Fahad Razak1, Daniel J Corsi2, Arthur S Slutsky3, Anura Kurpad4, Lisa Berkman5, Andreas Laupacis3, S V Subramanian5. 1. Li Ka Shing Knowledge Institute of St Michael's Hospital, Toronto, Ontario, Canada2Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada3Harvard Centre for Population and Development Studies, Harvard University, Cambridge, Massachusetts. 2. Harvard Centre for Population and Development Studies, Harvard University, Cambridge, Massachusetts4Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada. 3. Li Ka Shing Knowledge Institute of St Michael's Hospital, Toronto, Ontario, Canada2Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada. 4. Division of Nutrition, St John's Medical College, Bangalore, Karnataka, India. 5. Harvard Centre for Population and Development Studies, Harvard University, Cambridge, Massachusetts.
Abstract
IMPORTANCE: Body mass index (BMI) lower than 16 is the most severe category of adult undernutrition and is associated with substantial morbidity, increased mortality, and poor maternal-fetal outcomes such as low-birth-weight newborns. Little is known about the prevalence and distribution of BMI lower than 16 in low- and middle-income countries (LMIC). OBJECTIVE: To determine the prevalence and distribution of BMI lower than 16 and its change in prevalence over time in women in LMIC. DESIGN, SETTINGS, AND PARTICIPANTS: Cross-sectional data analysis composed of nationally representative surveys from 1993 through 2012 from the Demographic and Health Surveys Program. Women aged 20 through 49 years from 60 LMIC (N = 500,761) and a subset of 40 countries with repeated surveys (N = 604,144) were examined. EXPOSURES: Wealth was measured using a validated asset index, age was categorized in deciles, education by highest completed level (none, primary, secondary, or greater), and place of residence as urban vs rural. MAIN OUTCOMES AND MEASURES: The primary outcome was BMI lower than 16. Analyses assessed the prevalence of BMI lower than 16, its association with sociodemographic factors, and change in prevalence. Logistic regression models were used to calculate odds ratios (ORs), adjusting for survey design and age structure. RESULTS: Among countries examined, the pooled, weighted, and age-standardized prevalence of BMI lower than 16 was 1.8% (95% CI, 1.7% to 1.8%) with the highest prevalence in India (6.2% [95% CI, 5.9% to 6.5%]), followed by Bangladesh (3.9% [95% CI, 3.4% to 4.3%]), Madagascar (3.4% [95% CI, 2.8% to 4.0%], Timor-Leste (2.9% [95% CI, 2.4% to 3.2%]), Senegal (2.5% [95% CI, 1.9% to 3.2%]), and Sierra Leone (2.2% [95% CI, 1.3% to 3.0%]); and 6 countries had prevalences lower than 0.1% (Albania, Bolivia, Egypt, Peru, Swaziland, and Turkey). The prevalence of BMI lower than 16 in women with a secondary or higher education level was 0.51% (95% CI, 0.47% to 0.55%), and in mutually adjusted models, a less than primary education level was associated with an OR of 1.4 (95% CI, 1.2 to 1.6). The prevalence of BMI lower than 16 was 0.43% (95% CI, 0.37% to 0.48%) in the highest wealth quintile with an OR of 3.0 (95% CI, 2.4 to 3.7) in the lowest wealth quintile. Among the 24 of 39 countries with repeated surveys, there was no decrease in prevalence. In Bangladesh and India, rates were declining with an average absolute change annually of -0.52% (95% CI, -0.58% to -0.46%) in Bangladesh and -0.11% (95% CI, -0.12% to -0.10%) in India. CONCLUSIONS AND RELEVANCE: Among women in 60 LMIC, the prevalence of BMI lower than 16 was 1.8%, and was associated with poverty and low education levels. Prevalence of BMI lower than 16 did not decrease over time in most countries studied.
IMPORTANCE: Body mass index (BMI) lower than 16 is the most severe category of adult undernutrition and is associated with substantial morbidity, increased mortality, and poor maternal-fetal outcomes such as low-birth-weight newborns. Little is known about the prevalence and distribution of BMI lower than 16 in low- and middle-income countries (LMIC). OBJECTIVE: To determine the prevalence and distribution of BMI lower than 16 and its change in prevalence over time in women in LMIC. DESIGN, SETTINGS, AND PARTICIPANTS: Cross-sectional data analysis composed of nationally representative surveys from 1993 through 2012 from the Demographic and Health Surveys Program. Women aged 20 through 49 years from 60 LMIC (N = 500,761) and a subset of 40 countries with repeated surveys (N = 604,144) were examined. EXPOSURES: Wealth was measured using a validated asset index, age was categorized in deciles, education by highest completed level (none, primary, secondary, or greater), and place of residence as urban vs rural. MAIN OUTCOMES AND MEASURES: The primary outcome was BMI lower than 16. Analyses assessed the prevalence of BMI lower than 16, its association with sociodemographic factors, and change in prevalence. Logistic regression models were used to calculate odds ratios (ORs), adjusting for survey design and age structure. RESULTS: Among countries examined, the pooled, weighted, and age-standardized prevalence of BMI lower than 16 was 1.8% (95% CI, 1.7% to 1.8%) with the highest prevalence in India (6.2% [95% CI, 5.9% to 6.5%]), followed by Bangladesh (3.9% [95% CI, 3.4% to 4.3%]), Madagascar (3.4% [95% CI, 2.8% to 4.0%], Timor-Leste (2.9% [95% CI, 2.4% to 3.2%]), Senegal (2.5% [95% CI, 1.9% to 3.2%]), and Sierra Leone (2.2% [95% CI, 1.3% to 3.0%]); and 6 countries had prevalences lower than 0.1% (Albania, Bolivia, Egypt, Peru, Swaziland, and Turkey). The prevalence of BMI lower than 16 in women with a secondary or higher education level was 0.51% (95% CI, 0.47% to 0.55%), and in mutually adjusted models, a less than primary education level was associated with an OR of 1.4 (95% CI, 1.2 to 1.6). The prevalence of BMI lower than 16 was 0.43% (95% CI, 0.37% to 0.48%) in the highest wealth quintile with an OR of 3.0 (95% CI, 2.4 to 3.7) in the lowest wealth quintile. Among the 24 of 39 countries with repeated surveys, there was no decrease in prevalence. In Bangladesh and India, rates were declining with an average absolute change annually of -0.52% (95% CI, -0.58% to -0.46%) in Bangladesh and -0.11% (95% CI, -0.12% to -0.10%) in India. CONCLUSIONS AND RELEVANCE: Among women in 60 LMIC, the prevalence of BMI lower than 16 was 1.8%, and was associated with poverty and low education levels. Prevalence of BMI lower than 16 did not decrease over time in most countries studied.