Harmonization of maternal balanced energy-protein supplementation studies for individual participant data (IPD) meta-analyses - finding and creating similarities in variables and data collection.

BACKGROUND: Public health and clinical recommendations are established from systematic reviews and retrospective meta-analyses combining effect sizes, traditionally, from aggregate data and more recently, using individual participant data (IPD) of published studies. However, trials often have outcomes and other meta-data that are not defined and collected in a standardized way, making meta-analysis problematic. IPD meta-analysis can only partially fix the limitations of traditional, retrospective, aggregate meta-analysis; prospective meta-analysis further reduces the problems. METHODS: We developed an initiative including seven clinical intervention studies of balanced energy-protein (BEP) supplementation during pregnancy and/or lactation that are being conducted (or recently concluded) in Burkina Faso, Ethiopia, India, Nepal, and Pakistan to test the effect of BEP on infant and maternal outcomes. These studies were commissioned after an expert consultation that designed recommendations for a BEP product for use among pregnant and lactating women in low- and middle-income countries. The initiative goal is to harmonize variables across studies to facilitate IPD meta-analyses on closely aligned data, commonly called prospective meta-analysis. Our objective here is to describe the process of harmonizing variable definitions and prioritizing research questions. A two-day workshop of investigators, content experts, and advisors was held in February 2020 and harmonization activities continued thereafter. Efforts included a range of activities from examining protocols and data collection plans to discussing best practices within field constraints. Prior to harmonization, there were many similar outcomes and variables across studies, such as newborn anthropometry, gestational age, and stillbirth, however, definitions and protocols differed. As well, some measurements were being conducted in several but not all studies, such as food insecurity. Through the harmonization process, we came to consensus on important shared variables, particularly outcomes, added new measurements, and improved protocols across studies. DISCUSSION: We have fostered extensive communication between investigators from different studies, and importantly, created a large set of harmonized variable definitions within a prospective meta-analysis framework. We expect this initiative will improve reporting within each study in addition to providing opportunities for a series of IPD meta-analyses.

Revisiting maternal and child undernutrition in low-income and middle-income countries: variable progress towards an unfinished agenda.

13 years after the first Lancet Series on maternal and child undernutrition, we reviewed the progress achieved on the basis of global estimates and new analyses of 50 low-income and middle-income countries with national surveys from around 2000 and 2015. The prevalence of childhood stunting has fallen, and linear growth faltering in early life has become less pronounced over time, markedly in middle-income countries but less so in low-income countries. Stunting and wasting remain public health problems in low-income countries, where 4·7% of children are simultaneously affected by both, a condition associated with a 4·8-times increase in mortality. New evidence shows that stunting and wasting might already be present at birth, and that the incidence of both conditions peaks in the first 6 months of life. Global low birthweight prevalence declined slowly at about 1·0% a year. Knowledge has accumulated on the short-term and long-term consequences of child undernutrition and on its adverse effect on adult human capital. Existing data on vitamin A deficiency among children suggest persisting high prevalence in Africa and south Asia. Zinc deficiency affects close to half of all children in the few countries with data. New evidence on the causes of poor growth points towards subclinical inflammation and environmental enteric dysfunction. Among women of reproductive age, the prevalence of low body-mass index has been reduced by half in middle-income countries, but trends in short stature prevalence are less evident. Both conditions are associated with poor outcomes for mothers and their children, whereas data on gestational weight gain are scarce. Data on the micronutrient status of women are conspicuously scarce, which constitutes an unacceptable data gap. Prevalence of anaemia in women remains high and unabated in many countries. Social inequalities are evident for many forms of undernutrition in women and children, suggesting a key role for poverty and low education, and reinforcing the need for multisectoral actions to accelerate progress. Despite little progress in some areas, maternal and child undernutrition remains a major global health concern, particularly as improvements since 2000 might be offset by the COVID-19 pandemic.