliers can occasionally give useful information about these unusual conditions, they can also skew studies and models if they are not adequately accounted for. This paper describes a hybrid method for detecting outliers in data, AQI data are used in this study. The model uses a stacked machine learning model that incorporates K-means clustering, Random Forest (RF), and Gradient Boosting Classifier (GBC). K-means is used for initial categorization, followed by RF model training, and ultimately, the RF output is used as input for the GBC to generate the final classification. The performance of this stacked machine learning model is examined and compared to single models using the Accuracy measure. The findings show that the suggested technique is efficient, with an accuracy of 0.99, showing its potential for effective outlier detection in data.

Background: Menstruation is a normal biological process experienced by more than 300 million women globally daily. Women need clean menstrual absorbents that can be changed as often as needed in private and safe place with proper hygiene and disposal facilities. All these needs must be met throughout the duration of the menstrual cycle. Access to menstrual needs of women is important for their health, wellbeing, and human dignity. This study assessed the prevalence and factors associated with unmet need for menstrual hygiene management (MHM) in Ethiopia, Kenya, Uganda, Burkina Faso, Ghana, and Niger. Methods: We used data from the performance monitoring for action (PMA) surveys. We defined the unmet need for MHM as “lack of resources, facilities and supplies for MHM.” Sample characteristics were summarised using frequencies and percentages while prevalence was summarised using proportions and their respective confidence intervals (CI). Factors associated with unmet need for MHM were assessed using a multilevel logistic regression model. Results: In the six countries, majority of women were aged 20-34 years, were married, or cohabiting and had never given birth. The prevalence of unmet need for MHM was high among the uneducated and multiparous women, those who reused MHM materials, practiced open defaecation and lived in rural areas in all the six countries. The prevalence of unmet need for MHM was highest in Burkina Faso (74.8%) and lowest in Ghana (34.2). Age, education level, wealth status and marital status were significantly associated with unmet need for MHM. Reuse of MHM materials and open defaecation increased the odds of unmet need for MHM. Conclusion: More than half of women in five of the six countries have unmet need for MHM withodds of unmet need significantly higher among younger women, those with low wealth status, the unmarried, and those with poor access to sanitary facilities. This study highlights the state of period poverty in Sub-saharan Africa. Efforts to end period poverty should approach MHM needs as a unit as each need is insufficient on its own.