Comparative analysis of generalized least squares and generalized inverse regression models for predicting neonatal birth weight using maternal anthropometric measures

https://doi.org/10.51867/scimundi.5.2.29

Authors

  • Stephen Waswa Department of Mathematics, Masinde Muliro University of Science and Technology, P. O. Box 190-50100, Kakamega, Kenya
  • Kennedy Nyongesa Department of Mathematics, Masinde Muliro University of Science and Technology, P. O. Box 190-50100, Kakamega, Kenya
  • Colleta Akinyi Department of Mathematics, Masinde Muliro University of Science and Technology, P. O. Box 190-50100, Kakamega, Kenya
  • Michael Onyango Ojiema Department of Mathematics, Masinde Muliro University of Science and Technology, P. O. Box 190-50100, Kakamega, Kenya https://orcid.org/0000-0001-9635-7597
  • Frankline Tireito Department of Mathematics, Masinde Muliro University of Science and Technology, P. O. Box 190-50100, Kakamega, Kenya https://orcid.org/0000-0002-4106-4022

Keywords:

Generalized Inverse Regression, Generalized Least Squares, Maternal Anthropometry, Measurement Error, Neonatal Birth Weight, Predictive Modeling

Abstract

This research presents a comparative analysis of two advanced statistical methodologies for predicting neonatal birth weight using maternal anthropometric measures. We developed and evaluated both Generalized Least Squares (GLS) and Generalized Inverse Regression (GIR) models to account for complex error structures and measurement uncertainties inherent in obstetric data. Data were collected from 150 mothers delivering full-term, singleton infants at a regional hospital, recording maternal weight, abdominal circumference, and neonatal birth weight. The GLS approach addressed correlated errors through covariance matrix transformation, while the GIR model incorporated measurement error adjustments using Stein estimation techniques. Both models demonstrated strong predictive capabilities, with the GLS model achieving slightly better accuracy (R² = 0.78, MAE = 0.15 kg) compared to the GIR model (R² = 0.75, MAE = 0.18 kg). However, the GIR model showed superior robustness in handling measurement errors. The study concludes that both methodologies offer valuable approaches for birth weight prediction, with GLS preferred for optimal accuracy and GIR preferred for enhanced robustness in settings with significant measurement uncertainties.

Downloads

Download data is not yet available.
Dimensions

Abrams, B. (1995). Factors associated with pattern of maternal weight gain. Journal of Obstetric Gynecology, 86, 170-176.

https://doi.org/10.1016/0029-7844(95)00119-C DOI: https://doi.org/10.1016/0029-7844(95)00119-C

Bajracharya, J., Shreshtha, N. S., & Karki, C. (2012). Accuracy of prediction of birth weight by fetal ultrasound. Kathmandu University Medical Journal, 10(38), 74-76.

https://doi.org/10.3126/kumj.v10i2.7349 DOI: https://doi.org/10.3126/kumj.v10i2.7349

Chang, T. C., & Robson, S. C. (1992). Prediction of the small for gestational age infant. Journal of Obstetric Gynecology, 80, 1030-1037.

Ekele, B. O., & Otubu, A. M. (2006). Maternal and perinatal mortality. In Textbook of Obstetrics and Gynaecology for Medical Students (Vol. 2, pp. 526-531). Ibadan, Nigeria.

Fawzia, A. (2002). Prediction of low birth weight infants from ultrasound measurements of placental diameter and thickness. King Khalid University Journal of Obstetrics and Gynecology, 5, 312-314. https://doi.org/10.5144/0256-4947.2002.312 DOI: https://doi.org/10.5144/0256-4947.2002.312

Fredrick, I. O., & William, M. A. (2008). Pre-pregnancy body mass index, gestational weight gain, and other maternal characteristics in relation to infant weight. Journal of Maternal and Child Health, 12, 557-567.

https://doi.org/10.1007/s10995-007-0276-2 DOI: https://doi.org/10.1007/s10995-007-0276-2

Godfrey, K. M. (2000). Fetal nutrition and adult disease. Journal of Clinical Nutrition, 7, 1344-1352.

https://doi.org/10.1093/ajcn/71.5.1344s DOI: https://doi.org/10.1093/ajcn/71.5.1344s

Himes, J. H. (1985). Parent-specific adjustments for evaluation of recumbent length and stature in children. Pediatrics, 75, 304-313. https://doi.org/10.1542/peds.75.2.304 DOI: https://doi.org/10.1542/peds.75.2.304

Kinira, S. B., & Davey, S. G. (2005). Early growth and childhood obesity: A historical cohort study. Archives of Disease in Childhood, 90, 1122-1127.

https://doi.org/10.1136/adc.2004.066712 DOI: https://doi.org/10.1136/adc.2004.066712

Kruciko, R. G. (1967). Classical and inverse regression method of calibration. Technometrics, 9(3), 213-218.

https://doi.org/10.2307/1266511 DOI: https://doi.org/10.2307/1266511

Martin, R. M., Davey, S. G., Frankel, S., & Gunnell, D. (2004). Parents' growth in childhood and the birth weight of their offspring. Journal of Epidemiology, 13, 308-316.

https://doi.org/10.1097/01.ede.0000120042.16363.e3 DOI: https://doi.org/10.1097/01.ede.0000120042.16363.e3

Nasim, M. (2012a). Parents' growth in childhood and the birth weight of their offspring. Journal of Epidemiology, 13, 308-316.

Nasim, M. (2012b). Adjusting errors in recumbent length at birth by inverse regression. International Journal of Statistical Sciences, 12, 71-78.

Parsons, T. J., Power, C., & Summerbell, C. D. (1999). Childhood predictors of adult obesity: A systematic review. International Journal of Obesity and Related Metabolic Disorders, 23(8), S1-S107.

Pervin, K. (2004). Adjusting recumbent length at birth. Unpublished M. Phil seminar, Department of Statistics, University of Rajshahi.

Rao, P. (2001). Influence of pre-pregnancy weight, maternal height and weight gain during pregnancy. Bahrain Medical Journal, 23, 22-26.

Rouche, B. (1989). Weight and recumbent length from 1-12 months of age. American Journal of Clinical Nutrition, 49, 599-607. https://doi.org/10.1093/ajcn/49.4.599 DOI: https://doi.org/10.1093/ajcn/49.4.599

Stevens, C. (1998). Adolescent maternal weight gain and low birth weight. American Journal of Clinical Nutrition, 47, 948-953.

https://doi.org/10.1093/ajcn/47.6.948 DOI: https://doi.org/10.1093/ajcn/47.6.948

Voorhorst, F. J. (1993). Maternal characteristics and expected birth weight. European Journal of Obstetrics & Gynecology, 50, 115-122. https://doi.org/10.1016/0028-2243(93)90175-C DOI: https://doi.org/10.1016/0028-2243(93)90175-C

Whitemore, A. S. (1989). Errors-in-variables regression using Stein estimates. The American Statistician, 43(4), 226-228.

https://doi.org/10.1080/00031305.1989.10475663 DOI: https://doi.org/10.1080/00031305.1989.10475663

Published

2025-11-24

How to Cite

Waswa, S., Nyongesa, K., Akinyi, C., Ojiema, M. O., & Tireito, F. (2025). Comparative analysis of generalized least squares and generalized inverse regression models for predicting neonatal birth weight using maternal anthropometric measures. SCIENCE MUNDI, 5(2), 304–316. https://doi.org/10.51867/scimundi.5.2.29

Issue

Vol. 5 No. 2 (2025): Jul-Dec 2025

Section

Articles
Copyright & Licensing

Copyright (c) 2025 Stephen Waswa, Kennedy Nyongesa, Colleta Akinyi, Michael Onyango Ojiema, Frankline Tireito

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Crossref
Scopus
Google Scholar
Europe PMC

Most read articles by the same author(s)