AGE-STRATIFIED PREDICTION OF BREAST CANCER RISK IN NIGERIAN WOMEN: A MULTIVARIATE LOGISTIC REGRESSION ANALYSIS HIGHLIGHTING YOUNG ADULT VULNERABILITY

Read full text Summarise with VIRAT
Authors: Kennedy Irikefe Ekerikevwe, SONIA CHINOR CHIGOZIE
Cited by: 0
References: 5

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Breast cancer (BC) is a critical public health concern globally, particularly in Nigeria, where high mortality rates are inextricably linked to systemic factors, notably delayed diagnosis and a disproportionately high disease burden among Adolescent and Young Adult (AYA) females (aged 15–39 years). This study aimed to identify key demographic, reproductive, and lifestyle risk factors and develop an age-stratified predictive model using logistic regression to inform targeted early intervention strategies. A retrospective analysis of secondary data was performed using Electronic Health Records (EHRs) from a specialist hospital in Nigeria, utilizing a convenience sample of 60 female patients. The analysis examined the association between BC diagnosis (binary outcome) and nine predictor variables, including age group, family history, obesity, alcohol consumption, and age at first childbirth. Multivariate binary logistic regression was the primary analytical technique employed to model disease likelihood across distinct age cohorts. Results show that the overall model demonstrated a moderate level of explanatory power, achieving a McFadden Pseudo R2 of 0.363. The regression estimates exhibited significant statistical instability, characterized by extremely large standard errors for age group dummy variables, preventing reliable interpretation of individual Odds Ratios for most predictors, including marital status and smoking status. However, the derived predicted probabilities revealed a crucial finding: the 25–29 age group exhibited the highest predicted probability of BC diagnosis (0.60), confirming the substantial vulnerability of young Nigerian women. Alcohol consumption showed a borderline positive coefficient (beta = 2.7770, P = 0.079), suggesting a potentially influential, though non-significant, association, while age at first childbirth showed a negative relationship (beta = -0.2528, P = 0.121) that contradicted established epidemiological norms. The findings validate existing cancer registry data regarding the urgency of BC in young Nigerian women and provide empirical evidence highlighting the 25–29 age bracket as critically vulnerable. This mandates a fundamental reassessment of public health strategies, necessitating tailored screening and awareness campaigns that prioritize early detection and improved diagnostic access for AYA populations in Sub-Saharan Africa.

Fields of Study & Subjects

Cancer Risks and Factors Global Cancer Incidence and Screening Women's cancer prevention and management

Publication details

Journal / source
International Journal of Convergent and Informatics Science Research
Publication date
1 April 2026
Language
EN