Glucose tolerance at age 58 and the decline of glucose tolerance in comparison with age 50 in people prenatally exposed to the Dutch famine
- First Online:
- Cite this article as:
- de Rooij, S.R., Painter, R.C., Roseboom, T.J. et al. Diabetologia (2006) 49: 637. doi:10.1007/s00125-005-0136-9
- 1k Downloads
People who were small at birth have an increased risk of type 2 diabetes in later life. People who were in utero during the Dutch famine had decreased glucose tolerance and raised insulin concentrations at age 50. We aimed to evaluate whether prenatal famine exposure leads to more rapid progression of impaired glucose/insulin homeostasis with increasing age.
We performed an OGTT in 702 men and women at age 50 and in 699 men and women at age 58, all born as term singletons immediately before, during or after the 1944–1945 Dutch famine.
People who had been exposed to famine in utero had significantly higher 120-min glucose concentrations at age 58 compared with people who had not been exposed to famine (difference=0.4 mmol/l, 95% CI 0.1 to 0.7, adjusted for sex and BMI). Glucose tolerance deteriorated between the age of 50 and 58. The unadjusted 120-min glucose concentrations rose by 0.2 mmol/l (95% CI 0.0 to 0.4), while 120-min insulin concentrations had increased by 64 pmol/l (95% CI 48 to 82). There were no differences in the rates of glucose and insulin level increase between the famine-exposed group and the unexposed group (p=0.28 for the difference in increase in glucose concentrations and p=0.09 for insulin concentrations).
Although we confirmed that undernutrition during gestation is linked to decreased glucose tolerance, the effect does not seem to become more pronounced at age 58 as compared with age 50.
KeywordsGlucose tolerance Progression Prenatal famine
People who were small at birth have an increased risk of type 2 diabetes in later life [1, 2, 3, 4, 5, 6, 7]. One interpretation of this association is that an adverse intrauterine environment permanently impairs glucose homeostasis, primarily by inducing insulin resistance [8, 9]. This interpretation is strongly supported by animal studies, which have consistently shown that experimental undernutrition during gestation leads to impaired glucose homeostasis . The Dutch famine was a period of extreme food shortage in the west of the Netherlands that occurred during the last 5–6 months of World War II. The famine offers a unique opportunity to study the effects of prenatal undernutrition on health in later life. In 1998, we found the first direct evidence in humans that exposure to undernutrition during gestation reduced glucose tolerance in later life .
The prevalence of type 2 diabetes and IGT increases with age, mainly as a result of the increase in BMI, insulin resistance and a fall in insulin secretion [12, 13, 14]. It is not yet known whether the age-related decline in glucose tolerance is increased in people who experienced an adverse intrauterine environment. In animal experiments the effects of prenatal undernutrition on glucose tolerance have been shown to increase with age. Rats that were undernourished during gestation showed a decrease in glucose tolerance between 3 and both 12 and 15 months of age [15, 16]. In humans the prenatal influences on blood pressure have been shown to increase with age [17, 18]. The present study was designed to assess glucose tolerance in the famine cohort at age 58 and to evaluate whether the previously demonstrated effect of prenatal famine exposure on glucose tolerance in the Dutch famine birth cohort progresses with ageing.
Subjects and methods
Exposure to famine
We defined the famine period based on the official daily rations for the general population older than 21 years. These rations were about 7,560 kJ/day in December 1943 and gradually decreased to about 5,880 kJ in October 1944. On 26 November 1944 the rations fell below 4,200 kJ and after 12 May 1945 they rose above 4,200 kJ again. In June 1945 rations were over 8,400 kJ. The rations during the famine did not apply to children younger than 1 year, who were relatively protected. Their official daily rations were always higher than 4,200 kJ, which is adequate according to The Oxford Nutrition Survey . We considered foetuses to be exposed to famine if the average daily rations of the mother during any 13-week period of gestation were below 4,200 kJ. Babies born between 7 January 1945 and 8 December 1945 were thus exposed. We defined periods of 16 weeks each to differentiate between those who had been exposed in late gestation (born between 7 January and 28 April 1945), in mid-gestation (born between 29 April and 18 August 1945) and in early gestation (born between 19 August and 8 December 1945). Babies born before 7 January 1945 and babies conceived and born after 8 December 1945 were considered as unexposed to famine in utero and acted as a control group.
Birth measurements and information about the health and status of the mother were taken from the medical birth records . Trained nurses carried out the medical examinations and the interview. Methods for OGTT and anthropometry at age 50 have been described in detail elsewhere . At age 58, a standard 75-g OGTT was performed after an overnight fast. People with pre-existent diabetes, defined as taking oral or injected glucose-lowering medication, were excluded from the test. Plasma glucose concentrations were measured by a standardised enzymatic photometric assay on a Modular P Analyzer (Roche, Basel, Switzerland) and plasma insulin concentrations by an immunoluminometric assay on an Immulite 2000 Analyzer (Diagnostic Product Corporation, Los Angeles, CA, USA). We measured height with a fixed or portable stadiometer, weight with SECA and portable Tefal scales and waist and hip circumferences with a flexible tape measure. We interviewed all participants and asked them about their socio-economic status, their medical history, lifestyle and use of medication, using standardised questions. Current socio-economic status was coded according to ISEI-92, which is based on the person’s, or their partner’s occupation, whichever status was highest.
In order to make the results comparable with those of our previous study, we defined IGT as a 120-min glucose concentration of 7.8–11.0 mmol/l and type 2 diabetes as a 120-min glucose concentration of >11.0 mmol/l. Logarithmic transformations were applied to glucose, insulin and BMI values, because they had skewed distributions. We used linear regression analysis to determine the effect of prenatal exposure to famine at different stages of gestation on glucose tolerance at age 58. To detect a possible age-related deterioration in glucose tolerance, we only included data of subjects participating at age 50 as well as at age 58, using repeated-measures analysis. We first calculated differences between the exposed and unexposed groups. Then differences were calculated between unexposed subjects and subjects prenatally exposed to famine in late, mid- or early gestation. We adjusted for maternal and birth characteristics, sex, BMI, smoking and socio-economic status. For the last three variables, measurements at age 50 and age 58 were both included in the repeated-measures analysis.
Study group characteristics
Maternal, birth and adult characteristics according to timing of prenatal exposure to the Dutch famine
Exposure to famine
In late gestation
In early gestation
Proportion of men
Age at delivery (years)
Proportion of primiparous women
Proportion doing manual labour
Weight gain 3rd trimester (kg)
Weight at last antenatal visit (kg)
Gestational age (days)
Birth weight (g)
Birth length (cm)
Head circumference (cm)
Ponderal index (kg/m3)
Placental area (cm2)
Current smoker (%)
Current SES (ISEI-92)
Proportion of diabetic patients excluded from OGTT
Famine exposure and plasma glucose and insulin concentrations at age 58
Means of plasma glucose and insulin concentrations and prevalences of IGT and type 2 diabetes according to timing of prenatal exposure to the Dutch famine
Exposure to famine
In late gestation
In early gestation
Prevalence of diabetes based on OGTT (%)
Prevalence of IGT based on OGTT (%)
Effects of mother’s weight and birthweight
The weight of the mother at the last prenatal visit, birthweight, birth length and head circumference were all inversely related to 120-min glucose and insulin concentrations. Adjusted for sex and BMI the 120-min glucose concentrations increased by 0.8% (95% CI 0.5 to 1.1) per kilogram decrease in mother’s last weight, 13.2% (95% CI 7.4 to 19.2) with each kilogram decrease in birthweight, 2.6% (95% CI 1.4 to 3.9) with each centimetre decrease in birth length and 1.9% (95% CI 0.3 to 3.7) with each centimetre decrease in head circumference, all adjusted for sex and adult BMI. The 120-min insulin concentrations increased by 17 pmol/l (95% CI 10 to 23) per kilogram decrease in the mother’s last weight. Additional adjusting for birthweight showed that the effect of prenatal exposure to famine on 120-min glucose concentrations was larger than could be explained by the famine-related differences in birthweight. After adjustment for sex, BMI and birthweight, 120-min glucose concentrations were 0.2 mmol/l (95% CI −0.1 to 0.6) higher among the exposed than among unexposed people. Adjusting for other possible confounding variables including maternal age at delivery, current smoking status, current socio-economic status or alcohol consumption had little effect on the association between famine exposure and 120-min glucose concentrations. The prevalence of IGT and type 2 diabetes based on the OGTT did not differ significantly between the exposed and the unexposed people.
Effects of age
Progression of 120-min glucose and insulin levels and BMI between age 50 and age 58, according to timing of prenatal exposure to the Dutch famine
Exposure to famine
Born before (CI)c
In late gestation (CI)c
In mid-gestation (CI)c
In early gestation (CI)c
Conceived after (CI)c
Proportion of men (%)
120-min glucose (mmol/l)a
0.3 (0.0 to 0.6)
−0.1 (−0.6 to 0.2)
−0.1 (−0.5 to 0.3)
0.2 (−0.4 to 0.8)
−0.1 (−0.4 to 0.3)
0.1 (−0.1 to 0.2)
120-min insulin (pmol/l)a
71 (48 to 92)
35 (−1 to 66)
33 (−9 to 7)
54 (−2 to 98)
47 (21 to 70)
51 (38 to 64)
1.1 (0.5 to 1.6)
0.8 (−3.0 to 4.1)
1.3 (−0.3 to 2.8)
0.9 (−0.5 to 2.6)
1.3 (0.7 to 1.9)
1.1 (0.9 to 1.3)
In this unique cohort study of subjects born before, during and after the Dutch famine, we found that undernutrition during gestation was associated with reduced glucose tolerance and raised insulin concentrations at age 58. This confirms our previous findings at age 50 . We found that both 120-min glucose and insulin concentrations were higher in people exposed to famine at any stage of foetal development than in unexposed people. Importantly, this effect was larger than could be explained by the lower birthweight of babies born during the famine and by the low weight gain of their mothers. As with our previous study, this association was independent of people’s current BMI .
This is a population-based study of men and women recruited from the original famine cohort and still living in the Netherlands. At this age a significant number of individuals have overt type 2 diabetes and related diseases, especially cardiovascular disease. Selective participation of the more healthy subjects in our cohort could have influenced our results. Although we maximised response rates by visiting and studying subjects in their homes if necessary, people with type 2 diabetes, cardiovascular disease, high cholesterol levels and high blood pressure were less likely to attend the survey clinic (data not shown). However, there was no evidence of a difference in response rates between those who were famine-exposed or non-exposed. Consequently, we do not think that selective participation has affected the relative differences in glucose and insulin levels among the exposed and the unexposed groups. It could, however, have affected our power to detect effects of famine on type 2 diabetes.
Glucose tolerance deteriorated between the age of 50 and 58. A large part of this decline could be attributed to an increase in BMI. The decline in glucose tolerance was not more marked in the famine-exposed groups than in the control groups (Table 3). People who were born before the famine had the relatively largest increase in 120-min glucose and insulin concentrations. A possible explanation is that these people experienced the famine during their first year of life, which may have affected their postnatal growth pattern. Reduced early growth has been shown to be associated with later IGT and type 2 diabetes [20, 21]. Although infants were relatively protected during the famine, their official daily rations always being higher than 4,200 kJ, we do not know whether the famine affected their growth . Early mortality rates were highest for people born before the famine and the deaths were mainly related to undernutrition or infections, indicating that growth was probably also affected .
We found that people with low birthweight had a greater age-related progression of glucose intolerance. A larger age-related progression of raised blood pressure has already been demonstrated for people with low birthweight, but to the best of our knowledge this is the first direct evidence of an age-related amplification of the effect of reduced foetal growth on glucose tolerance in humans [17, 18].
The effect of prenatal exposure to famine on glucose tolerance could be mediated through a number of mechanisms. Animal experiments indicate that undernutrition during gestation affects the development of the pancreas, which leads to an impaired function of the beta cell and consequently insulin deficiency . In contrast, most human evidence points to the importance of insulin resistance. People who were thin at birth or had low birthweights were shown to be more insulin resistant than adults [24, 25]. The raised insulin concentrations that we found in famine-exposed individuals (Table 2) are consistent with insulin resistance acting as a mediator of the effect of famine exposure on glucose tolerance. Increasing evidence suggests that skeletal muscle is a key site for programming of insulin resistance. Muscle is a major site of glucose uptake and associations between a low ponderal index and altered metabolism of adult skeletal muscle have been found [26, 27]. Recently, an association between low birthweight and specific changes in muscle insulin-signalling protein expression was found . Another mechanism that could explain the association between prenatal exposure to famine and later glucose intolerance involves the hypothalamic-pituitary-adrenal (HPA) axis. Maternal sheep and guinea pigs that were undernourished in late gestation gave birth to offspring in which HPA function was altered in adult life [29, 30]. In humans, low birthweight is associated with elevated basal plasma cortisol concentrations and increased adrenocortical responsiveness to adrenocorticotrophin at adult age [31, 32]. It has been hypothesised that undernutrition during gestation alters the setpoint of the HPA axis resulting in an increased activity and consequently an increased secretion of glucocorticoids, which is associated with glucose intolerance and insulin resistance [33, 34].
In summary, although we confirmed that poor nutrition in utero is related to decreased glucose tolerance in later life, the effect of prenatal exposure to the Dutch famine does not seem to become more pronounced at age 58 as compared with age 50. Progression of glucose intolerance was found to be related to famine exposure during the first year of life and to low birthweight. This suggests that prenatal famine exposure, famine exposure during the first year of life and low birthweight may contribute to glucose intolerance by different mediating mechanisms.
This study was funded by the Netherlands Heart Foundation (Grant Number 2001B087), the Academic Medical Centre (Amsterdam, the Netherlands) and the Medical Research Council (UK). We thank the participants for their willing cooperation.