Ontology highlight
ABSTRACT: Objective
To derive dietary patterns (DP) among women of reproductive age (WRA) and older women in urban Sub-Saharan Africa (SSA), and to examine their association with body mass index (BMI), overweight and obesity.Design
We used principal component analysis to derive DP. Their association with BMI, overweight and obesity was assessed using linear and multinomial logistic regression models controlling for age, marital status, education and wealth.Setting
Cross-sectional data from prospective studies in Accra, Ghana (2008-2009), Dar es Salaam, Tanzania (2014) and Lilongwe and Kasungu, Malawi (2017-2018) were used.Participants
We compared WRA in Ghana (18-54 years, n 1762) and Malawi (19-48 years, n 137), and older women in Ghana (≥55 years, n 514) and Tanzania (≥50 years, n 134).Results
Among WRA, protein and healthy DP were identified in both Ghana and Malawi. In Ghana, the protein DP was associated with higher odds of overweight or obesity (adjusted OR 1·82, 95 % CI 1·27, 2·60 for quintile 2). Among older women, three DP were identified in Ghana (cereal, protein and healthy) and two DP in Tanzania (protein and healthy). The protein DP was associated with higher BMI in Ghana (adjusted mean difference 2·83, 95 % CI 0·95, 4·71 for quartile 3).Conclusions
Higher quintiles of the protein DP were associated with higher BMI and odds of overweight or obesity among women in urban Ghana, but not in Malawi or Tanzania. Further research is needed to understand how DP influence overweight and obesity among adult women in urban SSA.
SUBMITTER: Bliznashka L
PROVIDER: S-EPMC10195588 | biostudies-literature | 2021 Apr
REPOSITORIES: biostudies-literature
Bliznashka Lilia L Danaei Goodarz G Fink Günther G Flax Valerie L VL Thakwalakwa Chrissie C Jaacks Lindsay M LM
Public health nutrition 20200511 6
<h4>Objective</h4>To derive dietary patterns (DP) among women of reproductive age (WRA) and older women in urban Sub-Saharan Africa (SSA), and to examine their association with body mass index (BMI), overweight and obesity.<h4>Design</h4>We used principal component analysis to derive DP. Their association with BMI, overweight and obesity was assessed using linear and multinomial logistic regression models controlling for age, marital status, education and wealth.<h4>Setting</h4>Cross-sectional d ...[more]