Study Hypotheses:

1. Anaemia, which is frequent before conception as well as during early pregnancy, affects metabolism and foetal growth trajectories, influencing the risk of NCDs in the offspring.
2. Anaemia from conception till end of 2nd trimester is most detrimental for foetal and newborns' health, compared to 3rd trimester anaemia.
3. Anaemia from conception till end of 2nd trimester affects foetal and newborns health through poor placental development reflected in increased villous branching and changed umbilical and uterine blood flow.
4. Anaemia in early pregnancy disrupts the vascular endothelial growth factor A (VEGF-A)/placental growth factor (PlGF) balance and the insulin-like growth factor (IGF) axis resulting in poor placental development, and poor health of newborns. This may be reflected in specific methylation patterns.
5. Anaemia's impact on the risk for NCDs in the offspring may be mediated via epigenetic mechanisms, including changes in DNA methylation patterns.

Specific objectives are:

1. Characterize the health of women aged 18-40 years before conception and during pregnancy, focusing on prevalence of anaemia, infections, nutritional status, and Non-Communicable Diseases (NCDs).
2. Describe how 1st and 2nd trimester anaemia compared to 3rd trimester anaemia alters foetal growth and newborns' body composition.
3. Evaluate how 1st and 2nd trimester anaemia compared to 3rd trimester anaemia affects villous branching in the placenta, as well as uterine and umbilical artery blood flow.
4. Characterize how 1st, 2nd and 3rd trimester anaemia differentially changes the vascular endothelial growth factor A (VEGF-A)/placental growth factor (PlGF) balance and the insulin-like growth factor axis.
5. Determine which markers for foetal programming such as methylation of regulatory genes related to metabolism and haematopoiesis may be discovered early after exposure to anaemia in1st and 2nd trimester compared to 3rd trimester anaemia.

Outcome parameter for the pre-pregnancy study (i.): 1) Prevalence and severity of anaemia among non-pregnant Tanzanian women at fertile age and their nutritional status. 2) Incidence of conception among such women.

Outcome parameter for the pregnancy study (ii.): 1) Foetal growth rate in 2nd and 3rd trimester in offspring of anaemic mothers vs Foetal growth rate in non-anaemic mothers' offspring, 2) Comparison of anaemic mothers' vs non-anaemic mothers' newborn's body composition, placental villous branching, umbilical and uterine blood flow, VEGF-A/PlGF levels, and insulin-like growth factor axis.

For all analyses main exposure variable will be anaemia (Hb<5mmol/L) in 1st, 2nd and/or3rd trimester of pregnancy. Confounding exposure variables will be chronic health conditions (e.g. HIV), all temporary conditions mentioned above (e.g. malaria) as well as socioeconomic status and paternal characteristics. Regression modelling using both multiple linear and logistic regressions as well as modelling for repeated measures will be used for statistical analyses. For genetic and epigenetic changes, findings in the subgroup of 180 other/newborns, remaining significant after correction for multiple comparisons, will be validated in the remaining cohort. This will enable us within one study to replicate initial findings in one representative and distinct subgroup in another larger comparable study group from an overall homogenous study. Indeed, findings surviving both corrections for multiple testing in the initial hypothesis generating array approaches, as well as being replicated in the remaining study group with single target validation techniques, are highly likely to be biological important. Possible findings from the candidate gene approach will be corrected for multiple testing and the significant results will be compared with previous findings of the same genes from studies in muscle and adipose tissue conducted by Diabetes and Metabolism, Copenhagen University Hospital.

Inclusion Criteria:

• wishing to get pregnant
• negative pregnancy test
• not using family planning

Exclusion Criteria:

• not wishing to get pregnant
• being pregnant
• using family planning

Amendment October 2014: replacement of two-phase case-control study by two sub-projects, running in parallel, and following two distinct groups of women, 1) a cohort of 1500 non-pregnant women, disregarding Hb level, of whom 270 followed throughout pregnancy disregarding Hb level and gestational age at first contact, and 2) a case control study of 480 pregnant women followed from 1st trimester with a 160:160:160 ratio of women with Hb≤8g/dL: 8.1-10.9g/dL: ≥11g/dL.

The Danish Council for Strategic Research notified and accepted in November 2014.