Dietary diversity of women from soybean and non-soybean farming households in rural Zambia

Ndashe Philemon Kapulu, Christian Chomba, Chewe Nkonde, Melvin J Holmes, Simon Manda, Harriet Smith, Jennie Macdiarmid, Caroline Orfila* (Corresponding Author)

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Downloads (Pure)


Introduction: Soybean farming in Zambia is promoted to increase farm productivity and diversification away from maize, and improve cash income and livelihoods for farmers. However, the impact of soybean farming on women's dietary intake is not clear. This study compares the dietary diversity of women from soybean (S) and non-soybean (NS) farming households as a pathway to understanding policy efficacy.

Methods: A cross-sectional survey involving 268 women of reproductive age from 401 rural households was conducted in two soybean-producing districts of Central Province, Zambia. Data from a qualitative 7-day food frequency questionnaire (FFQ) was used to calculate dietary diversity scores (DDS), women's dietary diversity scores (WDDS-10) and assess dietary patterns. Information on household sociodemographic and agricultural characteristics was used to explore determinants of dietary diversity.

Results: Results show there were no significant differences in the mean DDS (S: 10.3 ± 2.4; NS:10.3 ± 2.6) and WDDS-10 (S:6.27 ± 1.55; NS:6.27 ± 1.57) of women from soybean and non-soybean farming households. Both cohorts had similar dietary patterns, plant-based food groups with additional fats and oils. Agricultural diversity was not associated with dietary diversity. Household wealth status was the most important determinant of dietary diversity, as women from wealthier households were more likely to have higher DDS (β = 0.262, 95% CI = 0.26 to 0.70, P < 0.001) and WDDS-10 (β = 0.222, 95% CI = 0.08 to 0.37, P < 0.003) compared to those from poorer households. Women from households that spent more on food had a higher DDS (β = 0.182, 95% CI = 0.002 to 0.07), but not WDDS-10 (β = 0.120, 95% CI = −0.01 to 0.03); for every additional dollar spent on food in the past 7 days, the DDS increased by 0.18. Meanwhile, soyabean farming was not statistically associated with higher wealth.

Conclusions: Policymakers and promoters of agricultural diversification and nutrition-sensitive agriculture need to consider how women can benefit directly or indirectly from soybean farming or other interventions aimed at smallholder farmers.
Original languageEnglish
Article number1115801
Number of pages15
JournalFrontiers in Sustainable Food Systems
Publication statusPublished - 12 Apr 2023

Bibliographical note

This research was funded by the Biological Sciences Research Council (BBSRC) under the Global Challenges Research Fund (GCRF) project Agricultural and Food-system Resilience: Increasing Capacity and Advising Policy (GCRF-AFRICAP) grant number BB/P027784/1, and by Research England through a Ph.D. stipend to NK (project reference number 95522727).

The findings presented in this paper forms part of a Ph.D. thesis by NK at the University of Leeds [Kapulu, Ndashe Philemon (2022)] Soybean Agricultural Expansions and Implications for Food and Nutrition Security in Zambia. Ph.D. thesis, University of Leeds, accessible via The authors wish to thank Prof. Tim Benton and Dr. Stephen Whitfield for leading the GCRF-AFRICAP project and suggesting the general area of study; Bertha Munthali at FARNPAN for discussions on the dietary assessment; Dr. Njongile Nyoni at FARNPAN for assistance with data collection and the ODK tool; the enumerators who assisted in data collection and the participants for taking part in the study.

Data Availability Statement

The Supplementary Material for this article can be found online at:

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.


Dive into the research topics of 'Dietary diversity of women from soybean and non-soybean farming households in rural Zambia'. Together they form a unique fingerprint.

Cite this