点击显示 收起
1 From the Departments of Clinical Epidemiology and Biostatistics (RCP, SRdR, PMB, and TJR), Cardiology (TAS), and Obstetrics and Gynecology (OPB), Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands, and the MRC Epidemiology Resource Centre (CO) and the Developmental Origins of Adult Disease Centre (DJB), University of Southampton, Southampton, United Kingdom
2 The Dutch Famine Birth Cohort Study is funded by the Diabetes Fonds (Netherlands), the Netherlands Heart Foundation (grant number 2001B087), Wellbeing (United Kingdom), the Medical Research Council (United Kingdom), and the Academic Medical Centre (Netherlands).
3 Address reprint requests to RC Painter, PO Box 22660, 1100 DD, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands. E-mail: r.c.painter{at}amc.uva.nl.
See corresponding editorial on page 271.
ABSTRACT
Background: Limited evidence suggests that maternal undernutrition at the time of conception is associated with increased cardiovascular disease risk in adult offspring.
Objective: We investigated whether persons conceived during the Dutch famine of World War II had an early onset of coronary artery disease (CAD).
Design: We compared the age at onset and cumulative incidence of CAD between persons born as term singletons who were exposed to the 1944–1945 Dutch famine during late (n = 160), mid- (n = 138), or early (n = 87) gestation and 590 unexposed subjects at age 50 or 58 y. Age at CAD onset was defined as the age at which angina pectoris was identified (according to the Rose questionnaire), Q waves were observed on an electrocardiogram (Minnesota codes 1–1 or 1–2), or coronary revascularization was performed (by angioplasty or bypass surgery).
Results: Of the 83 CAD cases identified, persons conceived during the famine were 3 y younger than the unexposed persons at the time of CAD diagnosis (47 y compared with 50 y) and had a higher cumulative incidence of CAD [13%; hazard ratio (HR) adjusted for sex: 1.9; 95% CI: 1.0, 3.8] than did the unexposed persons. The HR changed little after adjustment for smoking (HR: 1.8), social class (HR: 2.0), or size at birth (HR: 2.0).
Conclusions: We found an earlier onset of CAD among persons conceived during the famine, which suggests that maternal nutrition in early gestation may play a role in the onset of CAD. This finding agrees with evidence from animal experiments that identify periconceptional maternal diet as important in the offspring's adult health.
Key Words: Coronary artery disease • age at onset • maternal nutrition • maternal starvation • pregnancy • cardiovascular programming
INTRODUCTION
Restricted intrauterine growth has been proposed as an important contributor to later coronary artery disease (CAD) and its biological risk factors (1). Developing organ systems respond negatively to the reduced availability of nutrients, particularly during periods of rapid development—so-called critical periods (2).
Most studies in humans have access only to indirect measures of intrauterine nutrition, such as birth weight. Substantial changes in cardiovascular function can result from maternal or fetal undernutrition without affecting birth weight (3). To gain more insight into the mechanisms of disease in later life in humans after restricted prenatal nutrition, the sequelae of restricted maternal nutrition during gestation have been studied in the Leningrad Siege Study (4) and the Dutch Famine Birth Cohort Study. The Dutch famine was a 5-mo period of extreme food shortage during the winter of 1944–1945 in World War II. The Leningrad Study reported no effect of maternal malnutrition on the adult offspring's CAD prevalence. The Dutch famine, however, was relatively short compared with the Leningrad Siege Study, which allowed the effects to be studied by trimester of prenatal famine exposure. The previous findings from the Dutch Famine Birth Cohort Study support the hypothesis that the timing of the nutritional insult is important in determining its effect in later life; exposure to the Dutch famine in late gestation was associated with decreased glucose tolerance (5), whereas more microalbuminuria (6) was present among subjects exposed during midgestation. The most marked effects were described in the group of subjects conceived during the famine and include a more atherogenic lipid profile (7), altered clotting (8), more obesity (9), and a tripling of CAD prevalence at age 50 y (10).
The cluster of cardiovascular disease risk factors previously described in persons conceived in famine is in line with studies in animals, which have highlighted the importance of periconceptional maternal nutrition in programming cardiovascular disease risk (11–14). The effects of maternal periconceptional diet on the course of adult disease have not been investigated. We hypothesized that CAD manifests at an earlier age in persons exposed to famine during early gestation. We reexamined the findings of a study conducted at age 50 y and included information from a subsequent study 8 y later.
SUBJECTS AND METHODS
Selection procedure
The Dutch Famine Birth Cohort consists of 2414 live-born term singletons born in the Wilhelmina Gasthuis in Amsterdam, Netherlands. All infants were born between 1 November 1943 and 28 February 1947. The selection procedure for the study conducted at age 50 y was described in detail elsewhere (5), as was loss to follow-up because of mortality, emigration, and other reasons (15, 16). In short, cohort members were eligible for participation if they were living in the Netherlands at the start of the study (January 1995 and September 2002), and their address was known to the Dutch Famine Birth Cohort Study researchers. All eligible subjects were asked to participate at ages 50 and 58 y. Council registries helped trace people who had had a change of address since they were last traced at age 50 y. All participants provided written informed consent. The local Medical Ethics Committee approved the study. The study conformed to the Declaration of Helsinki.
Exposure to famine
We defined famine exposure according to the daily official food rations for adults. In addition to the official rations, food from other sources, such as church organizations, central kitchens, and the "black market," was also available and the people may have had access to up to double the rationed amount at the peak of the famine. The rations do, however, adequately reflect the fluctuation of food availability during the famine (17). A person was considered prenatally exposed to famine if the average daily rations for adults during any 13-wk period of gestation were <1000 kcal. Therefore, persons born between 7 January 1945 and 8 December 1945 were considered exposed prenatally to famine. Cohort members born between 1 November 1943 and 6 January 1945 (born before the famine) and between 9 December 1945 and 28 February 1947 (conceived after the famine) were unexposed to famine. We defined periods of 16 wk each to differentiate between those who were exposed in late gestation (born between 7 January and 28 April 1945), midgestation (born between 29 April and 18 August 1945), and early gestation (born between 19 August and 8 December 1945), in correspondence with previous publications on this cohort (5, 10). Persons exposed in early gestation were conceived during the famine. The famine ended in May 1945, with the advance of the allied armies into Holland. Food supplies were rapidly restored, and the average caloric intake in June 1945 was >2000 kcal.
Data collection
Medical birth records provided information about the mother, the course of gestation, and the size of the infant and the placenta at birth (5). Socioeconomic status (SES) at birth was defined according to the occupation of the head of the family and was classified as either manual or nonmanual on the basis of the information provided by the birth records.
Consenting cohort members came to the hospital. We measured height using a fixed or a portable stadiometer, weight with Seca scales (Hamburg, Germany) or Tefal portable scales (Groupe SEB Nederland BV, Veenendaal, Netherlands). Body mass index was calculated by dividing weight in kilograms by the square of height in meters. Blood pressure was measured twice on 2 occasions (morning and afternoon) with an automated device: a Profilomat (Disentronic Medical Systems AG, Burgdorf, Switzerland) at age 50 y and an Omron 705CP/IT (Omron Healthcare United Kingdom, West Sussex, United Kingdom) at age 58 y. Mean blood pressure was calculated from both the morning and afternoon measurements. Standard 12-lead electrocardiograms (ECGs) were used for all participants. Trained technicians blinded to the clinical data scored the ECGs according to the Minnesota criteria. Nondiabetic participants underwent standard 75-g oral glucose tolerance testing. Blood was drawn for the measurement of LDL, HDL, and triacylglycerol concentrations. Total cholesterol, HDL, and triacylglycerol concentrations were measured with the use of an enzymatic colorimetric reagent (Roche Diagnostics, Switzerland) on a P-800 Modular (Roche, Switzerland). LDL was calculated by using the Friedewald formula.
Participants were interviewed to obtain information about their medical history, including operations, lifestyle, and use of medication. We defined current SES according to the participant's or their partner's occupation, whichever was highest, using the ISEI-92 (18). The ISEI-92 scale ranges from 16 (minimum score; lowest status) to 87 (maximum score; highest status). Trained nurses carried out all measurements and interviews.
The presence of CAD was defined as the presence of one or more of the following: angina pectoris according to the Rose/World Health Organization questionnaire, Q waves on the ECG (Minnesota codes 1–1 or 1–2), or history of coronary revascularization (angioplasty or bypass surgery).
Statistical methods
For the investigation of age at onset of CAD, all subjects that had participated at age 50 or 58 y were included. To study associations between the progression of CAD and the timing of famine exposure during gestation and size at birth, we used the Cox regression model of the cumulative incidence and age of manifestation of CAD and calculated hazard ratios (HRs) and 95% CIs for subjects exposed in late, mid-, and early gestation and compared them with unexposed subjects. We constructed a Kaplan-Meier curve showing the cumulative incidence of CAD as a function of age per famine exposure group.
The time of event was defined as the age at onset of angina pectoris according to the Rose/World Health Organization questionnaire. If no age at onset of angina pectoris was stated, the age at the time of the first coronary revascularization procedure was used, and, in cases where both ages were missing, the age at the time of registration of Q waves on the ECG was used. Subjects who had only participated at age 50 y were censored at the age at that visit. When adjusting for covariates in the Cox model, we used the most recently collected available measurement before the event. If the event had occurred between the time points of participation, an estimation of the covariate at the time of the event was made with the use of linear interpolation.
We used logistic regression analysis to compare the characteristics of persons with and without CAD. Because of the left skewed distribution of age at first occurrence of CAD, this variable is reported in means after we applied a quadratic transformation. Body mass index, SES, the ratio of LDL to HDL, and glucose were log transformed because of their skewed distributions. These variables are reported as geometric means ± SDs; all other variables are reported as means ± SDs. All statistical analyses were performed by using SPPS 12.0.2 (SPSS Inc, Chicago, IL). We considered differences to be statistically significant if P values were <0.05.
RESULTS
Study population
The cohort contained 2414 members. Loss to follow-up was described previously (15, 16). At age 50 y, 1527 (63%) persons were eligible for participation. At age 58 y, 1423 (59%) persons were eligible for participation. A total of 975 subjects participated in this study. At age 50 y (range: 48–53 y), 736 persons participated, of whom 491 participated again at age 58 y. At age 58 y (range: 56–61 y), 732 persons participated in the study, 239 of whom had not participated at age 50 y. The participation rates among those exposed to famine (49%) and among those not exposed to famine (40%) in utero did not differ significantly (P = 0.7). The birth weights of persons included in the study (3353 g) did not differ significantly from the birth weights of those not included in the study (3341 g; P = 0.6).
Infants born after exposure to famine in late and midgestation were lighter and smaller than the unexposed infants, and their mothers weighed less at the end of gestation (Table 1).
View this table:
TABLE 1. Maternal, birth, and coronary artery disease characteristics of men and women who participated in the Dutch Famine Birth Cohort Study at age 50 or 58 y
Age at onset of coronary artery disease
A total of 83 subjects had developed CAD by the end of follow-up. We found an overall cumulative incidence of CAD of 9% in men and 8% in women. Persons exposed to famine in early gestation had the highest cumulative incidence of CAD (13%; sex-adjusted HR compared with unexposed persons: 1.9; 95% CI: 1.0, 3.8) (Table 1). The cumulative incidence of CAD in those exposed to famine in late gestation (sex-adjusted HR: 0.8; 95% CI 0.4 to 1.6) and midgestation (sex-adjusted HR: 1.1; 95% CI: 0.6, 2.1) did not differ significantly from that of those unexposed to famine (8% after exposure in mid- and late gestation compared with 8% in unexposed persons). On average, CAD manifested 3 y earlier in those exposed to famine in early gestation (mean age: 47 y; interquartile range: 45–51 y) than in those unexposed to famine (mean age: 50 y; interquartile range: 45–57 y) (Figure 1).
FIGURE 1.. Kaplan-Meier curve of the cumulative incidence of coronary artery disease (CAD) in persons born before the famine (n = 24); exposed to famine in late (n = 12), mid- (n = 11), or early (n = 11) gestation; or conceived after the famine (n = 25). The cumulative incidence of CAD was significantly greater in persons exposed to famine in early gestation than in those born before or conceived after the famine, P < 0.05 (Cox regression).
Men and women with CAD were lighter (3275 compared with 3360 g) and thinner (25.9 compared with 26.3 kg/m3) at birth and had a smaller head circumference (32.5 compared with 32.8 cm) at birth, although none of these differences was significant. After size at birth was adjusted for, the association between famine exposure in early gestation and CAD was not attenuated (HR: 2.0; 95% CI: 1.0, 3.8).
Coronary artery disease risk factors
The distribution of cardiovascular disease risk factors according to famine exposure during various stages of gestation among subjects at age 58 y is shown in Table 2.
View this table:
TABLE 2. Characteristics of men and women who participated in the Dutch Famine Birth Cohort Study at age 58 y1
In addition to the results shown in Table 2, famine exposure during any period of gestation was associated with elevated glucose concentrations at 120 min (P = 0.04; adjusted for sex and body mass index) and an elevated ratio of LDL to HDL (P = 0.03; adjusted for sex). Adjustment for the 2 social risk factors in Table 2, smoking (adjusted HR: 1.8; 95% CI: 0.9, 3.5) and low SES (adjusted HR: 2.0; 95% CI: 1.0, 3.8), had little effect on the association between famine exposure in early gestation and CAD.
Maternal constitution and fertility
There were no significant differences in maternal weight, age, parity, or SES at birth between persons with or without CAD. When these variables were entered into a multivariable Cox model, the association between exposure to famine in early gestation and CAD was little changed (multivariable-adjusted HR: 1.8; 95% CI: 0.9, 3.6).
DISCUSSION
We found that the risk of CAD before the age of 61 y in persons conceived during the Dutch famine was double that of unexposed persons. This association was independent of size at birth and of smoking and low SES. Of the 83 persons with CAD, those who were conceived during the famine were 3 y younger at diagnosis. Ours was the first study to describe the course of CAD in the offspring of mothers nutritionally deprived during early gestation.
Women were less fertile during the famine (19). Those who did conceive may have been of a different constitution. However, the correction for markers of maternal constitution or fertility, including maternal weight, age, parity, and SES, did not change the association of prenatal famine exposure with CAD.
Selective participation of persons who were fit enough to attend the clinic and prior excess mortality among the most seriously affected persons may have led to an underestimation of the effect of prenatal famine on subsequent CAD progression. However, we believe that the estimate reported in this article is relatively accurate, because analyses of the prevalence of angina pectoris and history of coronary revascularization surgery among persons who were not able to visit the clinic, but who agreed to a home or telephone interview, yielded results in the same direction (RC Painter, SR de Rooij, and TJ Roseboom, unpublished observations, 2005). Moreover, there was no excess all cause or CAD mortality among people conceived in the famine (16).
Although not statistically significant, persons with CAD were also lighter at birth than were persons without CAD. This finding agreed with results from other studies (1, 20).
Suboptimal intrauterine growth has been described to have programming effects on many cardiovascular disease risk factors, including hypertension (21), impaired glucose tolerance (22, 23), and lipid metabolism (24). Consistent with our previous study of the Dutch Famine Birth Cohort, persons conceived during the famine had higher plasma glucose concentration at 120 min and higher ratios of LDL to HDL cholesterol than did persons who had not been exposed to famine in utero. It is possible that the effects of famine on CAD are mediated through these 2 biological risk factors. It was not possible for us to explore the effect of these risk factors on CAD incidence because, for many subjects, we did not have measurements from before the onset of disease. Moreover, many of the subjects were being treated for type 2 diabetes or hypercholesterolemia.
There are many possible processes by which persons conceived in famine could have increased rates of CAD. Slow intrauterine growth has been shown to be associated with hormonal axis programming (25, 26), alterations in cardiovascular control mechanisms (11, 12, 27), altered myocardial structure (28), endothelial dysfunction (29), and accelerated atherogenesis (30). In future studies we hope to elucidate the role of these factors in the pathophysiology of coronary artery disease after prenatal famine exposure.
Persons conceived in famine not only had a higher cumulative incidence of CAD, but the disease occurred at an earlier age. Models in which animals were prenatally nutrient restricted had premature aging (31) and more rapid age-related progression of the biological risk factors of CAD (32, 33). There is some evidence of an association between low birth weight and increased aging rates in human studies too (34, 35). Although little research has been carried out elucidating the underlying mechanisms, Jennings et al (36) suggest that telomere shortening induced by prenatal undernutrition may be responsible for the premature senescence of tissues such as the liver and kidney. These studies also pointed out that catch-up growth, such as that which may have occurred in fetuses conceived during famine but exposed to adequate nutrition during the remainder of gestation, could result in further telomere shortening.
In summary, our findings suggest that maternal nutrition in early gestation may play an important role in the course of CAD. This suggestion is in line with evidence from animal experiments that identified preconceptional and preimplantation maternal diet as important for the offspring's adult health (11–14).
ACKNOWLEDGMENTS
We are grateful for the willing cooperation of all participants.
OPB, TJR, PMB, DJB, and CO conceived of and planned the study. RCP, TJR, SRdR, and TAS carried out the study. RCP, CO, and DJB performed the statistical analyses. All authors critically discussed the results. None of the authors had a conflict of interest.
REFERENCES
Related articles in AJCN: