Drinking tea and coffee during pregnancy-REDUCES-baby-size

Alpha · 2018-11-23 17:32:40 UTC

Coffee is evil

PaulL · 2018-11-24 04:17:34 UTC

Here’s a link to the abstract:

ncbi.nlm.nih.gov

Associations of maternal caffeine intake with birth outcomes: results from the Lifeways Cross Generation Cohort Study.

LW Chen, R Fitzgerald, CM Murrin, J Mehegan, CC Kelleher and CM Phillips, The American journal of clinical nutrition, Dec 2018 01

Maternal caffeine intake is associated with adverse birth outcomes, but in most studies the primary caffeine source is coffee; the influence of tea caffeine remains unclear.The aim of the study was to examine the association between maternal caffeine intake and birth outcomes in a population with tea as the predominant caffeine source.Data from 941 Irish mother-child pairs of the Lifeways Cross Generation Cohort Study were examined. Maternal dietary intakes in early pregnancy were assessed using a validated food-frequency questionnaire. Caffeine intake was derived from coffee, tea, soft drinks, and cocoa-containing foods and beverages. Associations of maternal caffeine intake with continuous (birth weight, birth length, and gestational age) and binary [low birth weight (LBW) (<2500 g) and preterm birth (PB) (<37 wk gestational age)] birth outcomes were investigated using multiple linear and logistic regressions, respectively, with adjustment for potential confounders.Tea was the predominant caffeine source (48%), followed by coffee (39%). In the fully adjusted model, maternal caffeine intake was associated with lower birth weight [β (95% CI): -71.9 (-105.4, -38.4) g · 100 mg-1 · d-1 caffeine increment], shorter birth length [-0.30 (-0.49, -0.11) cm], smaller head circumference [-0.12 (-0.24, -0.01) cm], and shorter gestational age [-0.13 (-0.25, -0.02) wk]; higher risks for LBW [OR (95% CI): 1.47 (1.14, 1.90)] and PB [1.36 (1.07, 1.74)] were also observed (all P < 0.05). The associations were robust to the exclusion of participants with pregnancy complications and in never smokers. Similar higher risks of adverse birth outcomes were observed for the highest caffeine intake categories from coffee [ORLBW: 3.10 (1.08, 8.89); ORPB: 2.74 (1.05, 7.16)] and tea [ORLBW: 2.47 (1.02, 6.01); ORPB: 2.56 (1.14, 5.75)], compared with the lowest intake categories (all P < 0.05).Maternal caffeine intake from both coffee and tea is associated with adverse birth outcomes. This prospective observational study was registered at ISRCTN Registry as ISRCTN16537904.

I don’t know. This is epidemiology, they used a food-frequency questionnaire, the effects are small, and the p-values are quite high (0.05). On the other hand, it’s a reasonably powerful study (n = 941). I just worry about confounders.

For example, I read something the other day claiming that the study showing an association between alcohol consumption and fetal development problems was confounded by economic status: the low drinkers were also the women who could afford better health care. Apparently, correcting for income level causes the association to vanish. So I wonder if there’s an overlooked factor contributing to the results in this study,

Alpha · 2018-11-24 05:08:20 UTC

I agree with you

studies based on questioners have too many holes.

lots of unknowns and variables.

I like studies based on measurements and large statiscal samples rather than asking if a woman drank coffee during pregnancy and how many cups a day without a mention of other foods.

but I am disparate to blame coffee for anything

Tmdlkwd · 2018-11-24 05:33:12 UTC

Lol
Fair enough

CatGirl · 2018-11-25 12:25:21 UTC

I’m a coffee and 2 pack a day cigarette baby of the late 60s. Being a Midwesterner, my mom drinks (weak-ish) coffee with every meal and at any and all times otherwise. Would I do it? Never. But my brother was 12lbs, my sister and I (twins) a total of 12lbs and my little sister was 8lbs.

I have an oddly high tolerance for caffeine and other stimulant type drugs. Not sure if connected.

ctviggen · 2018-11-25 15:51:11 UTC

As soon as they start using models, it’s over. There’s no way to know whether the model is correct, completely wrong, partially correct, etc.