点击显示 收起
1 From the Unit for Nutrition Research (IG, BEB, and IT) and the Department of Medicine (RB), LandspitaliUniversity Hospital, Reykjavik, Iceland; the Department of Food Science (IG, BEB, and IT) and the Faculty of Medicine (RB), University of Iceland, Reykjavik, Iceland; and The Icelandic Heart Association (VG and RB), Reykjavik, Iceland.
1 Supported by a research grant from The Icelandic Research Council and Research Fund of the University of Iceland (to IT). The collection of data on adults was supported by the Icelandic Heart Association. 1 Address reprint requests to I Gunnarsdottir, Unit for Nutrition Research, LandspitaliUniversity Hospital, 101 Reykjavik, Iceland. E-mail: ingigun{at}hi.is.
ABSTRACT
Background: Epidemiologic studies suggest a link between fetal and childhood growth and later coronary artery disease (CAD). The influence of adult body size on the relation between birth size and CAD has not been thoroughly studied.
Objective: We investigated the association between birth and adult sizes and CAD within a population with higher birth weight and a lower incidence of and mortality rate from CAD than those seen in other Scandinavian populations.
Design: Fatal or nonfatal CAD was ascertained in 2399 men and 2376 women born in the Greater Reykjavik area between 1914 and 1935. Birth size was obtained from the National Archives. Anthropometric measurements in adults were obtained from the randomized prospective Reykjavik Study.
Results: CAD was inversely related to birth length (P for trend = 0.029) in men but was not significantly related to birth weight or ponderal index (kg/m3). In men who were born short ( 50.5 cm) and who became tall adults (either 175180.5 or > 180.5 cm), the odds ratios (95% CI) for CAD were 1.9 (1.1, 3.1) and 2.2 (1.2, 4.0), respectively, when compared with men in the reference group (those born 52.554.0 cm long). A U-shaped relation between birth size and CAD was found for women.
Conclusions: Size at birth has an effect on CAD, but the effect is modified by adult body size. This confirms that environmental factors operate in both the prenatal and postnatal periods with regard to the development of CAD. The large birth size seen among Icelanders may explain the lower incidence and mortality rate of CAD in Iceland than are seen in other white populations.
Key Words: Birth weight infants coronary artery disease nutrition adults fetal growth retardation Iceland
INTRODUCTION
Some epidemiologic studies have found an association between small size at birth and impaired adult health (1, 2). Thus, among both men and women, low weight or fetal disproportion predicts an increased prevalence of both fatal and nonfatal coronary artery disease (CAD) (35). Poor nutrition has been suggested to be the main adverse environmental influence underlying these associations (2). Fetal adaptations to undernutrition are thought to lead to permanent, adverse changes in common cardiovascular risk factors such as blood pressure and glucose tolerance (6), and these effects seem to be most marked in very small fetuses. Genetics also is reported to play an important role in determining fetal growth and disease (7, 8).
More recent studies have investigated whether postnatal growth is associated with adult disease (4, 9). Small size at birth followed by accelerated growth during childhood has been shown to be associated with an increased risk of CAD and other diseases in adult life (4, 9, 10). It is therefore important to be able to include measurements of weight and height in adulthood in the analysis of the association between size at birth and the occurrence of CAD later in life. Opportunities have been scarce to examine the influence of adult body size on this relation, although measurements of adult weight and height have frequently been available for assessments of the relation between size at birth and other diseases (11, 12).
The Icelandic population is relatively homogeneous genetically (13), which minimizes the genetic influence of the association between birth size and adult diseases. The average birth weight of Icelanders is among the highest in the world (1416), and it could therefore be expected that the association found between size at birth and the occurrence of CAD later in life could be weaker in this population than that found in other populations in which small size at birth is more frequent. The large birth size of Icelanders might also have an effect on the lower CAD incidence and mortality rate than are seen among other white populations (17, 18).
The aim of the present study was to investigate the effect of the association between size at birth and the occurrence of CAD in a genetically homogeneous population with higher birth weight and lower incidence and mortality rate of CAD than those of other Scandinavian populations. The influence of size at birth on the occurrence of CAD and the relation of adult weight and height to that influence were also investigated.
SUBJECTS AND METHODS
Study population
The study included men and women who were born in the Greater Reykjavik area between 1914 and 1935 and who were still living in Reykjavik in 1967 (n = 6120). The study population was a subcohort of the randomly selected participants of the prospective Reykjavik Study of the Icelandic Heart Association. In that study, which began in 1967, information was collected on adult health, diseases, and mortality (19). In the present study, birth data were obtained from midwives original birth records, which were obtained from the National Archives of Iceland. Birth records were found and matched for 4828 persons. Twins (n = 53 persons) were excluded. Thus, in the present study, 4775 singletons, or 78% of the original study population, were included. Approval of the study protocol was granted by the Icelandic National Bioethics Committee and the Data Protection Commission, which provided the birth-size information after written informed consent was obtained from the subcohort of participants in the Reykjavik Study.
Collection of birth data
Midwives birth records included data on sex and birth weight as well as on singleton or multiple births. Guidelines to midwives at the time included instructions on measuring length in cm from crown to heel with the use of a nonelastic tape. From information in midwives birth records, the ponderal index (kg/m3) was calculated.
Collection of adult data
In the longitudinal Reykjavik Study, conducted 19671997, information on adult health was collected at the Icelandic Heart Association Heart Preventive Clinic. The mean (± SD) age at examination was 50 ± 8 y. Each subjects weight and height were recorded. With the use of a questionnaire, information on educational levels and smoking habits was collected; for the latter, subjects were categorized as either former smokers, current smokers, or never smokers.
Since 1981, data regarding the incidence of myocardial infarction have been collected as a part of the World Health Organization MONICA (monitoring trends and determinants in cardiovascular disease) Project (20). Hospital records for those who had myocardial infarction before 1981 were reviewed and evaluated according to the criteria used in the MONICA Project. For a nonfatal event to be considered definite as a myocardial infarction, the records had to show either a progression of Minnesota codes on serial electrocardiograms or cardiac enzyme values twice the normal limit, with either typical symptoms and an electrocardiogram that was not normal or an electrocardiogram progression labeled "probable" and lesser symptoms (21).
Information on death due to CAD was obtained from death certificates from 1967 to 1999 on file with the Statistical Bureau of Iceland by a search for the following codes of the International Classification of Diseases: 19671970 (7th revision), code 420; 19711980 (8th revision), codes 410413; and 19811999 (9th revision), codes 410414. All death certificates were reviewed and coded by an official government pathologist.
Statistical analysis
Mean ± SD values were used to describe the data. Logistic regression was used for calculating odds ratios (ORs) and 95% CIs were used for the birth size categories (22). Birth weight, birth length, and figures for the ponderal index were categorized into quartiles. The third category was used as the reference group: ie, 3.754.25 kg for birth weight, 52.554.0 cm for birth length, and 26.228.3 for ponderal index. Trend tests for birth weight, birth length, and ponderal index were calculated on the basis of the continuous variable. Multiple logistic regression was used to assess the combined effect of size at birth and potential confounding variables such as educational levels, smoking habits, and body mass index (BMI). Kendalls correlation was used to describe the association between size at birth and adult size. Adjustments were made for correlation of the year of birth with birth weight (r = -0.057, P < 0.001) and birth length (r = 0.187, P < 0.001). The t test was used to compare birth size and adult anthropometric data for men and women. Significance was defined as P < 0.05. The statistical software used was SPSS for WINDOWS, version 9 (SPSS Inc, Chicago).
RESULTS
In total, 440 men and 134 women were diagnosed as having CAD. The mean age at myocardial infarction was 61 y for men and 65 y for women. The number of deaths due to CAD during follow-up was 229 in men and 54 in women.
Anthropometric measurements for the subjects at birth and in adulthood are shown in Table 1. Men were heavier and taller than women on both occasions and had a significantly higher adult BMI (P < 0.001), although the ponderal index at birth did not differ significantly between the sexes (P = 0.466).
View this table:
TABLE 1 . Characteristics of the subjects1
ORs for CAD are shown in Table 2. Among men, there was an inverse relation between birth length and the occurrence of CAD (P = 0.029). There was also a trend toward a decreased risk of CAD with greater birth weight, although this trend was not significant (P = 0.095). Ponderal index was not associated with CAD. Among women, the association between size at birth and CAD was U-shaped (Table 2). For birth length, the value for the quadratic term in regression was P = 0.006, that for birth weight was P = 0.031, and that for ponderal index was P = 0.020. Adjustment for adult BMI, educational level, and smoking habits did not affect the results.
View this table:
TABLE 2 . Odds ratios (ORs) for coronary artery diseases (CAD) among men and women by 4 categories of birth weight, birth length, and ponderal index
Further analysis of the relation between birth length and the occurrence of CAD in men showed that the men who were born short and became tall as adults had an association between birth length and CAD (Figure 1). Men were divided into 3 groups according to adult height, and the relation between birth length and CAD was assessed in each height group. Birth length of 52.554.0 cm was used as reference. Among the men who were born 50.5 cm long, the ORs for CAD in those who reached a height of 175180.5 or > 180.5 cm as adults were 1.9 (95% CI: 1.1, 3.1) and 2.2 (1.2, 4.0), respectively, when compared with the men who were born 52.554.0 cm long (OR: 1.0, P < 0.05). The same pattern seen in the men was seen in the women: ie, women who were born short (< 52.5 cm) and who became tall as adults had the highest risk of CAD. However, this trend was not significant. For the group with the greatest birth length, a different pattern was seen, with the highest odds ratio in the group with intermediate adult height (2.8; 95% CI: 1.1, 6.9). If the analysis was confined to mortality from CAD, an inverse relation between birth length and mortality (P = 0.009) was observed in men. Mortality due to CAD was not related to birth weight or ponderal index. No association was found between birth size and mortality due to CAD among women.
FIGURE 1. . Men (n = 2381) were divided into 3 groups according to adult height, and the relation between birth length and coronary artery disease (CAD) was assessed in each height group. Among the men who were born 50.5 cm long, the odds ratios (95% CIs) for CAD in those who reached a height of 175180.5 or > 180.5 cm as adults were 1.9 (1.1, 3.1) and 2.2 (1.2, 4.0), respectively, when compared with the reference group of men who were born 52.554.0 cm long (P < 0.05).
Adult height was also clearly related to CAD. The women diagnosed with CAD were significantly shorter as adults than were the healthy women (162.7 cm compared with 164.8 cm; P = 0.004). The same trend was seen for men, but the difference was not significant (176.8 cm compared with 177.8 cm, P = 0.079). Birth length correlated with adult height (r = 0.173, P < 0.001), and birth weight correlated with adult weight (r = 0.114, P < 0.001). Birth weight was positively associated with adult BMI (r = 0.044, P = 0.002). Adult BMI or weight alone was not associated with an increased risk of CAD.
DISCUSSION
There was an inverse association between birth length and CAD in Icelandic men, independent of smoking habits or educational level. There was also a trend, although it was not statistically significant, toward a lower prevalence of CAD with greater birth weight. The association of birth size and CAD was U-shaped among women in the present study.
The strongest association found in the present study was between CAD and birth length among men. The genetic homogeneity of the studied population implies that the association found is influenced by environmental factors. It could be suggested, however, that length is more influenced by genetics than is birth weight (23, 24), but body length has been increasing during the last century, which suggests that it is also affected by environmental factors (25). Nutrient intake varies as much in the population studied as in the populations of other countries (26). Length at birth is a result of nutritional factors operating throughout gestation rather than at specific times during pregnancy, as the growth in length is more even throughout pregnancy, unlike the growth in fetal weight (27). Important novel findings are those that relate birth length and adult height to the occurrence of CAD. Studies from Finland show that less fetal growth followed by accelerated growth during childhood increased the risk of CAD (4, 9). In a study of men born in Uppsala, Sweden, the highest blood pressures were found in those who had low birth weight but who were tall as adults (28). In the present study, for the first time, men who were born small (< 50.5 cm) and who became tall as adults (> 175 cm) were shown to be more likely to have myocardial infarction or to die of CAD than were other men. The results of the present study indicate that the tallest adult males that are born short might have experienced nutritional inadequacy in utero, which increases the risk of CAD in adulthood, and these results strengthen the current view from findings that postnatal growth as well as fetal growth is important in the development of CAD (4, 9, 10).
The lack of a significant relation between birth weight and CAD in this population may be due to the generally high birth weight of Icelanders. The birth weight in Iceland is significantly higher than that in other Scandinavian populations and is among the highest in the world (1214). Only 1.3% of Icelandic subjects weighed 2500 g at birth. In the present study, 26% of the study population weighed > 4250 g at birth, whereas only 9% of Swedish men born from 1920 to1924 did so (29). It is suggested that the mechanism linking impaired fetal growth and CAD is associated with metabolic risk factors for CAD (6), and some of these associations are steepest at the lower end of the birth-weight spectrum (30, 31), which is almost never seen in the population studied. However, it is now known that poor nutrition in utero may lead to permanent changes even if the effect on fetal growth is small (32).
It is well known that maternal prepregnancy weight is an important determinant of birth weight (3335). The mean BMI of women of childbearing age in Iceland was 24.4 kg/m2 at the time of the study, which shows that our subjects were not born during any crisis or period of possible undernutrition (25). On average, Icelandic women gain 15 kg during pregnancy (36), and weight gain in pregnancy is related to the birth weight of the offspring (16). The almost universal use of cod liver oil by the population under study could also play a part in the high birth weight by its known effect of delaying parturition, which results in larger children (37, 38). Furthermore, the Icelandic diet is high in protein and rich in fish and milk products (37), and such a diet has been found to be related to high birth weight (39, 40).
Other possible explanations for the less significant relation between birth weight and CAD in this study are unlikely. The study includes a satisfactory number of endpoints, and therefore a lack of statistical power due to a relatively small number of cases is not likely to be an explanation. Missing gestational age data might also be suggested to explain the lack of significance, but, overall, it is not likely that the number of premature babies is of any significant importance (41).
The U-shaped association found for women between CAD and birth size could be assumed to be due to maternal diabetes, as in a previous study that described an increased risk of cardiovascular disease-related death in persons with birth weight > 4.3 kg (42). However, in Iceland, the prevalence of gestational diabetes is low (43), and therefore it cannot be the explanation for the U-shaped relation observed.
A strong relation has been found in other populations between low ponderal index and the occurrence of CAD later in life (5). In the present study, the year of birth is correlated with birth weight (positively) and with birth length (negatively). These results mean that Icelandic newborns are getting thinner. In the present study, ponderal index was not associated with fatal or nonfatal myocardial infarction, which indicates that, at the higher end of the birth-weight spectrum, thinness at birth is not a risk factor for CAD. The mean ponderal index in this study was not unlike that in other similar studies (4, 5).
Because the birth weight in Iceland is among the highest in the world, it could be expected from the fetal-origins hypothesis that the prevalence of CAD would be low in this population. In fact, the CAD incidence and mortality rate are lower than those in other countries with white populations (17, 18), which indicates a protective effect of high birth weight. Because many fetal-origin studies have been performed in Finland and the United Kingdom, it is of interest to compare the CAD incidence and mortality rate in Iceland to those in Finland and the United Kingdom. According to the results of the MONICA Project, 19851990, the CAD incidence was 484 per 100 000 per year in Iceland but 700 (range: 533818) per 100 000 per year in 5 cohorts in Finland and the United Kingdom (21). Furthermore, the CAD mortality rate was 167 per 100 000 per year in Iceland and > 300 (range: 270395) per 100 000 per year in 5 cohorts in Finland and the United Kingdom. Iceland also has a lower prevalence of type 2 diabetes than is seen in other Scandinavian countries (44), despite its higher prevalence of obesity (45). Recently, an association was found between low birth weight and length and glucose intolerance in Iceland (46), but this relation was much weaker than that found in other similar studies. As for CAD, it could be that the high birth weight of Icelanders protects them from diseases related to impaired fetal growth. Therefore, it is possible that reducing the number of low-birth-weight infants might result in a lower prevalence of type 2 diabetes worldwide and might lower the incidence and mortality rate of CAD.
A protective effect of increased birth length against CAD was seen, and, for the first time, the importance of adult height was shown, because men who were born short but who became tall as adults were at greatest risk. The large birth size may explain the lower incidence and mortality rate of CAD in Iceland than in other countries with white populations.
The fact that the population studied is genetically homogeneous supports the hypothesis that the association found between size at birth and the occurrence of CAD is not of genetic origin only, and it points to the importance of nutritional factors. The results described in this study support the notions that the environment is an important contributor to CAD morbidity and mortality and that the environment exerts this influence both prenatally and postnatally.
ACKNOWLEDGMENTS
We thank Orn Olafsson, Helgi Sigvaldason, and Nikulas Sigfusson for their valuable help with data preparation and calculations and the employees at the National Archives of Iceland and the Genetical Committee at the University of Iceland for their assistance in the preparation of the project and with data collection.
REFERENCES