NLM5008 Chap.7 Diet in Pregnancy and the Postpartum Period
Diet in Pregnancy and the Postpartum Period
The Week 4 question
Week 4 looks at how eating during pregnancy and after delivery affects both the pregnancy itself and the child's health years later, asking which nutrients and dietary patterns matter and what an optimal diet might look like.
Its two readings answer with two very different designs, and the contrast between them is the most useful thing to carry from the week.
A whole diet, scored
The first reading pools two prospective mother-child cohorts, Project Viva in Massachusetts with 997 pairs and Rhea in Crete with 569. Diet in pregnancy was scored on a Mediterranean diet score from 0 to 9: one point for eating above the recommendation for each of eight beneficial components, and one point for eating red and processed meat at or below the threshold.
Both cohorts averaged in the low band. Children were measured by study staff at a median of 7.7 years in Viva and 4.2 years in Rhea.
In the pooled analysis, each three-point increase in the score was associated with a BMI z-score about 0.14 units lower, a waist about 0.39 cm smaller, smaller skin-fold thicknesses and slightly lower systolic and diastolic blood pressure.
The associations were not explained by birth or infant characteristics.
The authors list self-reported diet, an equally weighted score, different child ages across cohorts and likely lower-income families lost to follow-up, and they call for intervention trials because an observational design cannot establish cause.
Randomising support rather than feeding
The second reading is PROBIT, a cluster-randomised trial in Belarus from June 1996 to December 1997. Observational breastfeeding studies are open to selection and confounding, and mothers cannot ethically be assigned to breast or formula feeding, so the trial randomised 31 maternity hospitals and polyclinics to a breastfeeding promotion programme modelled on the Baby-Friendly Hospital Initiative, or to usual practice.
Among 17,046 mother-infant pairs, the intervention raised breastfeeding at 12 months and exclusive breastfeeding at three months, and reduced gastrointestinal infection and atopic eczema in the first year. Respiratory infection did not differ.
Reading the two together
The cohorts capture a real diet over a real pregnancy but can only adjust for what they measured.
The trial balances unmeasured factors by design but tests a promotion programme rather than breastfeeding itself. Knowing which question each design can answer is the point of the week.
What this chapter covers
- 01
Nutrition in pregnancy and after birth as the Week 4 question
- 02
How the Mediterranean diet score is built
- 03
Pooled results from Project Viva and Rhea
- 04
Limits of an observational diet study
- 05
PROBIT and the logic of cluster randomisation
- 06
Reading weaning odds ratios in the right direction
Worked example · free
Reading a breastfeeding odds ratio that runs backwards
- 1Interpret the exclusive breastfeeding percentages.
- 2State what the odds ratio measures.
- 2Interpret the ratio and its interval.
Key terms
- Mediterranean Diet Score
- A 0 to 9 score of adherence built from nine food and nutrient components.
- Cluster Randomised Trial
- A trial that randomises groups such as hospitals rather than individuals.
- Exclusive Breastfeeding
- Feeding only breast milk, with no solids, other milk, water or other liquids apart from medicines.
- BMI Z-Score
- A child's BMI expressed in standard deviations from the reference mean for age and sex.
- Baby-Friendly Hospital Initiative
- A WHO and UNICEF programme supporting the start and continuation of breastfeeding.
- Metabolic Programming
- The idea that nutrition in early life sets metabolic function in ways that persist.
- Intention To Treat
- Analysing participants in the groups they were randomised to, whatever they later did.
Diet in Pregnancy and the Postpartum Period FAQ
Did the Mediterranean diet in pregnancy cause lower child adiposity?
The study cannot say. It found that higher scores were associated with lower child BMI z-score, waist and skin-folds in two pooled cohorts, but the design is observational, so the authors call for intervention trials to test whether the diet itself is responsible.
Why did PROBIT randomise hospitals rather than mothers?
Assigning mothers to breast or formula feeding would be neither feasible nor ethical, and randomising individual mothers within one hospital would let the programme spill over. Randomising whole sites to a promotion programme avoids both problems.
What did PROBIT not find?
It found no reduction in respiratory tract infection, with 39.2% and 39.4% of infants affected in the two arms. Mortality differences were small and not statistically significant, and the authors note the trial was not powered for them.
Why report diet effects per three points of score?
Three points is the width of each conventional band, low, moderate or high adherence, and both cohorts averaged in the low band, so a per-increment estimate describes a realistic improvement rather than a jump from the lowest to the highest scores.
What are the main limitations of the diet cohorts?
Self-reported diet, a score that weights every component equally and depends on chosen cut-offs, children measured at different ages in the two cohorts, and loss to follow-up that probably removed lower-income families, which may mean the true associations are underestimated.
Assessment move
Write the two Week 4 readings side by side under the same headings: design, unit randomised or observed, exposure, outcome, direction and what the design cannot show. The comparison is more examinable than either paper alone.
Working through Diet in Pregnancy and the Postpartum Period in NLM5008? Sia is AskSia’s AI Health and Medicine tutor — ask any NLM5008 Diet in Pregnancy and the Postpartum Period question and get a clear, step-by-step explanation grounded in how NLM5008 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
