- Email: [email protected]
The metabolic consequences of prematurity
Growth Hormone & IGF Research 14 (2004) S136–S139 www.elsevier.com/locate/ghir
The metabolic consequences of prematurity P.L. Hofman *, F. Regan, M. Harris, E. Robinson, W. Jackson, W.S. Cutfield Liggins Institute, University of Auckland, 2-6 Park Avenue, Auckland, New Zealand
Abstract An association between low birth weight, commonly a reflection of an adverse in utero environment, and the subsequent development of diseases such as type 2 diabetes and hypertension in later life is now generally accepted – as is an association between an adverse perinatal environment and a permanent reduction in insulin sensitivity. This and other metabolic abnormalities have been demonstrated from childhood through to adulthood in subjects who were born full-term but small for gestational age (SGA). Less is known about children born prematurely into an adverse neonatal environment. We present data demonstrating that premature infants also have metabolic abnormalities similar to those observed in full-term, SGA children, and that these occur irrespective of whether the premature infants are SGA or appropriate for gestational age (AGA). Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Metabolism; Prematurity; Small for gestational age; Insulin resistance; Insulin sensitivity
1. Introduction Long-term survival for infants born extremely prematurely (<32 weeks) has become an almost expected outcome over the past two decades. Progressive improvements in neonatal care have expanded the premature population so that it now comprises approximately 2.5% of total annual births. With this improvement in survival rate, the focus has shifted to the later consequences of prematurity. Not surprisingly, there has been an emphasis on improving developmental outcome and minimising behavioural problems in children born prematurely. However, their very low birth weight has lead to comparisons with SGA subjects and the associated risks of adult disease. The link between low birth weight and adult disease was first described by Barker et al., and their results were subsequently confirmed by several other groups [1–4]. This association has been recently reviewed [5]. Low birth weight has been significantly correlated with an increased risk of type 2 diabetes mellitus, hypertension, obesity, ischaemic heart disease and cerebrovascular accidents (CVAs). Associated metabolic *
Corresponding author. Tel.: +64-9-373-999x86453; fax: +64-9-3738763. E-mail address: [email protected] (P.L. Hofman). 1096-6374/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ghir.2004.03.029
abnormalities in these subjects have included dyslipidaemia, a procoagulation state, and insulin resistance. Of these abnormalities, insulin resistance is a universal finding. It is a fundamental and early abnormality found in subjects at high risk of type 2 diabetes mellitus and, when present in non-diabetic, genetically predisposed individuals, increases the risk of developing diabetes by more than 40% over a 20-year period [6]. Similarly, insulin resistance is postulated to be an early finding in essential hypertension and is observed in normotensive, first-degree relatives of hypertensive subjects (see review [7]). One recent paper utilised insulin-suppression tests – a well-validated measure of insulin sensitivity – to categorise 208 otherwise-healthy adult subjects into normal (steady-state plasma glucose (SSPG) <4.4 mmol/L), moderate (SSPG 4.4–7.6 mmol/ L) and marked (SSPG >7.8 mmol/L) insulin resistance [8]. After a follow-up period of 4–11 years, those subjects with moderate or severe insulin resistance had a marked increase not only in the rate of hypertension, coronary heart disease, type 2 diabetes and CVAs, but also in the rate of cancer. An isolated reduction in insulin sensitivity, therefore, was significantly linked to many later diseases characteristic of those associated with being born SGA. Insulin resistance, by definition, reflects the action of insulin in suppressing glucose production and increasing
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
glucose disposal. Thus, in the non-diabetic patient, increased insulin secretion is required to compensate for reduced sensitivity to insulin. However, other actions of insulin may not be impaired and this compensatory hyperinsulinism may cause pathological activation in other sites. This can lead to such diverse effects as increased sympathetic nervous system activation, renin– angiotensin system stimulation and vascular smooth muscle cell proliferation [9]. Consistent associations have also been observed between insulin resistance/ hyperinsulinism and dyslipidaemia that are believed to be causal [9].
2. Foetal programming The finding of metabolic abnormalities in SGA subjects has led to the concept of foetal programming. According to this hypothesis, during critical periods of foetal and early infant life, subjects exposed to an adverse environment develop compensatory survival responses that become permanent or ‘‘programmed’’. When the adverse environment is removed, these responses can become maladaptive and lead to later disease. For instance, in a nutrient-poor environment such as that experienced in a pregnancy complicated by utero-placental insufficiency, glucose availability can become critical and ideally should be preferentially diverted to essential organs (such as brain, heart, etc.). The development of insulin resistance could facilitate this process by reducing glucose uptake in muscle and fat. Although suppression of insulin secretion could result in the same effect, maintaining insulin secretion may be important in other roles in foetal homeostasis separate from carbohydrate metabolism. Evidence that prematurity is associated with later adult disease is scant, primarily due to a lack of older survivors. High survival rates following more extreme prematurity are a modern phenomenon (in the past 20–30 years), and these survivors are not yet in the atrisk age group for the diseases identified from SGA studies. There has been one publication about young adults who were born prematurely (mean age, 24 years) in which it was reported that these prematurely born adults had elevated systolic and diastolic blood pressures compared with a cohort of subjects who were born full-term and with normal birth weights [10]. These prematurely born subjects also had slightly, but significantly, elevated fasting insulin levels, suggesting they may have reduced insulin sensitivity. Results of a subsequent study in 9- to 12-year-old children using a modified oral glucose tolerance test (OGTT) with measurement of baseline and 30-min glucose and insulin levels also suggested there were abnormalities in glucose homeostasis [11]. Higher glucose levels following the glucose load occurred in patients with lower birth
S137
weights, although insulin levels were only associated with current body mass index (BMI). The two main variables known to influence insulin sensitivity in otherwise healthy children are BMI and puberty. In this study, 23% of boys and 42% of girls were in puberty, making interpretation of glucose/insulin interactions difficult. Nevertheless, this study also suggested there was a perturbation in insulin action/glucose regulation.
3. Assessing reduced insulin sensitivity in children born premature If insulin resistance is important in the pathogenesis of the adult diseases observed in subjects born full-term and SGA, and if this change in insulin sensitivity is secondary to the foetal and early infant environment, then measurable differences should be observed from birth. Relative hyperinsulinism in the neonatal period in prematurely born infants – an indirect but reasonably accurate marker of reduced insulin sensitivity – has been demonstrated [12,13]. However, more sophisticated and reliable assessments of insulin sensitivity are difficult to perform in infants for both technical and ethical reasons. We have previously demonstrated that healthy, prepubertal children (age, 4–10 years) who were born at term and SGA, are more likely to be insulin resistant compared with a short (<10th percentile) otherwise normal term AGA control group [14]. We utilised the minimal model approach pioneered by Richard Bergman [15] that involves performing an intravenous glucose tolerance test (IVGTT) after an overnight fast, followed by frequent blood sampling. A rapid (30second) infusion of dextrose (0.3 g/kg) is given at time 0, and a further rapid infusion of the insulin secretagogue, tolbutamide (5 mg/kg), is given at 20 min. Twenty-seven samples are collected and kept on ice until they can subsequently be analysed for insulin and glucose levels. The glucose and insulin values are then entered into a computer program utilising a mathematical model of glucose homeostasis. Three parameters are derived from this analysis: the insulin sensitivity index (SI ), the acute insulin response (AIR) and glucose effectiveness (Sg). The SI reflects insulin-stimulated glucose disposal and insulin suppression of glucose output; AIR represents beta cell function and Sg represents the ability of glucose to suppress glucose output and increase glucose uptake (approximately equivalent to insulin-independent glucose uptake). Both SI and Sg have been previously validated against the gold standard insulin clamp with an r value of 0.88 for SI when tolbutamide is used in the IVGTT [16]. To examine more precisely the question of impaired insulin sensitivity in premature infants, we recently completed a similar study using the minimal model in healthy, prepubertal children (age 4–10 years) who had
S138
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
Table 1 Patient characteristics Variables
Preterm
n Age in years Males (%) Gestational age in weeks BWtSDS HtSDS MPHSDS BMISDS *
Term
AGA
SGA
AGA
SGA
37 6.7 (1.3) 35.1 27.6 (2.2) )0.09 (0.81) )0.09 (1.37) 0.28 (1.07) 0.40 (2.22)
11 6.4 (1.8) 55.5 30.6 (3.2) )2.25 (0.79) )1.32 (1.23) )0.62 (1.34) )0.97 (1.45)
20 7.1 (1.4) 70.0 39.1 (1.3) )0.25 (0.72) )1.71 (1.52) )0.64 (0.82) )0.35 (1.50)
13 8.9 (2.3) 61.5 38.6 (1.8) )2.65 (1.29) )2.45 (0.80) )0.47 (1.29) )0.72 (1.23)
BWtSDS, birth weight SDS; MPHSDS, midparental height SDS; HtSDS, height SDS; BMISDS, body mass index SDS. Data on WtSDS was not available for all patients. However, the correlation between WtSDS and BMISDS in patients studied was an r of 0.96.
been born prematurely. Of the 48 children born prematurely (gestation <32 weeks), 11 were SGA (defined as a birthweight <10th percentile) and 37 were AGA. A comparison group of 20 healthy, prepubertal, term AGA children were recruited. The characteristics of each group are summarised in Table 1. Data from our previously published study on term SGA children are included for comparison. The groups were compared using a general linear regression model adjusting for known and suspected variables affecting SI . These included body mass index standard deviation score (BMISDS), height standard deviation score (HtSDS), age and sex. Our results confirmed that children born prematurely were insulin resistant, with both premature AGA and SGA children having a comparable reduction in insulin sensitivity (approximately 50% the SI of term AGA controls) and similar to levels reported for children born at term and SGA (Table 2). As expected, the premature groups had elevated AIRs, reflecting the compensatory insulin release required to maintain euglycaemia. Finally, there was a small but significant reduction in Sg, a finding not observed in the cohort of term, SGA children. Interestingly, there was no association with either birth weight or gestational age, with all premature children being insulin resistant (data not shown). The finding of reduced insulin sensitivity in premature infants suggests that a critical window of Table 2 Assessment of insulin sensitivity in healthy, prepubertal children (age 4–10 years) born prematurely Assessment
Premature AGA
Premature SGA
Term controls
SI AIR Sg Kg
13.9 6.7 510 643 2.01 0.74 2.82 0.86
18.9 10.0 286 189 2.06 0.61 2.42 0.87
30.9 21.9 188 144 2.41 0.89 2.59 0.77#
#
P < 0:05 either premature group versus controls. P < 0:01. ** P < 0:005. *
programming occurs early in the third trimester. Survival at such a premature age requires weeks in neonatal intensive care. This is an extremely stressful time when immaturity causes problems in multiple organ systems. Nutritional support during this time is also far from ideal and is invariably inadequate. We hypothesise that the metabolic abnormalities observed reflect programming due to this adverse extrauterine environment, similar to that seen in subjects born at term and SGA who experienced an adverse in utero environment. It remains to be elucidated which particular aspects of preterm care influence metabolic programming and whether the environment prior to the premature delivery is important. However, as all prematurely born children had reduced insulin sensitivity, and as there was no change in insulin sensitivity with gestational age, a later – rather than earlier – critical window for programming (possibly 30–32 weeks) is more likely. Another observation from this study was the lack of any additive effect on reduced insulin sensitivity from being both premature and SGA. This suggests there may be a maximal effect from the adverse environmental insult occurring in the third trimester, and that the combination of both prematurity and SGA does not add to the magnitude of these later metabolic abnormalities. Alternatively, it can be argued that all premature births are abnormal and involve some degree of growth restraint. It would have been interesting to investigate whether the cause of the prematurity influenced insulin sensitivity. However, this was not assessed in our cohort. Differences may exist between the metabolic abnormalities observed in term SGA and prematurely born individuals. An example of this was the finding of a reduced Sg in prematurely born subjects, which suggests that the metabolic changes linked with premature birth may be associated with a higher risk of developing diseases later in adulthood. It has been demonstrated that an isolated defect in insulin sensitivity, insulin secretion or glucose effectiveness is unlikely to cause diabetes mellitus, and that major defects in at least two of these parameters are generally needed [16]. The finding of
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
defects in two of these parameters in childhood is of concern and may increase the risk of type 2 diabetes mellitus. Other metabolic abnormalities have been described in SGA subjects. These include an increased risk of early adrenarche and pubarche, early menarche, polycystic ovarian syndrome (PCOS), ovarian hyporesponsiveness to follicle-stimulating hormone (FSH) and reduced uterine and ovarian size in postmenarchal adolescent girls [17–23]. In both girls and boys aged 3–6 months, elevated FSH levels have been documented compared with infants born full-term and AGA [24]. To date, there is no published data to indicate that prematurity is associated with these abnormalities, but as some abnormalities such as PCOS have insulin resistance as an recognised pathogenic factor, it would not be surprising if similar changes are observed.
4. Conclusion Premature birth is associated with metabolic abnormalities that are similar to those observed in subjects born at term and SGA. The cause for the reduced insulin sensitivity remains uncertain, but likely reflects exposure to the adverse postnatal/early neonatal environment with these children. It is important that further research identifies the pre- and postnatal variables that can influence this programming since therapeutic interventions – at least postnatally – hold great potential. References [1] D.J. Barker, C. Osmond, C.M. Law, The intrauterine and early postnatal origins of cardiovascular disease and chronic bronchitis, J. Epidemiol. Commun. Health 43 (1989) 237–240. [2] D.J. Barker, C. Osmond, J. Golding, D. Kuh, M.E. Wadsworth, Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease, Br. Med. J. 298 (1989) 564– 567. [3] V. Yiu, S. Buka, D. Zurakowski, M. McCormick, B. Brenner, K. Jabs, Relationship between birthweight and blood pressure in childhood, Am. J. Kidney Dis. 33 (1999) 253–260. [4] J. Leger, C. Levy-Marchal, J. Bloch, A. Pinet, D. Chevenne, D. Porquet, D. Collin, P. Czernichow, Reduced final height and indications for insulin resistance in 20 year olds born small for gestational age: regional cohort study, Br. Med. J. 315 (1997) 341– 347. [5] D.J. Barker, The malnourished baby and infant (Review), Br. Med. Bull. 60 (2001) 69–88. [6] B.C. Martin, J.H. Warram, A.S. Krolewski, R.N. Bergman, J.S. Soeldner, C.R. Kahn, Role of glucose and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-up study (Comment), Lancet 340 (1992) 925–929.
S139
[7] F.S. Facchini, N. Hua, F. Abbasi, G.M. Reaven, Insulin resistance as a predictor of age related disease, J. Clin. Endocrinol. Metab. 86 (2001) 3574–3579. [8] G.M. Reaven, H. Lithell, L. Landsberg, Hypertension and associated metabolic abnormalities – the role of insulin resistance and the sympathoadrenal system (Review), N. Engl. J. Med. 334 (1996) 374–381. [9] B.V. Howard, Insulin resistance and lipid metabolism (Review), Am. J. Cardiol. 84 (1999) 28J–32J. [10] R.J. Irving, N.R. Belton, R.A. Elton, B.R. Walker, Adult cardiovascular risk factors in premature babies, Lancet 355 (2000) 2135–2136 (Comment, Erratum appears in Lancet 356 (2000) 514). [11] M.S. Fewtrell, C. Doherty, T.J. Cole, M. Stafford, C.N. Hales, A. Lucas, Effects of size at birth, gestational age and early growth in preterm infants on glucose and insulin concentrations at 9–12 years, Diabetologia 43 (2000) 714–717. [12] I.P. Gray, P.A. Cooper, B.J. Cory, M. Toman, N.J. Crowther, The intrauterine environment is a strong determinant of glucose tolerance during the neonatal period, even in prematurity, J. Clin. Endocrinol. Metab. 87 (2002) 4252–4256. [13] L. Jackson, A. Burchell, A. McGeechan, R. Hume, An inadequate glycaemic response to glucagon is linked to insulin resistance in preterm infants? (Fetal & Neonatal Edition), Arch. Dis. Child. 88 (2003) F62–F66. [14] P.L. Hofman, W.S. Cutfield, E.M. Robinson, R.N. Bergman, R.K. Menon, M.A. Sperling, P.D. Gluckman, Insulin resistance in short children with intrauterine growth retardation, J. Clin. Endocrinol. Metab. 82 (1997) 402–406. [15] R.N. Bergman, D.T. Finegood, M. Ader, Assessment of insulin sensitivity in vivo, Endocrine Rev. 6 (1985) 45–86. [16] R.N. Bergman, Lilly lecture 1989. Toward physiological understanding of glucose tolerance. Minimal-model approach, Diabetes 38 (1989) 1512–1527. [17] L. Ibanez, C. Valls, E. Miro, M.V. Marcos, F. de Zegher, Early menarche and subclinical ovarian hyperandrogenism in girls with reduced adult height after low birth weight, J. Pediatr. Endocrinol. Metab. 15 (2002) 431–433. [18] L. Ibanez, N. Potau, A. Ferrer, F. Rodriguez-Hierro, M.V. Marcos, F. de Zegher, Reduced ovulation rate in adolescent girls born small for gestational age, J. Clin. Endocrinol. Metab. 87 (2002) 3391–3393. [19] L. Ibanez, N. Potau, D. Dunger, F. de Zegher, Precocious pubarche in girls and the development of androgen excess, J. Pediatr. Endocrinol. Metab. 13 (Suppl. 5) (2000) 1261–1263. [20] L. Ibanez, A. Ferrer, M.V. Marcos, F.R. Hierro, F. de Zegher, Early puberty: rapid progression and reduced final height in girls with low birth weight, Pediatrics 106 (2000) E72. [21] L. Ibanez, N. Potau, F. de Zegher, Ovarian hyporesponsiveness to follicle stimulating hormone in adolescent girls born small for gestational age, J. Clin. Endocrinol. Metab. 85 (2000) 2624–2626. [22] L. Ibanez, N. Potau, G. Enriquez, F. de Zegher, Reduced uterine and ovarian size in adolescent girls born small for gestational age, Pediatr. Res. 47 (2000) 575–577. [23] L. Ibanez, N. Potau, I. Francois, F. de Zegher, Precocious pubarche, hyperinsulinism, and ovarian hyperandrogenism in girls: relation to reduced fetal growth, J. Clin. Endocrinol. Metab. 83 (1998) 3558–3562. [24] L. Ibanez, C. Valls, M. Cols, A. Ferrer, M.V. Marcos, F. de Zegher, Hypersecretion of FSH in infant boys and girls born small for gestational age, J. Clin. Endocrinol. Metabol. 87 (2002) 1986–1988.
The metabolic consequences of prematurity P.L. Hofman *, F. Regan, M. Harris, E. Robinson, W. Jackson, W.S. Cutfield Liggins Institute, University of Auckland, 2-6 Park Avenue, Auckland, New Zealand
Abstract An association between low birth weight, commonly a reflection of an adverse in utero environment, and the subsequent development of diseases such as type 2 diabetes and hypertension in later life is now generally accepted – as is an association between an adverse perinatal environment and a permanent reduction in insulin sensitivity. This and other metabolic abnormalities have been demonstrated from childhood through to adulthood in subjects who were born full-term but small for gestational age (SGA). Less is known about children born prematurely into an adverse neonatal environment. We present data demonstrating that premature infants also have metabolic abnormalities similar to those observed in full-term, SGA children, and that these occur irrespective of whether the premature infants are SGA or appropriate for gestational age (AGA). Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Metabolism; Prematurity; Small for gestational age; Insulin resistance; Insulin sensitivity
1. Introduction Long-term survival for infants born extremely prematurely (<32 weeks) has become an almost expected outcome over the past two decades. Progressive improvements in neonatal care have expanded the premature population so that it now comprises approximately 2.5% of total annual births. With this improvement in survival rate, the focus has shifted to the later consequences of prematurity. Not surprisingly, there has been an emphasis on improving developmental outcome and minimising behavioural problems in children born prematurely. However, their very low birth weight has lead to comparisons with SGA subjects and the associated risks of adult disease. The link between low birth weight and adult disease was first described by Barker et al., and their results were subsequently confirmed by several other groups [1–4]. This association has been recently reviewed [5]. Low birth weight has been significantly correlated with an increased risk of type 2 diabetes mellitus, hypertension, obesity, ischaemic heart disease and cerebrovascular accidents (CVAs). Associated metabolic *
Corresponding author. Tel.: +64-9-373-999x86453; fax: +64-9-3738763. E-mail address: [email protected] (P.L. Hofman). 1096-6374/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ghir.2004.03.029
abnormalities in these subjects have included dyslipidaemia, a procoagulation state, and insulin resistance. Of these abnormalities, insulin resistance is a universal finding. It is a fundamental and early abnormality found in subjects at high risk of type 2 diabetes mellitus and, when present in non-diabetic, genetically predisposed individuals, increases the risk of developing diabetes by more than 40% over a 20-year period [6]. Similarly, insulin resistance is postulated to be an early finding in essential hypertension and is observed in normotensive, first-degree relatives of hypertensive subjects (see review [7]). One recent paper utilised insulin-suppression tests – a well-validated measure of insulin sensitivity – to categorise 208 otherwise-healthy adult subjects into normal (steady-state plasma glucose (SSPG) <4.4 mmol/L), moderate (SSPG 4.4–7.6 mmol/ L) and marked (SSPG >7.8 mmol/L) insulin resistance [8]. After a follow-up period of 4–11 years, those subjects with moderate or severe insulin resistance had a marked increase not only in the rate of hypertension, coronary heart disease, type 2 diabetes and CVAs, but also in the rate of cancer. An isolated reduction in insulin sensitivity, therefore, was significantly linked to many later diseases characteristic of those associated with being born SGA. Insulin resistance, by definition, reflects the action of insulin in suppressing glucose production and increasing
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
glucose disposal. Thus, in the non-diabetic patient, increased insulin secretion is required to compensate for reduced sensitivity to insulin. However, other actions of insulin may not be impaired and this compensatory hyperinsulinism may cause pathological activation in other sites. This can lead to such diverse effects as increased sympathetic nervous system activation, renin– angiotensin system stimulation and vascular smooth muscle cell proliferation [9]. Consistent associations have also been observed between insulin resistance/ hyperinsulinism and dyslipidaemia that are believed to be causal [9].
2. Foetal programming The finding of metabolic abnormalities in SGA subjects has led to the concept of foetal programming. According to this hypothesis, during critical periods of foetal and early infant life, subjects exposed to an adverse environment develop compensatory survival responses that become permanent or ‘‘programmed’’. When the adverse environment is removed, these responses can become maladaptive and lead to later disease. For instance, in a nutrient-poor environment such as that experienced in a pregnancy complicated by utero-placental insufficiency, glucose availability can become critical and ideally should be preferentially diverted to essential organs (such as brain, heart, etc.). The development of insulin resistance could facilitate this process by reducing glucose uptake in muscle and fat. Although suppression of insulin secretion could result in the same effect, maintaining insulin secretion may be important in other roles in foetal homeostasis separate from carbohydrate metabolism. Evidence that prematurity is associated with later adult disease is scant, primarily due to a lack of older survivors. High survival rates following more extreme prematurity are a modern phenomenon (in the past 20–30 years), and these survivors are not yet in the atrisk age group for the diseases identified from SGA studies. There has been one publication about young adults who were born prematurely (mean age, 24 years) in which it was reported that these prematurely born adults had elevated systolic and diastolic blood pressures compared with a cohort of subjects who were born full-term and with normal birth weights [10]. These prematurely born subjects also had slightly, but significantly, elevated fasting insulin levels, suggesting they may have reduced insulin sensitivity. Results of a subsequent study in 9- to 12-year-old children using a modified oral glucose tolerance test (OGTT) with measurement of baseline and 30-min glucose and insulin levels also suggested there were abnormalities in glucose homeostasis [11]. Higher glucose levels following the glucose load occurred in patients with lower birth
S137
weights, although insulin levels were only associated with current body mass index (BMI). The two main variables known to influence insulin sensitivity in otherwise healthy children are BMI and puberty. In this study, 23% of boys and 42% of girls were in puberty, making interpretation of glucose/insulin interactions difficult. Nevertheless, this study also suggested there was a perturbation in insulin action/glucose regulation.
3. Assessing reduced insulin sensitivity in children born premature If insulin resistance is important in the pathogenesis of the adult diseases observed in subjects born full-term and SGA, and if this change in insulin sensitivity is secondary to the foetal and early infant environment, then measurable differences should be observed from birth. Relative hyperinsulinism in the neonatal period in prematurely born infants – an indirect but reasonably accurate marker of reduced insulin sensitivity – has been demonstrated [12,13]. However, more sophisticated and reliable assessments of insulin sensitivity are difficult to perform in infants for both technical and ethical reasons. We have previously demonstrated that healthy, prepubertal children (age, 4–10 years) who were born at term and SGA, are more likely to be insulin resistant compared with a short (<10th percentile) otherwise normal term AGA control group [14]. We utilised the minimal model approach pioneered by Richard Bergman [15] that involves performing an intravenous glucose tolerance test (IVGTT) after an overnight fast, followed by frequent blood sampling. A rapid (30second) infusion of dextrose (0.3 g/kg) is given at time 0, and a further rapid infusion of the insulin secretagogue, tolbutamide (5 mg/kg), is given at 20 min. Twenty-seven samples are collected and kept on ice until they can subsequently be analysed for insulin and glucose levels. The glucose and insulin values are then entered into a computer program utilising a mathematical model of glucose homeostasis. Three parameters are derived from this analysis: the insulin sensitivity index (SI ), the acute insulin response (AIR) and glucose effectiveness (Sg). The SI reflects insulin-stimulated glucose disposal and insulin suppression of glucose output; AIR represents beta cell function and Sg represents the ability of glucose to suppress glucose output and increase glucose uptake (approximately equivalent to insulin-independent glucose uptake). Both SI and Sg have been previously validated against the gold standard insulin clamp with an r value of 0.88 for SI when tolbutamide is used in the IVGTT [16]. To examine more precisely the question of impaired insulin sensitivity in premature infants, we recently completed a similar study using the minimal model in healthy, prepubertal children (age 4–10 years) who had
S138
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
Table 1 Patient characteristics Variables
Preterm
n Age in years Males (%) Gestational age in weeks BWtSDS HtSDS MPHSDS BMISDS *
Term
AGA
SGA
AGA
SGA
37 6.7 (1.3) 35.1 27.6 (2.2) )0.09 (0.81) )0.09 (1.37) 0.28 (1.07) 0.40 (2.22)
11 6.4 (1.8) 55.5 30.6 (3.2) )2.25 (0.79) )1.32 (1.23) )0.62 (1.34) )0.97 (1.45)
20 7.1 (1.4) 70.0 39.1 (1.3) )0.25 (0.72) )1.71 (1.52) )0.64 (0.82) )0.35 (1.50)
13 8.9 (2.3) 61.5 38.6 (1.8) )2.65 (1.29) )2.45 (0.80) )0.47 (1.29) )0.72 (1.23)
BWtSDS, birth weight SDS; MPHSDS, midparental height SDS; HtSDS, height SDS; BMISDS, body mass index SDS. Data on WtSDS was not available for all patients. However, the correlation between WtSDS and BMISDS in patients studied was an r of 0.96.
been born prematurely. Of the 48 children born prematurely (gestation <32 weeks), 11 were SGA (defined as a birthweight <10th percentile) and 37 were AGA. A comparison group of 20 healthy, prepubertal, term AGA children were recruited. The characteristics of each group are summarised in Table 1. Data from our previously published study on term SGA children are included for comparison. The groups were compared using a general linear regression model adjusting for known and suspected variables affecting SI . These included body mass index standard deviation score (BMISDS), height standard deviation score (HtSDS), age and sex. Our results confirmed that children born prematurely were insulin resistant, with both premature AGA and SGA children having a comparable reduction in insulin sensitivity (approximately 50% the SI of term AGA controls) and similar to levels reported for children born at term and SGA (Table 2). As expected, the premature groups had elevated AIRs, reflecting the compensatory insulin release required to maintain euglycaemia. Finally, there was a small but significant reduction in Sg, a finding not observed in the cohort of term, SGA children. Interestingly, there was no association with either birth weight or gestational age, with all premature children being insulin resistant (data not shown). The finding of reduced insulin sensitivity in premature infants suggests that a critical window of Table 2 Assessment of insulin sensitivity in healthy, prepubertal children (age 4–10 years) born prematurely Assessment
Premature AGA
Premature SGA
Term controls
SI AIR Sg Kg
13.9 6.7 510 643 2.01 0.74 2.82 0.86
18.9 10.0 286 189 2.06 0.61 2.42 0.87
30.9 21.9 188 144 2.41 0.89 2.59 0.77#
#
P < 0:05 either premature group versus controls. P < 0:01. ** P < 0:005. *
programming occurs early in the third trimester. Survival at such a premature age requires weeks in neonatal intensive care. This is an extremely stressful time when immaturity causes problems in multiple organ systems. Nutritional support during this time is also far from ideal and is invariably inadequate. We hypothesise that the metabolic abnormalities observed reflect programming due to this adverse extrauterine environment, similar to that seen in subjects born at term and SGA who experienced an adverse in utero environment. It remains to be elucidated which particular aspects of preterm care influence metabolic programming and whether the environment prior to the premature delivery is important. However, as all prematurely born children had reduced insulin sensitivity, and as there was no change in insulin sensitivity with gestational age, a later – rather than earlier – critical window for programming (possibly 30–32 weeks) is more likely. Another observation from this study was the lack of any additive effect on reduced insulin sensitivity from being both premature and SGA. This suggests there may be a maximal effect from the adverse environmental insult occurring in the third trimester, and that the combination of both prematurity and SGA does not add to the magnitude of these later metabolic abnormalities. Alternatively, it can be argued that all premature births are abnormal and involve some degree of growth restraint. It would have been interesting to investigate whether the cause of the prematurity influenced insulin sensitivity. However, this was not assessed in our cohort. Differences may exist between the metabolic abnormalities observed in term SGA and prematurely born individuals. An example of this was the finding of a reduced Sg in prematurely born subjects, which suggests that the metabolic changes linked with premature birth may be associated with a higher risk of developing diseases later in adulthood. It has been demonstrated that an isolated defect in insulin sensitivity, insulin secretion or glucose effectiveness is unlikely to cause diabetes mellitus, and that major defects in at least two of these parameters are generally needed [16]. The finding of
P.L. Hofman et al. / Growth Hormone & IGF Research 14 (2004) S136–S139
defects in two of these parameters in childhood is of concern and may increase the risk of type 2 diabetes mellitus. Other metabolic abnormalities have been described in SGA subjects. These include an increased risk of early adrenarche and pubarche, early menarche, polycystic ovarian syndrome (PCOS), ovarian hyporesponsiveness to follicle-stimulating hormone (FSH) and reduced uterine and ovarian size in postmenarchal adolescent girls [17–23]. In both girls and boys aged 3–6 months, elevated FSH levels have been documented compared with infants born full-term and AGA [24]. To date, there is no published data to indicate that prematurity is associated with these abnormalities, but as some abnormalities such as PCOS have insulin resistance as an recognised pathogenic factor, it would not be surprising if similar changes are observed.
4. Conclusion Premature birth is associated with metabolic abnormalities that are similar to those observed in subjects born at term and SGA. The cause for the reduced insulin sensitivity remains uncertain, but likely reflects exposure to the adverse postnatal/early neonatal environment with these children. It is important that further research identifies the pre- and postnatal variables that can influence this programming since therapeutic interventions – at least postnatally – hold great potential. References [1] D.J. Barker, C. Osmond, C.M. Law, The intrauterine and early postnatal origins of cardiovascular disease and chronic bronchitis, J. Epidemiol. Commun. Health 43 (1989) 237–240. [2] D.J. Barker, C. Osmond, J. Golding, D. Kuh, M.E. Wadsworth, Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease, Br. Med. J. 298 (1989) 564– 567. [3] V. Yiu, S. Buka, D. Zurakowski, M. McCormick, B. Brenner, K. Jabs, Relationship between birthweight and blood pressure in childhood, Am. J. Kidney Dis. 33 (1999) 253–260. [4] J. Leger, C. Levy-Marchal, J. Bloch, A. Pinet, D. Chevenne, D. Porquet, D. Collin, P. Czernichow, Reduced final height and indications for insulin resistance in 20 year olds born small for gestational age: regional cohort study, Br. Med. J. 315 (1997) 341– 347. [5] D.J. Barker, The malnourished baby and infant (Review), Br. Med. Bull. 60 (2001) 69–88. [6] B.C. Martin, J.H. Warram, A.S. Krolewski, R.N. Bergman, J.S. Soeldner, C.R. Kahn, Role of glucose and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-up study (Comment), Lancet 340 (1992) 925–929.
S139
[7] F.S. Facchini, N. Hua, F. Abbasi, G.M. Reaven, Insulin resistance as a predictor of age related disease, J. Clin. Endocrinol. Metab. 86 (2001) 3574–3579. [8] G.M. Reaven, H. Lithell, L. Landsberg, Hypertension and associated metabolic abnormalities – the role of insulin resistance and the sympathoadrenal system (Review), N. Engl. J. Med. 334 (1996) 374–381. [9] B.V. Howard, Insulin resistance and lipid metabolism (Review), Am. J. Cardiol. 84 (1999) 28J–32J. [10] R.J. Irving, N.R. Belton, R.A. Elton, B.R. Walker, Adult cardiovascular risk factors in premature babies, Lancet 355 (2000) 2135–2136 (Comment, Erratum appears in Lancet 356 (2000) 514). [11] M.S. Fewtrell, C. Doherty, T.J. Cole, M. Stafford, C.N. Hales, A. Lucas, Effects of size at birth, gestational age and early growth in preterm infants on glucose and insulin concentrations at 9–12 years, Diabetologia 43 (2000) 714–717. [12] I.P. Gray, P.A. Cooper, B.J. Cory, M. Toman, N.J. Crowther, The intrauterine environment is a strong determinant of glucose tolerance during the neonatal period, even in prematurity, J. Clin. Endocrinol. Metab. 87 (2002) 4252–4256. [13] L. Jackson, A. Burchell, A. McGeechan, R. Hume, An inadequate glycaemic response to glucagon is linked to insulin resistance in preterm infants? (Fetal & Neonatal Edition), Arch. Dis. Child. 88 (2003) F62–F66. [14] P.L. Hofman, W.S. Cutfield, E.M. Robinson, R.N. Bergman, R.K. Menon, M.A. Sperling, P.D. Gluckman, Insulin resistance in short children with intrauterine growth retardation, J. Clin. Endocrinol. Metab. 82 (1997) 402–406. [15] R.N. Bergman, D.T. Finegood, M. Ader, Assessment of insulin sensitivity in vivo, Endocrine Rev. 6 (1985) 45–86. [16] R.N. Bergman, Lilly lecture 1989. Toward physiological understanding of glucose tolerance. Minimal-model approach, Diabetes 38 (1989) 1512–1527. [17] L. Ibanez, C. Valls, E. Miro, M.V. Marcos, F. de Zegher, Early menarche and subclinical ovarian hyperandrogenism in girls with reduced adult height after low birth weight, J. Pediatr. Endocrinol. Metab. 15 (2002) 431–433. [18] L. Ibanez, N. Potau, A. Ferrer, F. Rodriguez-Hierro, M.V. Marcos, F. de Zegher, Reduced ovulation rate in adolescent girls born small for gestational age, J. Clin. Endocrinol. Metab. 87 (2002) 3391–3393. [19] L. Ibanez, N. Potau, D. Dunger, F. de Zegher, Precocious pubarche in girls and the development of androgen excess, J. Pediatr. Endocrinol. Metab. 13 (Suppl. 5) (2000) 1261–1263. [20] L. Ibanez, A. Ferrer, M.V. Marcos, F.R. Hierro, F. de Zegher, Early puberty: rapid progression and reduced final height in girls with low birth weight, Pediatrics 106 (2000) E72. [21] L. Ibanez, N. Potau, F. de Zegher, Ovarian hyporesponsiveness to follicle stimulating hormone in adolescent girls born small for gestational age, J. Clin. Endocrinol. Metab. 85 (2000) 2624–2626. [22] L. Ibanez, N. Potau, G. Enriquez, F. de Zegher, Reduced uterine and ovarian size in adolescent girls born small for gestational age, Pediatr. Res. 47 (2000) 575–577. [23] L. Ibanez, N. Potau, I. Francois, F. de Zegher, Precocious pubarche, hyperinsulinism, and ovarian hyperandrogenism in girls: relation to reduced fetal growth, J. Clin. Endocrinol. Metab. 83 (1998) 3558–3562. [24] L. Ibanez, C. Valls, M. Cols, A. Ferrer, M.V. Marcos, F. de Zegher, Hypersecretion of FSH in infant boys and girls born small for gestational age, J. Clin. Endocrinol. Metabol. 87 (2002) 1986–1988.