PDF Summary:Expecting Better, by Emily Oster

...

Expecting Better is not generally meant as a prescriptive guide. Rather, just as Oster trains her economics students, you should 1) gather the right data about a decision, 2) weigh your personal costs and benefits, 3) make your personal decision based on all the information you have.

(Shortform note: in this summary, we’ll use “you” as though we’re speaking to the mother.)

...

Conception and the Ovulation Cycle

The ovulation cycle lasts 28 days on average. Starting from Day 1, when the period starts, the first 14 days are gearing up for ovulation. On Day 14, the egg is released and travels through the Fallopian tube toward the uterus. This journey takes around 6 days. If the egg is fertilized, it implants in the uterus around Day 23.

Critically, the egg survives only an average of 24 hours, well before it makes it to the uterus. Therefore sperm must be waiting around in the Fallopian tube, right as the egg is released.

Luckily, sperm live up to 5 days in the Fallopian tube. Therefore, the best chance of conception happens the day before or the day of ovulation. Conception can still happen up to 5 days before ovulation, with lower chances.

Here’s a graph of chances of conception depending on day of ovulation cycle, in women aged 26-35:

Myth: It’s best to have sex every other day, so you can hit at least one of the two best days and save up your sperm.

Reality: Chances of conception are nearly identical whether...

...

Here’s an example. Picture a study where you divide the population of mothers into abstainers and drinkers, then study their kids. If you picture the type of woman who has 0 drinks a week vs 2 drinks a week vs 10 drinks a week, you can already imagine differences in the child’s environment, such as their education, wealth, and location. When the kids grow up, if you find that children of drinkers end up with behavioral problems, it’s unclear whether it was the drinking or the other factors (like household conditions) that made a difference.

Similarly, people worry that coffee causes miscarriages. But women who drink more coffee tend to be older than those who don’t. What if it’s age that’s causing the miscarriages, and not the caffeine?

In one notable study on alcohol, Oster found a startling difference in cocaine use: 18% of women who didn’t drink reported using cocaine during pregnancy, compared to 45% of those having 1 drink per day!

Not only are the poor studies themselves confusing, their results are then cherry-picked by popular media and distorted into a shocking headline: “1 cup of coffee a day causes miscarriages.” Before you know it, myths start...

...

• Women with a previous miscarriage have a miscarriage chance of 25% in first trimester, compared to 4-5% for first pregnancies or women with a previous successful pregnancy.
• Age increases miscarriage rate: women below age 20 have a miscarriage rate of 4.4%; women aged 20-35 have a miscarriage rate of 6.7%; and women over 35 have a 19% miscarriage rate.
• IVF pregnancies have miscarriage rates of 30%, vs 19% for natural pregnancies.
• Vaginal bleeding signals increased risk: 13% of women with bleeding have miscarriages, vs 4.2% without.
• Lack of nausea signals a higher miscarriage rate.
• Low levels of progesterone may contribute to miscarriage.

...

Listeria grows well at fridge temperatures, so if you want to be safer, chuck out anything that’s been in the fridge for too long

The Best Fish to Eat

Fish have two competing effects on the fetus:

• High mercury lowers IQ.
  • A 1 ug/g increase in mercury sampling causes a decrease of 0.7 IQ points. In other words, the average American woman and the highest mercury-exposed woman produces a 3.5 IQ point difference in kids.
  • Predator fish and long-living fish (listed below) have the highest levels of mercury. The former eats other fish harboring mercury, and the latter accumulates mercury over a long lifetime.
• Omega-3 fatty acids increase IQ.
  • Increasing DHA fatty acids by 1 g/day would increase IQ by 1.3 points.

The solution is to eat high omega-3 fish with low mercury. The best fish to eat are: salmon, herring, sardines, pollock, and mackerel.

The worst fish that you should AVOID are: orange roughy, canned tuna, king mackerel, grouper, shark, and swordfish. These contain low omega-3 but have high mercury.

These fish are somewhere in the middle: tilefish, halibut, sushi-grade tuna, flounder, and snapper.

...

Pregnant women have had scares with anti-nausea medication before—thalidomide caused birth defects, and Bendectin (B6 + Unisom) was pulled off the market. However, Oster argues that the mother’s comfort is quite important, particularly if it affects her diet and well-being, and OTC anti-nausea medications are not shown to be bad.

Oster’s recommendations for nausea cures, in declining order of preference:

1. Eat smaller meals
2. Vitamin B6
3. Vitamin B6 + Unisom
4. Zofran

...

Noninvasive Test: Cell-free Fetal DNA

Cell-free fetal DNA tests use fetal DNA that is circulating in the mother’s bloodstream.

The mother’s bloodstream contains not just fetal cells but also fetal DNA outside of cells. In fact, 10-20% of DNA isolated from maternal plasma is from the fetus.

Theoretically, if the fetal DNA could be isolated fully from the mother’s DNA, then full genomic sequencing could occur. We’re not quite there yet, but bulk defects can still be detected. For instance, if a higher-than-expected proportion of chromosome 21 is in the mix, there’s a good chance it indicates trisomy 21.

The sensitivity (true positive rate, or accurate detection of a disorder) is remarkably high—in one large-scale study, 99.1% of trisomies are detected with this procedure. The false negative rate is also low, at 0.05% (of 146,000 women without Down Syndrome, 61 were given a false positive).

Given your age and test result, then, what is the likelihood of your child having Down Syndrome? Here’s a helpful chart:

...

The recommended limit on radiation exposure for the duration of pregnancy is 1 mSv (millisievert). This is conservative, with experts suggesting there are no definitive data showing fetal harm at doses below 20 mSv. Oster does note that exposure to twice this limit might increase the risk of the child having a fatal cancer by 1 in 5,000.

How does this translate to flying? A single 4-hour flight would deliver at most 0.02 mSv, or 2% of the limit. And flight attendants receive around 1.5 mSv per year, so each individual flight is not a big deal.

How about airport security? A single screening has a maximum of 0.25 μSv, or 0.025% of the limit. In other words, you could get a screening every single day and that would represent 9% of your annual conservative limit.

In other words, unless you’re a flight attendant, you probably won’t reach the conservative limit. But you can always opt for the pat-down instead of the X-ray.

(Shortform note: to put these radiation doses into perspective, here are examples of radiation doses in normal life:

• 0.25μSv: maximum limit on dose from a single airport security screening
• 5-10μSv: one set of dental X-rays
• 0.3-0.5mSv: dosage...

...

• 42% of SGA babies have complications of difficulty breathing, irregular blood sugar regulation, and abnormal neurological signs
• A SGA baby born prematurely has higher risk of complications than an appropriate weight baby. In one study:
  • Chronic lung disease increased from 29% to 57%
  • Mortality increased from 17% to 33%

Risks of Too Much Weight Gain

Eating too much in pregnancy could cause the fetus to be large for gestational age (LGA):

• Gaining 10 lb more than recommendation doubles risk of LGA babies (from 5% to 10%) and halves risk of SGA babies (from 10% to 5%)
• LGA babies roughly double the risk of requiring C-section and instrument-assisted delivery, and roughly triple risk of needing resuscitation or ICU transfer. (Shortform note: we could not find the baseline rates from Oster’s references.)

Another concern is that too much weight gain may cause your child to be overweight later in life. This has been shown in animal studies, with a possible mechanism being insulin resistance in the womb leading to insulin resistance in life.

But this is hard to control from environmental factors and parent habits, and Oster argues the effects are...

...

• Some believe that Y sperm (leading to males) are fast swimmers but die off quickly, so if you want to have a girl, then have sex earlier. This is called the Shettles method. There is no evidence that this works.
  • Some believe female fetuses have faster heart rates. This is a myth: fetal heart rate does not differ between genders.

...

The recommendation is to sleep on the left side, not the back. The evidence is currently inconclusive - some studies show sleeping on the back or right side doubled stillbirth rate. However, a blood flow study found that lying on your back found no particularly bad impact on blood flow.

Some women feel faint on their backs - if so, then they should switch to their left side. If you don’t feel faint, it’s probably fine, though if you can sleep on your left, that’s preferred.

Taking OTC sleep aids like Unisom are fine. Ambien has mixed safety data, but Oster believes it’s fine occasionally.

...

*   Ambien (insomnia)
*   Prilosec (heartburn)
*   Ibuprofen (pain)

Category D

• Definition: Evidence of harm to fetus in human studies, but benefits may outweigh risks in certain situations.
• Once a drug falls in this bucket, it’s hard to get out given how difficult it is to prescribe and get new studies.
• Examples:
  • Paxil/paroxetine (depression)
  • Lorazepam (anxiety)
  • Prinivil/lisinopril (hypertension)
  • Tetracycline (antibiotic, acne)
  • Aspirin

Category X

• Definition: Should not be taken during pregnancy under any circumstances.
• Also includes drugs that have no purpose during pregnancy, like oral contraceptives
• Examples:
  • Accutane/isotretinoin (acne)
  • Leflunomide (arthritis)
  • Warfarin (anticoagulant)
  • Lipitor/atorvastatin, Zocor/simvastatin (high cholesterol)
  • Finasteride (enlarged prostate)

... </tr> 34 1.39% 1.1% 35 2.33% 0.8% 36 4.37% 0.6% 37+ 89.09% 0.2% </table>

Chances of a preterm baby are thus sizable, but still in the small minority.

Delaying Labor

If you go into labor early, what happens?

1. You receive tocolytics (eg magnesium sulfate) to suppress contractions and delay birth for 1-2 days.
2. You receive steroid shots, which speeds up fetal development.
   • Steroids result in 30% decrease in fetal death.
3. You can get transported to a hospital with a higher-level NICU, with more sophisticated equipment to handle very...

...

• Risk for stillbirth, neonatal metabolic problems

Management

• Glucose monitoring and control

Rh Alloimmunization

What: Baby has Rh+ blood, Mom has Rh-. Mother produces antibodies that can cross placenta and lead to destruction of fetal red blood cells.

Rate: 0.1%

Risk Factors

• Requirement: mother must be Rh- and father is Rh+ (homozygous or heterozygous)
• Amniocentesis/CVS causes blood mixing
• Bleeding in pregnancy
• Previous pregnancies with Rh incompatibility (first baby is less at risk)

Consequences

• Risk of fetal anemia and hyperbilirubinemia

Management

• Rhogam shot at 28 weeks and after delivery. This destroys fetal red blood cells before the mother can develop immunity to them. This dramatically reduces risk of Rh disease.

Cervical Insufficiency/Incompetence

What: Painless dilation of the cervix before reaching term

Rate: 1-2%

Risk Factors

• History of cervical biopsy
• Repeated procedures to cervix during pregnancy
• Significant trauma to cervix

Consequences

• 2nd trimester miscarriage
• Very preterm birth

Management

• Cervical length screening
• Progesterone treatment
• Cerclage (sutures...

...

• Cervical dilation
• Baby drops down in pelvis a few days/weeks prior to labor
• Bishop score - combines the above into a score from 0 to 13. >6 is fairly advanced

Bishop score and cervical effacement are predictive of vaginal deliveries.

At 42 weeks, you’re almost certain to have labor induced, likely with Pitocin (synthetic oxytocin), which stimulates contractions. This is to reduce the risk of stillbirth with babies who are born too late. Oster argues that Pitocin may increase pain in labor and increase risk of C-section, which is why she was willing to wait until the very last moment to induce.

...

Epidurals are very popular and are used in about 66% of births in the US.

For the baby, epidurals don’t have big differences in APGAR score, fetal distress, or NICU time. The epidural does raise the risk of a maternal fever during labor, which could prompt doctors to administer antibiotics to be safe and thus expose the baby to antibiotics.

Cons of an epidural:

• Increased risk of forceps or vacuum extractor in delivery
• Greater use of C-section for fetal distress
• Longer pushing time (15 minutes)
• Greater use of Pitocin in labor to compensate for slowing in contractions
• Higher chance of baby facing up at birth (possibly due to less walking)
• Greater chance of maternal hypotension
• Longer recovery postpartum
• Increased chance of fever in labor
• 1/200 epidurals cause a wet tap, where the needle goes into the spinal fluid and causes a headache for a few days

Oster ultimately chose to deliver both children without an epidural.

...

*   Oster thinks having one was one of her best pregnancy decisions.

• Monitoring should be done intermittently or mobile, not continuous or immobilized.
  • Doctors want to track fetal heart rate to detect fetal distress. More than 85% of women have monitoring during labor.
  • Options for type of monitoring include immobilizing monitoring, portable monitoring, or an internal monitor (that’s threaded through the cervix and screwed into the baby’s scalp).
  • Options for style of monitoring include continuous, or intermittent listening.
  • Continuous monitoring leads to much more intervention, increasing C-section risk by 60% and use of instruments, but without any difference in baby outcomes (APGAR scores, NICU admissions). Oster argues this is because continuous monitoring leads to false positives about fetal distress.
  • Some hospitals may have a non-negotiable policy on monitoring.
• If labor progression is slow, we prefer augmentation by amniotomy (breaking water) first, then Pitocin.
  • Both interventions generally don’t have other complications like increasing risk of C-section.
  • Oster preferred Pitocin last because of...

...

• Putatively this is useful for bone marrow transplants for leukemia. However, keep in mind a child cannot use her own cord blood; she must use a sibling’s. The chance of using banked cord blood is 1 in 20,000.
• Another sales pitch is that cord blood could be used for regenerative medicine in the future. Oster argues that by that point, we’ll probably have better stem cell technologies in general, including techniques to derive stem cells from ordinary cells.
• Consider banking in a public blood bank instead to help a stranger out.

...

If you do choose home birth, Oster suggests working with a certified nurse-midwife, who will have the most training and certification. Certified professional midwives don’t have a nursing degree, and direct entry midwives don’t have a college degree and no license.

A good compromise: your hospital may have a more comfortable birthing center, a comfier room with a tub, a bigger bed, and less fetal monitoring.