sign in sign up
<<
Home , Metabolic syndrome, Weight gain

International Journal of (1999) 23, 656±659 ß 1999 Stockton Press All rights reserved 0307±0565/99 $12.00 http://www.stockton-press.co.uk/ijo Relative and obesity as a child predict as an adult

MJ Vanhala, MD1*, PT Vanhala2, SM KeinaÈnen-Kiukaanniemi3, EA Kumpusalo MD4 and JK Takala MD4

1Community Centre of Imatra Town, Virastokatu 2, 55100 Imatra; 2PieksaÈmaÈki District Health Centre, P.O.B. 85, 76101 PieksaÈmaÈki; 3Oulu University, Department of Public Health Science and General Practice, Oulu University Hospital, Kajaanintie 52A, 90220 Oulu; and 4Kuopio University, Department of Public Health and General Practice, P.O.B. 1627, FIN 70211 Kuopio, Finland

OBJECTIVE: To examine whether birth weight, weight gain from birth to the age of seven or body-mass index at the age of seven have any association with metabolic syndrome as an adult. DESIGN: A population study. SUBJECTS: 210 men and 218 women out of a total 712 subjects aged 36, 41 or 46 years in PieksaÈmaÈ ki town, Finland. MAIN OUTCOME MEASURES: Weight at birth and weight and height at the age of seven and metabolic syndrome de®ned as a clustering of , (hypertriglyceridaemia or low high-density-lipoprotein choles- terol), and resistance (inferred by abnormal tolerance or hyperinsulinaemia). RESULTS: No association was found between birth weight and the metabolic syndrome as an adult. Among obese children at the age of seven (body-mass index in the highest quartile), the odds ratio (OR) for the metabolic syndrome in adulthood was 4.4 (95% CI 2.1-9.5) as compared to the other children (the three other quartiles combined). After adjustment for age, sex and current obesity, the risk of the syndrome still was 2.4 (95% CI 2.1-9.5). CONCLUSION: We could not replicate the close association between low birth weight and the metabolic syndrome in adulthood as has been shown in some earlier studies. Obesity at the age of seven predicts the metabolic syndrome in adulthood.

Keywords: Metabolic syndrome; obesity; children; ; non-insulin-dependent ; cardiovascular risk factors

Introduction Material and methods

Insulin resistance has been suggested to be the All subjects (n ˆ 1008) born in the years 1947, 1952, primary abnormality in metabolic syndrome; a clus- and 1957 in PieksaÈmaÈki, Finland were invited to tering of cardiovascular risk factors such as hyper- attend examinations for the prevalence of metabolic triglyceridaemia, low high-density-lipoprotein (HDL)- syndrome, and 712 (71%) subjects participated. In the cholesterol, abnormal glucose , hyperin- health examination, body-mass-index (BMI), - sulinaemia and often hypertension.1±5 It has also been to-hip ratio (WHR), blood pressure, and non-insulin- shown that obesity is closely associated with insulin dependent diabetes (NIDDM) of ®rst degree relatives resistance and that it could be the `generator' for were recorded. In those subjects having either a close metabolic syndrome.6 Furthermore, we have recently relative with NIDDM, or BMI 30 kg=m2, or WHR shown that the risk of metabolic syndrome is espe- 1.0 in men and 0.88 in women, or using any cially high among those subjects who have been antihypertensive drug or having a systolic blood permanently obese from the age of seven to middle- pressure 160 mmHg, or a diastolic blood pressure age.7 Since there is also evidence to suggest that a low 95 mmHg, blood sample, were taken for measure- birth weight predicts the later occurrence of metabolic ment of fasting blood lipids (cholesterol, HDL-cho- syndrome,8±10 we were interested in ®nding whether lesterol, ), fasting plasma insulin, and two obesity or a high relative weight gain as a child we hour (75 g of glucose) according re¯ected in the occurrence of the syndrome as an to the protocol described elsewhere.11,12 adult. In the present study, metabolic syndrome was de®ned as a clustering1±4,11,12 of: (1) high blood pressure13 (either a systolic blood pressure 140 mmHg, or a *Correspondence: Dr. M. Vanhala Community Health Centre of diastolic blood pressure90 mmHg, or antihyperten- Imatra Town Virastokatu 2, 55100 Imatra, Finland. sive drug treatment), (2) dyslipidemia14,15 (hypertri- E-mail: mauno.vanhala@imatra.® Received 3 June 1998; revised 12 January 1999; accepted glyceridaemia 1.70 mmol=l, HDL-cholesterol < 1.0 1 February 1999 in men and < 1.20 in women, or both), and (3) insulin Childhood weight and adult metabolic syndrome MJ Vanhala et al 657 resistance16,17 (inferred by abnormal glucose metabo- weight (Table 1). Adjustment for age, sex, and the lism according to the WHO criteria or fasting hyper- current BMI did not change this result. insulinaemia 13.0 mU=l). As we have shown in this Among the 439 subjects with data of BMI at the age population earlier,11 by using the above described of seven, the metabolic syndrome was present in method the metabolic syndrome as de®ned 18=219 (8%) men, and in 12=220 (6%) women. In was detectable with a sensitivity of 96%. the individuals having the syndrome, BMI at the Weight and height at age of seven were traced for age of seven was statistically signi®cantly (P ˆ 439=712 (62%), records of birth weight for 428 0.001) higher 16.3Æ 2.4 kg=m2 as compared to (60%), and length at birth for 163 (23%) individuals. 15.3Æ 1.4 kg=m2 in the subjects without the syn- Records of both birth weight and BMI at age of seven drome. In the highest quartile of BMI at the age of were available for 344 (48%) subjects. seven, the crude OR for metabolic syndrome as an was de®ned as the sex-speci®c adult was 4.4 (95% CI 2.1±9.5), as compared to the highest quartile of BMI at age of seven, and adulthood three lowest quartiles combined. After taking into obesity as the sex-speci®c highest quartile of the account age, sex, and the current obesity, OR for the current BMI, meaning in this study BMI > 28.2 kg=m2 metabolic syndrome in adulthood still was 2.4 (95% in men, and BMI > 27.9 kg=m2 in women. The rela- CI 1.0±5.6) among the obese children (Table 1). tive weight was de®ned to be increased in childhood, The OR for the metabolic syndrome was 2.3 (95% if the quartile of BMI at the age of seven was higher CI 1.0±5.5) in those subjects whose relative weight than quartile of weight at birth. had been increased, as compared to the individuals Statistical analysis was performed by using SPSS whose relative weight was the same or had been for Windows 95. Means of the variables are presented reduced from birth to age of seven. Adjustment for with standard deviation (Æ sd). The relative risk age, sex, and the current obesity, reduced the OR for (Odds ratio) of the syndrome in different groups metabolic syndrome as an adult to 1.6 (95% CI 0.7± according to the relative weight change from birth 3.7) (Table 2). to the age of seven, and according to the quartiles of BMI at age of seven were calculated by using the logistic regression analysis. Discussion

The small numbers of cases and variability of effects Results are weaknesses of our study, but the presented results however are in line with result of Hulman et al.18 showing that the origins of adult risk of cardiovascular Among the 428 subjects with records of birth weight, disease and non-insulin-dependent diabetes can be the metabolic syndrome was present in 8% (n ˆ 34) of related to somatic growth as a child, but not necessa- men and women. The means of birth weights did not rily intrauterine growth. Our study therefore provides differ for each other in the groups of individuals with further evidence that the occurrence of the metabolic or without the metabolic syndrome (3557 grams vs syndrome as an adult may have an association with 3511 grams. P ˆ 0.624), and the prevalence of the childhood obesity and that a relative body weight gain syndrome ranged without any statistical signi®cance in childhood also may re¯ect in the occurrence of the from 5.6% to 11.3% in different quartiles of birth metabolic syndrome in adulthood. However, we could

Table 1 Prevalence (%) of metabolic syndrome (MBS) in the middle-age according to the sex-spesi®c quartiles of birth weight, and body-mass index at age of seven

Quartile I II III IV

Birth weight (grams) Boys (nˆ 210) < 3300 3300±3590 3600±3930 > 3930 Girls (n ˆ 218) < 3050 3051±3380 3381±3730 > 3730 Number of subjects in different quartiles 108 96 115 109 Prevalence % of MBS as an adult (n) 5.6% (6) 7.3% (7) 11.3% (13) 7.3% (8) BMI at age of seven (kg/m2) Boys (nˆ 219) < 14.6 14.7±15.2 15.3±16.3 > 16.3 Girls (n ˆ 220) < 14.4 14.4±15.1 15.2±16.1 > 16.1 Number of subjects in different quartiles 110 109 110 110 Prevalence % of MBS as an adult (n) 4.5%(5) 2.8%(3) 4.5%(5) 15.5%(17) Crude Odds Ratio (OR) for MBS as an adult 1 4.4 95% CI (2.1±9.5) OR for MBS, adjusted for sex, and age 1 4.5 (2.1±9.6) OR for MBS, adjusted for sex, age, and obesity1 1 2.4 (1.0±5.6)

1 Obesity as an adult: body-mass index > 28.2 kg/m2 in men, > 27.9 kg/m2 in women. Childhood weight and adult metabolic syndrome MJ Vanhala et al 658 Table 2 Prevalence % (n) of metabolic syndrome (MBS) according to change from quartile of birth weight to the quartile of body-mass index (BMI) at age of seven

Lower quartile of BMI Same quartile of BMI at Higher quartile of BMI at age of 7 than quartile age of 7 as quartile of at age of 7 than quartile of birth weight birth weight of birth weight

Number of subjects 121 95 128 Prevalence of MBS (n) 2%(3) 7%(7) 10%(13) Odds (OR) ratio for MBS as an adult 95% CI 1 2.3 (1.0±5.5) OR for MBS, adjusted for age, and sex 1 2.3 (0.9±5.4) OR for MBS, adjusted for age, sex and obesity1 1 1.6 (0.7±3.7)

1 Obesity as an adult: body-mass index > 28.2 kg/m2 in men, > 27.9 kg/m2 in women.

not replicate the close association between low birth prevention in the families of an obese child to reduce weight and metabolic syndrome in adulthood as NIDDM and in later life. described by Barker et al 8 and also shown in some other studies.9,10 Insulin resistance with the compensatory hyperin- Acknowledgements We are grateful to Pirjo Halonen, Statistician in sulinaemia is the primary defect in metabolic Kuopio University, for giving advice on statistical syndrome.1±4 It is also known that insulin is a analysis of data for this study. potential growth factor.19 In our study, 17 of the 30 adults (57%) with the syndrome had been obese at the age of seven. The crude OR for the syndrome was 2.3 References among those individuals whose relative weight had 1 Reaven GM. Banting Lecture 1988: role of insulin resistance been increased before the age of seven. Although the in human disease. Diabetes 1998; 37: 1595±1607. statistical signi®cance disappeared in the adjustment 2 De Fronzo RA, Ferrannini E. Insulin Resistance. A multi- faceted syndrome responsible for NIDDM, obesity, hyperten- for age, sex, and current obesity, these results suggest sion, dyslipidemia, and atherosclerotic cardiovascular disease. that there could be an association between the meta- Diab Care 1991; 14: 173±194. bolic syndrome and the relative weight gain in child- 3 Laakso M. The possible pathophysiology of insulin resistance hood. Furthermore, those subjects obese as children syndrome. Cardiovasc Risk Factors 1993; 1: 55±66. had a crude relative risk of 4.4 for the metabolic 4 Reaven GM. A Syndrome of resistance to insulin stimulated glucose uptake (Syndrome-X): De®nitions and implications. syndrome as an adult. Adjustment for age and sex did Cardiovasc Risk Factors 1993; 1: 2±11. not reduce the OR for the metabolic syndrome. Even 5 Raitakari OT, Porkka KVK, RoÈnnemaa T, Knip M, Uhari M, after taking into account current obesity, the OR for AÊ kerblom HK et al. The role of insulin in clustering of serum the syndrome occuring in adulthood for those obese as lipids and blood pressure in children and adolescents. The children was 2.4 (95% CI 1.0±5.6). We know that Cardiovascular Risk in Young Finns Study. Diabetologia 1995; 38: 1042±1050. BMI alone is not a perfect measure of adiposity and 6 BjoÈrntorp P. `Portal' as a generator of risk therefore, the data of WHR in childhood would have factors for cardiovascular disease and diabetes. Arteriosclero- given more information of the possible association sis 1990; 10: 493±496. between the metabolic syndrome and the type of 7 Vanhala MJ, Vanhala PT, Kumpusalo EA, Halonen P, obesity. Our ®nding however suggests that childhood Takala JK. Relation between obesity from childhood to adult- hood and the metabolic syndrome: population based study. obesity de®ned according to the BMI predicts meta- BMJ 1998; 17: 319. bolic disturbances in adults. One possible explanation 8 Barker DJP, Hales CN, Fall CHD, Osmond C, Phipps K, is that insulin resistance is present already in early Clark PMS. Type 2 (non-insulin-dependent) diabetes mellitus, childhood and might also result in weight gain of the hypertension and (syndrome-X): relation to child. Another explanation for our result is that both reduced fetal growth. Diabetologia 1993; 36: 62±67. 9 Phillips DIW, Barker DJP, Hales CN, Hirst S, Osmond C. obesity and metabolic syndrome may have some Thinness at birth and insulin resistance in adult life. Diabeto- common, yet unknown etiologic factor. logia 1994; 37: 150±154. 10 Lithell HO, McKeigue PM, Berglund L, Mohsen R, Lithell U-B, Leon D. Relation of size at birth to non-insulin- dependent diabetes and insulin concentrations in men aged 50±60 years. BMJ 1996; 312: 406±410. 11 Vanhala MJ, Kumpusalo EA, PitkaÈjaÈrvi TK, Takala JK. `Metabolic syndrome' in a middle-aged Finnish population. Conclusion J Cardiovascular Risk 1997; 4: 291±295. 12 Vanhala MJ, PitkaÈjaÈrvi TK, Kumpusalo EA, Takala JK. Obesity type and clustering of insulin resistance-associated We conclude that obesity and also a relative weight cardiovascular risk factors in middle-aged men and women. gain in childhood may carry an additional burden for Int J Obes 1998; 22: 369±374. the metabolic syndrome as an adult. Since we know 13 Joint National Committee on Detection, Evaluation and Treat- that gaining weight in childhood like adulthood is ment of High Blood Pressure. The ®fth report of the Joint 20 National Committee on Detection, Evaluation and Treatment increasing in westernized countries, our ®nding of High Blood Pressure (JNC-V). Arch Intern Med 1993; 153: should alert health care professionals to start early 154±183. Childhood weight and adult metabolic syndrome MJ Vanhala et al 659 14 Castelli WP. (1992) Epidemiology of Triglycerides: A View 18 Hulman S, Kushner H, Katz S, Falkner B. Can cardiovascular from Framingham. Am J Cardiol 70: 3H±9H. risk be predicted by newborn, childhood, and adolescent body 15 Prevention of coronary heart disease: Scienti®c background size? An examination of longitudinal data in urban African and new clinical quidelines. Recommendations of the Eur- Americans. J Pediatr 1998; 132: 90±97. opean Society prepared by the international 19 Froesch ER, Schmid C, Schwander J, Zapf J. Actions of task for the prevention of coronary heart disease. Nutr Metab insulin-like growth factors. Ann Rev Phys 1985; 47: 443± Cardiovasc Dis 1992; 113±156. 467. 16 World Health Organization. Report of WHO study group on 20 Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, diabetes mellitus. Geneva, Switzerland: World Health Orga- Johnson CL. Owerweight prevalence and trends for children nization; 1985: Technical report series 1985; No 727. and adolescents: the National Health and Examina- 17 Laakso M. How good a marker is insulin level for insulin tion Surveys, 1963 to 1991. Arch Pediatr Adolesc Med 1995; resistance? Am J Epidemiol 1993; 9: 959±965. 149: 1085±1091.