from the Helsinki Birth Cohort Study

Symposium doi: 10.1111/j.1365-2796.2007.01798.x Epidemiology, genes and the environment: lessons learned from the Helsinki Birth Cohort Study J. G. Eriksson From the Department of Public Health, University of Helsinki, Helsinki, Finland Abstract. Eriksson JG (University of Helsinki, Finland). Epidemiology, genes and the environment: lessons learned from the Helsinki Birth Cohort Study (HBCS) (Symposium). J Intern Med 2007; 261: 418 425. Nonoptimal growth during fetal life and infancy is associated with an increased risk of coronary heart disease and type 2 diabetes later in life. This early pattern of growth is associated with an increased disease risk especially when followed by a relative gain in body size later in childhood. Genetic factors are closely involved in growth and disease pathogenesis and gene-early life environmental interactions will be described affecting adult health outcomes. This overview will primarily focus upon findings from the Helsinki Birth Cohort Study consisting of 15846 subjects born 1924 1944 on whom growth data and information on adult health are available. Keywords: coronary heart disease, fetal growth, genes, infancy, interactions, obesity, type 2 diabetes. Introduction A large number of epidemiological studies from different geographical regions have established that there is a strong relationship between a small body size at birth and later health. People who were small at birth have an increased risk of developing e.g. coronary heart disease (CHD) and type 2 diabetes later in life. Not only a small body size at birth, but also nonoptimal growth during infancy is associated with an increased risk for the aforementioned diseases [1 43]. Fetal growth and CHD David Barker was one of the pioneers within the DOHaD field by suggesting that the epidemic of CHD in Western societies might have its origin in fetal life [1, 2]. In fact Barker and his study group was able to show that a small birth size was associated with higher death rates from CHD among both men and women [2, 6]. A study including 15 726 people born from 1911 to 1930 in Hertfordshire, UK, showed that death rates from CHD fell with increasing birth weight. This increased CHD risk associated with a small body size at birth was a consequence of growth restriction rather than prematurity, supporting the view that maternal and fetal undernutrition would be an important underlying factor for CHD [2, 6]. These original study findings were subject to criticism and a high degree of scepticism within the medical field. However, within a few years these findings were supported by similar observations in other populations and by other investigators [15 19, 24]. Strong support came from the US were the investigators conducting a study including 70 297 nurses found a twofold rise in the risk of CHD across the birthweight range [20]. Not only a small birth weight but also other measures of impaired intrauterine growth e.g. short length and low ponderal index at birth have been linked with an increased risk for CHD. Today a large number of studies point towards the importance of events during critical periods of growth and development in the pathogenesis of CHD and its risk factors [1 43]. 418 ª 2007 Blackwell Publishing Ltd

Although a small birth size, due to restricted growth, has originally been associated with unfavourable long-term health outcomes, the effect of programming is not necessarily affecting birth size. Findings from the Dutch Hunger Winter study show that programming or induction of a risk factor or disease can occur without manifest growth failure [23, 27, 41 43]. Infant growth and CHD In several aspects infant growth can be seen as a continuation of fetal growth and consequently a strong relationship between birth weight and weight during infancy exists. In fact growth during infancy seems to be of at least equal importance as prenatal growth in relation to later health outcomes. The pioneer work was once again done in the UK. Findings from men belonging to the Hertfordshire cohort proved that weight at 1 year was strongly and inversely associated with CHD [2]. Among men, born in Hertfordshire, UK, death rates from CHD doubled across the range of weight (11.8 8.2 kg) at 1 year of age. Helsinki Birth Cohort Study Using the unique resources of the Helsinki Birth Cohort Study (HBCS) we have been able to confirm several of these previously cited findings in relation to early growth and later adult health outcomes [19, 24 26, 28, 30, 31, 34, 39]. Two cohorts including 15 846 subjects born at Helsinki University Central Hospital and who grew up in the Helsinki area have been studied. Data from the older cohort (n ¼ 7086) born 1924 1933, includes birth characteristics and information on growth between 7 and 15 years of age [24]. The younger cohort, born 1934 1944, includes 8760 subjects. This cohort has data on birth characteristics as well as detailed data on childhood growth from birth to 12 years of age [28]. Information on growth and childhood characteristics has been abstracted from birth, child welfare and school health care records. From the year 1971 onwards these cohorts have been followed up by register linkage to national Finnish registers using the unique social security number assigned to all Finnish citizens. Furthermore, 2500 individuals from these cohorts have participated in a clinical study which has provided detailed information on metabolic and genetic aspects in relation to patterns of growth and adult health outcomes [36, 39]. A small body size at 1 year of age adds to the CHD risk independently of size at birth [28]. In fact body size at 1 year of age is a strong and independent predictor of later CHD (Table 1). The combination of poor prenatal growth and poor growth during infancy seems to be associated with the highest CHD morbidity and mortality in adult life. Interestingly, within the Hertfordshire study cohort low weight at 1 year has been associated with an atherogenic lipid profile [44, 45]. Similar associations between weight at 1 year of age and adverse adult lipid profiles have also been observed in HBCS. This finding would offer a plausible explanation for the epidemiological associations found between nonoptimal growth during infancy and the increased risk for CHD in later life. It has been argued that most published data on the relationship between infant growth and adult health outcomes are observational and retrospective making interpretations difficult and providing an insecure basis for clinical practice. Due to the natural history of CHD this is inevitably the case. It has even been proposed that a high-nutrient diet in infancy adversely programs the main components of the metabolic syndrome thereby suggesting that slower infant growth and relative undernutrition benefit later cardiovascular health [46, 47]. In some species accelerated early Table 1 Hazard ratios for coronary heart disease in the Helsinki Birth Cohort Study among Finnish men born 1934 1944 according to BMI at 1 year of age BMI at 1 year (kg m )2 ) Hazard ratio for CHD (95% CI) 16 1.83 (1.28 2.60) )17 1.61 (1.15 2.25) )18 1.29 (0.91 1.81) )19 1.12 (0.77 1.62) >19 1.00 P-value for trend 0.0004 BMI, body mass index. ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425 419

growth has short-term benefits but adverse long-term consequences. In humans these arguments and interpretations have largely been based upon studies performed in preterm infants followed up and studied during childhood and adolescence [46, 47]. Therefore these arguments cannot be generalized to full-term infants, or to adult health outcomes. Childhood growth and CHD Is the increased risk for CHD associated with a small body size at birth and nonoptimal growth during infancy modified by growth during the childhood years? The HBCS has largely contributed to our present knowledge in this field. Deaths from CHD are associated with a small body size at birth, nonoptimal growth during infancy and an around average body mass index (BMI) during childhood. Interestingly, measures of childhood growth often interact with birth measurements in the prediction of adult disease. This is illustrated in Fig. 1 showing that an increase in body size from birth to 11 years of age was associated with an increased risk of CHD in adult life primarily in those who were small or thin at birth. Boys who were not small at birth did not seem to be at increased risk of CHD risk despite their higher childhood BMI [28]. In other words, the consequences of achieved childhood body size seem to be conditioned by growth in utero and during infancy; they do not depend only on the absolute level of weight or body size attained. Early and childhood growth of men who developed CHD is shown in Fig. 2 [36]. In the figure the z-score for the whole cohort is set at zero and an individual maintaining a steady position as large or small in relation to the other subjects would follow a horizontal path on the figure. Those who later developed CHD had been small at birth and during infancy, they experienced accelerated gain in weight and BMI after 5 years of age; however, their heights remained below average. This is consistent with the known association between CHD and short adult stature. These findings also support the view that increasing weight and BMI during infancy are associated with a reduced risk of CHD. 0.2 0.1 z-score 0 0.1 0.2 Cohort Height BMI Weight 0.3 0 6 12 18 2 4 6 8 10 Age (months) Age (years) Fig. 1 Hazard ratios for death from coronary heart disease according to ponderal index at birth and body mass index at age 11 years, adjusted for length of gestation. Arrows indicate average values. Reproduced from Eriksson et al. [28]. Fig. 2 Mean z-scores for height, weight, and BMI from birth to 11 years of age among boys who developed coronary heart disease. The mean values for the whole cohort are set at zero, with deviations from the mean expressed as standard deviations (z-scores). Adapted with permission by the Massachusetts Medical Society; Copyright ª 2005, Massachusetts Medical Society. N Engl J Med 2005; 353: 1802 9. 420 ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425

Women are known to have lower rates of CHD than men. There are also fewer studies focusing on the relationship between early growth and CHD among women. However, the findings in women usually follow those in men and CHD among women seems to be associated with small size at birth [6, 20, 21, 25, 34, 38]. The paths of fetal growth among women and men differ; boys tended to be thin at birth while the girls tended to be short at birth [34]. The associations between early growth and CHD in adult life are strong and graded. A theoretical estimation based upon data from the HBCS suggests that men who had a ponderal index at birth >26 kg m )3 and whose height and BMI at 1 year of age were above the average for the cohort had half the risk of CHD occurring before age 65 years. These estimates are obviously based upon data from a historical cohort but despite this they give a picture about the magnitude of risk associated with various patterns of growth. Type 2 diabetes and early growth Coronary heart disease is commonly associated with disturbances in glucose metabolism. Therefore it is not surprising that both CHD and type 2 diabetes share some early life risk factors. Fetal and infant growth patterns of individuals who later develop type 2 diabetes resemble much the growth of those who develop CHD [3, 4, 10, 17, 26, 30, 31, 39]. Childhood growth of those who later developed type 2 diabetes showed similarities with the growth patterns associated with CHD. Obesity is the single most important risk factor for type 2 diabetes. However, the patterns of growth that lead to these two disorders differ [39, 48, 49]. People who become obese tend to be big at birth, whereas people who develop type 2 diabetes tend more often to have low birth weight and to be small or thin at birth. The traditional pattern of growth associated with type 2 diabetes later in life is characterized by a small body size at birth and thinness at 1 year of age [30, 31, 33, 39]. The adiposity of young children, as assessed by BMI, decreases to a minimum around 5 6 years of age before increasing again, i.e. the adiposity or BMI rebound [49]. The cumulative incidence of type 2 diabetes in adult life in relation to age at BMI rebound is shown in Table 2 [31]. An early timing of the BMI rebound was associated with a significantly higher cumulative incidence of type 2 diabetes. The cumulative incidence decreased from 8.6% in those in whom the BMI rebound occurred before the age of 5 years to 1.8% in those in whom it occurred after the age of 7 years [31]. Table 2 also shows the relation between age at BMI rebound, BMI at 1 and 12 years of age and the cumulative incidence of type 2 diabetes in adult life. An early age at BMI rebound was preceded by thinness at birth and during infancy, but associated with an above average BMI at 12 years of age and high rates of type 2 diabetes in adult life. An early timing of the BMI rebound seems to be associated with the development of both obesity and type 2 diabetes [31, 49]. A unique population-based longitudinal study of children born in India, studied around the age of 30 years, showed similar results as in the Finnish study. Those who showed impaired glucose regulation had a lower BMI up to 2 years of age, followed by an early age at BMI rebound and accelerated gain in BMI until adult life [33]. A U-shaped relationship between birth size and type 2 diabetes in later life suggests that in fact both extremes of the birth weight spectrum are associated Table 2 Age at BMI rebound in relation to BMI at 1 and 12 years of age and cumulative incidence (CI) of type 2 diabetes in adult life Age at adiposity rebound (years) BMI at 1 year (kg m )2 ) BMI at 12 years (kg m )2 ) CI of type 2 diabetes (%) 4 16.9 20.1 8.6 )5 16.9 17.9 4.4 )6 17.7 17.2 3.2 )7 18.2 17.1 2.2 8 18.4 16.9 1.8 P-value for trend <0.001 <0.001 <0.001 BMI, body mass index. ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425 421

Mean z-score 1 0.75 0.5 0.25 0 0.25 0.5 Birthweight > 3.5 kg Birthweight 3.5 kg Fastinginsulin (pmol L 1 ) 85 80 75 70 65 60 55 50 45 40 3000 3500 >3500 Birthweight (g) Pro12Pro Pro12Ala/Ala12Ala Fig. 4 Fasting insulin concentrations according to birth size and PPAR c gene polymorphism 0.75 0 1 2 3 4 5 6 7 8 9 10 11 12 Age (years) Fig. 3 Growth of 8 760 children in two birth weight groups above and below 3.5 kg. The solid lines represent the change in BMI of all children in the two birth weight groups. The dotted lines represent the change in BMI of the 290 children who later developed type 2 diabetes. The z-scores for the entire cohort are set at zero, represented by the dashed line. Copyright ª 2003 American Diabetes Association. From Diabetes Care, Vol. 26, 2003: 3006 3010. Reprinted with permission from The American Diabetes Association. with an increased risk for type 2 diabetes [9, 32]. In the HBCS the majority or two-thirds of the type 2 diabetics were born with a birth weight below 3500 g, but two separate pathways leading to type 2 diabetes were identified (Fig. 3). The existence of these different growth patterns leading to the same disease could perhaps explain some of the heterogeneity associated with the disease in clinical practice. This could also have major impact of prevention and treatment of type 2 diabetes. Genes and early growth with the DOHaD field In 1999 Andrew Hattersley put forward the fetal insulin hypothesis offering an alternative explanation for the association between a low birth weight and adverse adult health outcomes [50]. He suggested that the same genetic factors would alter both intrauterine growth and e.g. adult glucose metabolism. In other words a small body size at birth and impaired glucose regulation in adult life would be phenotypes of the same underlying genotype. The fetal insulin hypothesis was primarily supported by findings in various monogenic forms of diabetes e.g. in diabetes caused by glucokinase gene mutations. Not only the glucokinase gene has been focused upon in this respect but also the IGF-I and insulin VNTR genes have been suggested to be simultaneously related to both fetal growth and adult health outcomes [51, 52]. Today there is no strong evidence suggesting that any common gene or gene variant would be explaining the association between birth size and later health outcomes. Important to keep in mind is that these two alternatives i.e. the genes versus environment are not mutually exclusive in explaining the pathogenesis of several noncommunicable diseases. The intrauterine environment might well interact with genes affecting health later in life. The peroxisome proliferator-activated receptor genes or the PPARs play an important role in regulation of glucose, lipid and energy metabolism [53, 54]. There is a common missense mutation in the functional domain of the human PPAR-c-2 gene resulting in a substitution of proline by alanine in codon 12. This has been found to modulate the transcriptional activity of the gene. The Pro12Ala variant of the gene has been found to be associated with improved insulin sensitivity and a lower risk for type 2 diabetes compared with the Pro12Pro genotype [55 57]. Fasting insulin is widely used as a proxy for insulin sensitivity [58]. Elderly individuals, who were carriers 422 ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425

of the Ala allele of the PPAR-c-2 gene had lower fasting insulin and glucose concentrations i.e. they were more insulin sensitive compared with the carriers of the Pro12Pro genotype (59). No differences in body size at birth or childhood body size were observed in relation to genotypes. Figure 4 shows that the well-known association between a small body size at birth and insulin resistance, was present only in individuals with the high risk Pro12Pro genotype. This suggests that the negative effect of a small body size at birth was abolished by the presence of the Ala allele. In other words the effects of the PPAR-c-2 genotype, in elderly people, depended on their body size at birth [59]. There was a significant interaction between birth size and genotype. Similar interactions between birth size and the PPAR-c-2 genotype have been made in relation to manifest type 2 diabetes and dyslipidaemia [60, 61]. The PC-1 gene is another candidate gene for type 2 diabetes being involved in the insulin signalling pathway [62]. The 121Q variant of the gene has a greater inhibitory action on the insulin receptor than the 121K variant and has been associated with insulin resistance. Subjects carrying the 121Q allele had a significantly higher prevalence of type 2 diabetes and hypertension, but only in the presence of a small body size at birth [63]. In other words a similar interaction as the one described above for the PPAR-c-2 gene and birth size. These findings could be interpreted as manifestations of gene early environmental interactions and illustrate the importance of gene-early environment interaction in relation to risk factors for CHD and type 2 diabetes. Conclusion Acknowledging the interactions between early growth and genotypes might help us designing individual therapies as well as planning dietary and exercise interventions for high risk groups. These interventions need to take into account individual variability not only in genetic setup but also in early patterns of growth. Conflict of interest statement No conflict of interest was declared. References 1 Barker DJ, Osmond C. Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet 1986; 1: 1077 81. 2 Barker DJP, Osmond C, Winter PD, Margetts B, Simmonds SJ. Weight in infancy and death from ischaemic heart disease. Lancet 1989; 2: 577 80. 3 Hales CN, Barker DJP, Clark PMS et al. Fetal and infant growth and impaired glucose tolerance at age 64. Br Med J 1991; 303, 1019 22. 4 Barker DJ, Hales CN, Fall CH, Osmond C, Phipps K, Clark PM. Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 1993; 36: 62 7. 5 Barker DJ, Osmond C, Simmonds SJ, Wield GA. The relation of small head circumference and thinness at birth to death from cardiovascular disease in adult life. Br Med J 1993; 306: 422 6. 6 Osmond C, Barker DJ, Winter PD, Fall CH, Simmonds SJ. Early growth and death from cardiovascular disease in women. Br Med J 1993; 307: 1519 24. 7 Phipps K, Barker DJ, Hales CN, Fall CH, Osmond C, Clark PM. Fetal growth and impaired glucose tolerance in men and women. Diabetologia 1993; 36: 225 8. 8 McCance DR, Pettitt DJ, Hanson RL, Jacobsson LT, Bennett PH, Knowler WC. Glucose, insulin concentrations and obesity in childhood and adolescence as predictors of NIDDM. Diabetologia 1994; 37: 617 23. 9 McCance DR, Pettitt DJ, Hanson RL, Jacobsson LTH, Knowler WC, Bennett PH. Birth weight and non-insulin dependent diabetes: thrifty genotype, thrifty phenotype, or surviving small baby genotype? Br Med J 1994; 308: 942 5. 10 Phillips DI, Barker DJ, Hales CN, Hirst S, Osmond C. Thinness at birth and insulin resistance in adult life. Diabetologia 1994; 37: 150 4. 11 Valdez R, Athens MA, Thompson GH, Bradshaw BS, Stern MP. Birthweight and adult health outcomes in a biethnic population in the USA. Diabetologia 1994; 37: 624 31. 12 Barker DJ, Martyn CN, Osmond C, Wield GA. Abnormal liver growth in utero and death from coronary heart disease. Br Med J 1995; 310: 703 4. 13 Fall CH, Osmond C, Barker DJ et al. Fetal and infant growth and cardiovascular risk factors in women. Br Med J 1995; 310: 428 32. 14 Fall CH, Vijayakumar M, Barker DJ, Osmond C, Duggleby S. Weight in infancy and prevalence of coronary heart disease in adult life. Br Med J 1995; 310: 17 9. 15 Frankel S, Elwood P, Sweetnam P, Yarnell J, Smith GD. Birthweight, body-mass index in middle age, and incident coronary heart disease. Lancet 1996; 348: 1478 80. ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425 423

16 Frankel S, Elwood P, Sweetnam P, Yarnell J, Smith GD. Birthweight, adult risk factors and incident coronary heart disease: the Caerphilly Study. Public Health. 1996; 110: 139 43. 17 Lithell HO, McKeigue PM, Berglund L, Mohsen R, Lithell UB, Leon DA. Relation of size at birth to non-insulin dependent diabetes and insulin concentrations in men aged 50 60 years. Br Med J 1996; 312: 406 10. 18 Stein CE, Fall CH, Kumaran K, Osmond C, Cox V, Barker DJ. Fetal growth and coronary heart disease in south India. Lancet 1996; 348: 1269 73. 19 Forsén T, Eriksson JG, Tuomilehto J, Teramo K, Osmond C, Barker DJP. Mother s weight in pregnancy and coronary heart disease in Finnish men. Br Med J 1997; 315: 837 40. 20 Rich-Edwards JW, Stampfer MJ, Manson JE et al. Birth weight and risk of cardiovascular disease in a cohort of women followed up since 1976. Br Med J 1997; 315: 396 400. 21 Leon DA, Lithell HO, Vagero D et al. Reduced fetal growth rate and increased risk of death from ischaemic heart disease: cohort study of 15 000 Swedish men and women born 1915 29. Br Med J 1998; 317: 241 5. 22 McKeigue PM, Lithell HO, Leon DA. Glucose tolerance and resistance to insulin-stimulated glucose uptake in men aged 70 years in relation to size at birth. Diabetologia 1998; 41: 1133 8. 23 Ravelli AC, van der Meulen JH, Michels RP et al. Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998; 351: 173 7. 24 Eriksson JG, Forsen T, Tuomilehto J, Winter PD, Osmond C, Barker DJP. Catch-up growth in childhood and death from coronary heart disease: longitudinal study. Br Med J 1999; 318, 427 31. 25 Forsen T, Eriksson JG, Tuomilehto J, Osmond C, Barker DJ. Growth in utero and during childhood among women who develop coronary heart disease: longitudinal study. Br Med J 1999; 319: 1403 7. 26 Forsen T, Eriksson J, Tuomilehto J, Reunanen A, Osmond C, Barker D. The fetal and childhood growth of persons who develop type 2 diabetes. Ann Intern Med 2000; 133: 176 82. 27 Roseboom TJ, van der Meulen JH, Osmond C et al. Coronary heart disease after prenatal exposure to the Dutch famine, 1944 45. Heart 2000; 84: 595 8. 28 Eriksson JG, Forsen T, Tuomilehto J, Osmond C, Barker DJP. Early growth and coronary heart disease in later life: longitudinal study. Br Med J 2001; 322: 949 53. 29 Jensen CB, Storgaard H, Dela F, Holst JJ, Madsbad S, Vaag AA. Early differential defects of insulin secretion and action in 19-year-old caucasian men who had low birth weight. Diabetes 2002; 51: 1271 80. 30 Eriksson J, Forsen T, Osmond C, Barker D. Pathways of infant and childhood growth that lead to type 2 diabetes. Diabetes Care 2003; 26: 3006 10. 31 Eriksson JG, Forsen T, Tuomilehto J, Osmond C, Barker DJP. Early adiposity rebound in childhood and risk of type 2 diabetes in adult life. Diabetologia 2003; 46: 322 9. 32 Wei JN, Sung FC, Li CY et al. Low birth weight and high birth weight infants are both at an increased risk to have type 2 diabetes among schoolchildren in Taiwan. Diabetes Care 2003; 26: 343 8. 33 Barghava SK, Sachdev HS, Fall CHD et al. Relation of serial changes in childhood body-mass index to impaired glucose tolerance in young adulthood. N Engl J Med 2004; 350: 865 75. 34 Forsen T, Osmond C, Eriksson JG, Barker DJ. Growth of girls who later develop coronary heart disease. Heart 2004; 90: 20 4. 35 Yajnik CS. Early life origins of insulin resistance and type 2 diabetes in India and other Asian countries. J Nutr 2004; 134: 205 10. 36 Barker DJP, Osmond C, Forsen TJ, Kajantie E, Eriksson JG. Trajectories of growth among children who later have coronary event. N Engl J Med 2005; 353: 1802 9. 37 Ozanne SE, Jensen CB, Tingey KJ, Storgaard H, Madsbad S, Vaag AA. Low birthweight is associated with specific changes in muscle insulin-signalling protein expression. Diabetologia 2005; 48: 547 52. 38 Rich-Edwards JW, Kleinman K, Michels KB et al. Longitudinal study of birth weight and adult body mass index in predicting risk of coronary heart disease and stroke in women. Br Med J. 2005; 330: 1115 20. 39 Eriksson JG, Osmond C, Kajantie E, Forsén TJ, Barker DJP. Patterns of growth among children who later develop type 2 diabetes or its risk factors. Diabetologia 2006; 49: 2853 8. 40 Lawlor DA, Davey Smith G, Clark H, Leon DA. The associations of birthweight, gestational age and childhood BMI with type 2 diabetes: findings from the Aberdeen Children of the 1950s cohort. Diabetologia 2006; 49: 2614 7. 41 Painter RC, de Rooij SR, Bossuyt PM et al. Early onset of coronary artery disease after prenatal exposure to the Dutch famine. Am J Clin Nutr 2006; 84: 322 7. 42 de Rooij SR, Painter RC, Roseboom TJ et al. Glucose tolerance at age 58 and the decline of glucose tolerance in comparison with age 50 in people prenatally exposed to the Dutch famine. Diabetologia 2006; 49: 637 43. 43 Roseboom T, de Rooij S, Painter R. The Dutch famine and its long-term consequences for adult health. Early Hum Dev. 2006; 82: 485 91. 44 Fall CHD, Barker DJP, Osmond C, Winter PD, Clark PMS, Hales CN. Relation of infant feeding to adult serum cholesterol concentration and death from ischaemic heart disease. Br Med J 1991; 304: 801 5. 45 Barker DJP, Martyn CN, Osmond C, Hales CN, Fall CHD. Growth in utero and serum cholesterol concentrations in adult life. Br Med J 1993; 307: 1524 7. 46 Singhal A, Cole TJ, Fewtrell M, Deanfield J, Lucas A. Is slower early growth beneficial for long-term cardiovascular health? Circulation 2004; 109: 1108 13. 47 Singhal A, Lucas A. Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 2004; 363: 1642 5. 48 Eriksson J, Forsen T, Osmond C, Barker D. Obesity from cradle to grave. Int J Obes Relat Metab Disord 2003; 27, 722 7. 49 Rolland-Cachera MF, Deheeger M, Bellisle F et al. Adiposity rebound in children: a simple indicator for predicting obesity. Am J Clin Nutr 1984; 39: 129 35. 424 ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425

50 Hattersley AT, Tooke JE. The fetal insulin hypothesis: an alternative explanation of the association of low birth-weight with diabetes and vascular disease. Lancet 1999; 353: 1789 92 51 Ong KKL, Phillips DI, Fall C et al. The insulin gene VNTR, type 2 diabetes and birth weight. Nat Genet 1999; 21: 262 3. 52 Vaessen N, Janssen JA, Heutink P et al. Associations between genetic variation in the gene for insulin-like growth factor-i and low birth weight. Lancet 2002; 359: 1036 7. 53 Debril MB, Renaud JP, Fajas L, Auwerx J. The pleiotropic functions of peroxisome proliferator-activated receptor g. J Mol Med 2001; 79: 30 47. 54 Stumvoll M, Häring H. The peroxisome proliferator-activated receptor-gamma2 Pro12Ala polymorphism. Diabetes 2002; 51: 2341 7. 55 Tonjes A, Scholz M, Loeffler M, Stumvoll M. Association of Pro12Ala polymorphism in peroxisome proliferator-activated receptor c with pre-diabetic phenotypes: meta-analysis of 57 studies on nondiabetic individuals. Diabetes Care 2006; 29: 2489 97. 56 Scaglioni S, Verduci E, Salvioni M et al. PPAR-gamma2 Pro12Ala variant, insulin resistance and plasma long-chain polyunsaturated fatty acids in childhood obesity. Pediatr Res 2006; 60: 485 9. 57 Soriguer F, Morcillo S, Cardona F et al. Pro12Ala polymorphism of the PPARG2 gene is associated with type 2 diabetes mellitus and peripheral insulin sensitivity in a population with a high intake of oleic acid. J Nutr 2006; 136: 2325 30. 58 Laakso M. How good a marker is insulin level for insulin resistance? Am J Epidemiol 1993; 137: 959 65. 59 Eriksson JG, Lindi V, Uusitupa M. The effects of the Pro12Ala polymorphism of the peroxisome proliferator-activated receptorgamma2 gene on insulin sensitivity and insulin metabolism interact with size at birth. Diabetes 2002; 51: 2321 4. 60 Eriksson JG, Osmond C, Lindi V et al. Interactions between peroxisome proliferator-activated receptor gene polymorphism and birth length influence risk for type 2 diabetes. Diabetes Care 2003; 26: 2476 7. 61 Eriksson J, Lindi V, Uusitupa M et al. The effects of the Pro12Ala polymorphism of the PPARgamma-2 gene on lipid metabolism interact with body size at birth. Clin Genet 2003; 64: 366 70. 62 Gu HF, Almgren P, Lindholm E et al. Association between the human glycoprotein PC-1 gene and elevated glucose and insulin levels in a paired-sibling analysis. Diabetes 2000; 49: 1601 3. 63 Kubaszek A, Markkanen A, Eriksson JG et al. The association of the K121Q polymorphism of the plasma cell glycoprotein- 1gene with type 2 diabetes and hypertension depends on size at birth. J Clin Endocrinol Metab 2004; 89: 2044 7. Correspondence: Johan G. Eriksson, Department of Public Health, University of Helsinki, University of Helsinki 00014, Finland. (fax: +358 9 19127540; e-mail: johan.eriksson@helsinki.fi). ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 418 425 425