There is ample empirical evidence showing that conditions early in life have long-term consequences for health, with the prenatal months being a key critical period (Almond and Currie 2010, Almond et al. 2018, Conti et al. 2019). Such studies usually examine the exposure to adverse conditions due to unexpected shocks (severe or mild) that may occur during pregnancy and may have short-term, medium-term, and long-term effects. Examples of severe shocks include natural disasters, pandemics, famines, military conflicts, terrorist attacks, or even nuclear accidents. Pregnancy is a sensitive period for both mothers and their children and is naturally a period of high concern and worry for those who are going through this during a period of crisis, such as the COVID-19 pandemic (Schwandt 2020).
The famines of WWII had devastating consequences for civilian populations across Europe and globally (Ó Gráda 2019). War can affect physical and mental health via multiple channels, such as experience of hunger, dispossession, absence of the father, and stress from combats (Kesternich et al. 2014). Several papers have studied long-term impacts of episodes of famine at the time of WWII. For instance, Jürges (2013) finds negative effects of the German famine on both education and labour market outcomes, Neelsen and Stratmann (2011) find negative impacts on education and literacy for those exposed to the Greek famine, and Allais et al. (2021) find worse health outcomes for women exposed to episodes of hunger in France.
Evidence from the Dutch Hunger Winter
The liberation of the Netherlands from the German Nazi occupation during WWII started in September 1944, but it was only when the Nazi surrendered in May 1945 that the entire country was liberated. Until then, the Western part of the country (see Figure 1) experienced – especially in the big cities – a severe famine as a result of a Nazi blockade, triggered by a Dutch national railways strike aimed at facilitating the Allied liberation efforts. During the winter of 1944-45, the food and living circumstances rapidly worsened due to low temperatures, freezing of the canals, and the military stalemate of the Allied forces on the Dutch front. While throughout the occupation food rations were maintained at around 1,800 calories per day per person, they dropped to below 1,000 calories by November 1944 and to 500 calories by April 1945. The famine only ended with the liberation of the Western Netherlands in early May 1945. While the famine affected the entire population living in the Western Netherlands, more than 40,000 individuals were exposed in utero, making it a suitable ‘natural’ experiment to study the consequences of exposure to severe prenatal shocks.
Figure 1 The liberation of the Netherlands during the final stages of WWII
Source: Ekamper et al. (2020).
Previous studies have examined the Dutch famine of 1944-45. The landmark study of Ravelli et al. (1976) found an increase in the likelihood of obesity among those exposed in the first half of pregnancy; Scholte et al. (2015) find negative effects on health care use and employment at age 55 for those exposed in middle or late gestation; and Portrait et al. (2017) find a significant reduction in height for exposure between gestation and age 2.
The complexity of the mechanisms at play
While most previous work has compared cohorts born in different periods in regions with varying exposure to the famine, we argue that a more in-depth examination of the mechanisms at play can reveal important information about the co-occurring shocks which affected these cohorts. These may be the low-caloric and imbalanced nutritional content of the rations, but also the harsh weather conditions and the ongoing warfare. Additionally, such harsh conditions might have both scarring and selection effects, possibly acting in the opposite direction – for example, the weakest might be less likely to survive, resulting in a population with better health. Whether one or the other effect dominates is ultimately an empirical question. As argued by Bozzoli et al. (2009), while selection effects tend to dominate in very hard conditions with high infant mortality, scarring effects are more pronounced in milder adverse conditions with low infant mortality.
In our paper (Conti et al. 2021), we explore all these factors by re-examining the data on military recruits examined at 18 years old by Ravelli et al. (1976) and combining them with newly collected and digitised data from historical sources. These include information on caloric content and nutrient composition of the rations, average daily temperature, number of deaths due to warfare, along with city-level information on demographic and socioeconomic characteristics, number of stillbirths, neonatal and post-neonatal deaths, among others. Our analysis focuses on health outcomes at the time of the military examination at age 18, and examines the impacts of exposure during each of the three trimesters of pregnancy, as we know that different stages of fetal development entail susceptibility to different shocks (Koppensteiner and Manacorda 2016).
Our analysis confirms that famine exposure starting prenatally is more harmful than solely postnatal exposure. More precisely, we show greater negative health impacts among those exposed since early gestation. The cohorts affected by the Dutch famine had, on average, a significantly higher weight and a higher prevalence of obesity when they reached late adolescence. These results are primarily driven by a change in the nutritional content of the rations, through a reduction in the proportion of proteins. This suggests that looking at the fall in calories alone is not enough for a full assessment of the impacts occurring via the nutritional channel. Additionally, we show that exposure to warfare is a second mechanism – possibly operating via stress, or loss of income.
Lastly, we show that conditioning on the sample of survivors leads to a downward bias in estimating the impacts of the famine, and we provide evidence of both selection and scarring effects. On the one hand, those at the bottom of the body mass index (BMI) distribution are less likely to survive. On the other hand, even after accounting for selection, we observe substantial scarring effects that result in a higher prevalence of obesity among the survivors. Accounting for selection also reveals a significant effect of prenatal famine exposure on the likelihood of having an intellectual disability, which can be rationalized by the fact that the more vulnerable were less likely to survive in the first place.
Wars and famines are still a reality in the modern world and many children are given an unfair start in life, sometimes as early as in utero. Our results provide evidence on the importance of nutritional composition of the diet – above and beyond a measure of caloric intake – and on the role of both selection and scarring effects in explaining the impacts of severe prenatal shocks. Our study confirms the need for a deeper understanding of the various channels through which early conditions affect health later in life.
Allais, O, G Fagherazzi and J Mink (2021), “The long-run effects of war on health: Evidence from World War II in France”, Social Science & Medicine 276: 113812.
Almond, D and J Currie (2010), “The long-term impact of life before birth”, VoxEU.org, 24 June.
Almond, D, J Currie and V Duque (2018), “Childhood circumstances and adult outcomes: Act II”, Journal of Economic Literature 56(4): 1360-1446.
Bozzoli, C, A Deaton and C Quintana-Domeque (2009), “Adult height and childhood disease”, Demography 46(4): 647–669.
Conti, G, G Mason and S Poupakis (2019), “Developmental Origins of Health Inequality”, in Oxford Research Encyclopedia of Economics and Finance.
Conti, G, S Poupakis, P Ekamper, G E Bijwaard and L H Lumey (2021), “Severe Prenatal Shocks and Adolescent Health: Evidence from the Dutch Hunger Winter”, CEPR Discussion Paper 16633.
Ekamper, P, G E Bijwaard, F W A van Poppel and L H Lumey (2020), “War- and famine-related excess mortality among civilians in the Netherlands, 1944-1945”, Journal of Maps 16(1): 124-131.
Jürges, H (2013), “Collateral damage: The German food crisis, educational attainment and labor market outcomes of German post-war cohorts”, Journal of Health Economics 32(1): 286–303.
Kesternich, I, B Siflinger, J P Smith, and J K Winter (2014), “The effects of World War II on economic and health outcomes across Europe”, The Review of Economics and Statistics 96(1): 103–118.
Koppensteiner, M and M Manacorda (2016), “The effect of day-to-day violence on infant health”, VoxEU.org, 18 April.
Neelsen, S and T Stratmann (2011), “Effects of prenatal and early life malnutrition: Evidence from the Greek famine”, Journal of Health Economics 30(3): 479–488.
Ó Gráda, C (2019), “The famines of WWII”, VoxEU.org, 2 September.
Portrait, F, T van Wingerden, and D Deeg (2017), “Early life undernutrition and adult height: The Dutch famine of 1944-45”, Economics & Human Biology 27: 339–348.
Ravelli, G-P, Z A Stein, and M W Susser (1976), “Obesity in young men after famine exposure in utero and early infancy”, New England Journal of Medicine 295(7): 349–353.
Scholte, R S, G J Van den Berg, and M Lindeboom (2015), “Long-run effects of gestation during the Dutch hunger winter famine on labor market and hospitalization outcomes”, Journal of Health Economics 39: 17–30.
Schwandt, H (2020), “Pregnancy during the pandemic”, VoxEU.org, 2 June.