What is the effect of stress while pregnant on the health of the baby? This is a question faced by all women – both in developed and developing countries. Indeed, in developed countries, stress may be more prevalent than many physical stressors such as nutritional deficiencies. Despite the pervasiveness of stress in modern society, particularly during pregnancy, we have a very limited understanding of the effect on children’s health when they are born, and whether there are any long-term negative effects in terms of cognitive development and on labour market outcomes. Our research attempts to fill this void.
Mechanisms and stress
Much is now known about the effects of shocks to the physical health of pregnant women on the outcomes of the in-utero children (see Almond and Currie 2010 and Currie 2011 for surveys of some of this work). While evidence generally suggests that adverse health or nutrition shocks to pregnant women have significant and often long-lasting effects on the outcomes of their children, much less is known about the effects of psychological stresses.
There are a number of mechanisms through which stress could affect a developing foetus. The medical literature identifies several potential mechanisms:
- Stress triggers the production of a placental corticotrophin-releasing hormone, which has been shown to lead to reduced gestational age and low birth weight.
- Stress suppresses the immune system, thereby making pregnant mothers more susceptible to sickness, and can cause high blood pressure, which increases the chance of having pre-term labour or a low birth weight infant.
- There may be adverse behavioural responses to stress, such as smoking cigarettes or drinking alcohol, which can also have adverse effects on the health of the foetus.
In general, it is difficult to identify the causal effect of stress due to the difficulty of separating out the stress effect from the direct effect of the event causing the stress.
Using the death of a parent while pregnant
In our research, we use register data on the population of Norway to examine the effect of stress at the individual level that is induced by the death of a pregnant woman’s parent. We study approximately 25 Norwegian cohorts from 1967 onwards to assess the effects of maternal stress on short-term outcomes such a birth weight, APGAR scores, gestation length, birth length, and whether they were born through C-section (Black et al. 2014). We then evaluate longer-term effects such as on IQ scores for boys aged about 18 (from military records), as well as completed education and earnings from the age of 25.
A number of issues arise when one tries to estimate the effect of stress caused by the death of a pregnant woman’s parent on the outcomes of the in-utero children. While we address each individually in our research paper, we describe here a few of the more salient ones.
One potentially important issue that arises is that individuals who have a parent die younger are not a random sample of the population. To deal with this problem, we use two approaches in our estimation strategy. Our first approach involves comparing the outcomes of children who experience the death of a grandparent while in utero to the outcomes of similar children who experience the death of a grandparent in the period just before or just after the in-utero period. By making this comparison, we are able to isolate the effect of a death during pregnancy from the effect of a death in the general time period around pregnancy. The second approach compares two children born to the same mother, with the mother experiencing a parental death during only one of the pregnancies (this is a common approach in this literature – see Currie and Rosin-Slater 2013, Black et al. 2013). Reassuringly, our results from these two very different approaches are very similar.
Effects on child health
We find that, regardless of which approach we use, a parental death experienced while pregnant leads to small but statistically significant negative effects on birth outcomes, including birth weight and APGAR scores. We find that the effects are larger for boys than for girls, which is consistent with a more general literature that boys are more vulnerable to in-utero shocks (Eriksson et al. 2010). When we look by cause of death, we find that the effects are largest when cardiovascular disease is the cause of death, tentatively suggesting that more unexpected deaths have bigger adverse effects.
While the process through which bereavement affects birth outcomes may by purely physiological and be related to the body’s natural responses to stress and grief, there may also be behavioural responses of the mother that affect the foetus. One such possibility is that stress increases the likelihood or level of maternal smoking. We have information on smoking behaviour of mothers from 1998 onwards and we have used it to see whether bereavement affects smoking. We found no evidence of any effect. This suggests that behavioural responses may not be an important part of the story, but more research with more detailed data (for example, on other behaviours such as drinking and pre-natal care) will be required to verify this conclusion.
Some earlier work on the effect of physical insults in utero, such as those generated as a result of exposure to nuclear fallout from Chernobyl while pregnant, found little if any effect on short-run health outcomes, but did find longer-run cognitive effects (Almond et al. 2009), so it is not clear what our expectation is in terms of exposure to stress. In fact, we find no evidence for adverse effects on the children’s outcomes later in life such as cognitive test scores, educational attainment, height, and earnings. The confidence intervals imply that we can rule out negative effects on completing 12 years or more of schooling of more than 1.6%, negative effects on full-time earnings of more than 4%, negative effects on IQ score (for men) of more than 7% of a standard deviation, and negative effects on adult height (of men) of more than half a centimetre.
Using unique data from Norway, we are able to estimate the effect of stress induced by the death of a parent while pregnant on the outcomes – both short- and long-run – of the children. We find that maternal bereavement has small but statistically significant adverse effects on birth outcomes, and these effects are larger for boys than for girls. However, there is no evidence of any adverse effects on any of our later outcomes including cognitive test scores, educational attainment, and earnings. This suggests that, even though there may be measurable effects on birth outcomes, acute psychological stressors during pregnancy have limited adverse consequences on the child’s success in education and the labour market.
Almond, Douglas J (2006), “Is the 1918 Influenza Pandemic Over? Long‐Term Effects of In Utero Influenza Exposure in the Post‐1940 U.S. Population”, Journal of Political Economy, 114(4): 672–712.
Almond, Douglas J and Janet Currie (2010), “Human Capital Development Before Age Five”, Chapter 15 in Orley Ashenfelter and David Card (eds.), Handbook of Labor Economics, Volume 4b, North Holland: 1315–1486.
Black, Sandra E, Aline Bütikofer, Paul J Devereux, and Kjell G Salvanes (2013), “This Is Only a Test? Long-Run Impacts of Prenatal Exposure to Radioactive Fallout ”, NBER Working Paper 18987.
Black, Sandra, Paul J Devereux, and Kjell Salvanes (2014), “Does grief transfer across generations? In-utero deaths and child outcomes ”, NBER Working Paper 19979.
Currie, Janet (2011), “Inequality at Birth: Some Causes and Consequences”, American Economic Review, 101(3): 1–22.
Currie, Janet and Maya Rossin-Slater (2013), “Weathering the Storm: Hurricanes and Birth Outcomes”, Journal of Health Economics, 32(3): 487–503.
Eriksson, J G, G Kajantie, C Osmond, K Thornburg, and D J Barker (2010), “Boys Live Dangerously in the Womb”, American Journal of Human Biology, 22(3): 330–335.