Metabolic markers in relation to nutrition and growth in healthy 4-y-old children in Sweden

¹ From the Department of Paediatrics, Institute for Clinical Sciences, Sahlgrenska Academy, Göteborg University, Göteborg, Sweden

² Supported by grants from the Swedish Dairy Association, the Medical Faculty of Göteborg University, the Swedish Nutrition Foundation, the Mayflower Charity Foundation for Children, and the Children’s Hospital Research Foundation.

³ Address reprint requests to M Garemo, Department of Paediatrics, Sahlgrenska Academy, Göteborg University, Queen Silvia Children’s Hospital, S-416 85 Göteborg, Sweden. E-mail: malingaremo{at}hotmail.com or birgitta.strandvik{at}pediat.gu.se.

ABSTRACT  
Background: The worldwide increase in overweight and obesity probably involves dietary factors, and early indicators of risk must be identified.

Objective: We aimed to analyze metabolic markers in relation to dietary intake and anthropometry in healthy 4-y-old children.

Design: A cross-sectional study of nutritional intake was performed in 95 children by use of 7-d food records. Fasting blood samples were analyzed for glucose, insulin, and lipids.

Results: The study population was representative of Swedish children except that more parents than the average had a university education. The boys’ mean energy intake was higher (6.6 ± 0.75 MJ) than the girls’ (5.7 ± 0.79 MJ). Significant associations were found between the percentage of energy from carbohydrates and that from fat (r = –0.91) and sucrose (r = 0.59). High body mass index was associated with a low percentage of energy from fat (r = –0.32). Serum triacylglycerol, insulin, and the HOMA (homeostatic model assessment) index were higher in girls than in boys. In girls, HOMA ß-cell function was significantly negatively associated with fat intake and serum fasting insulin, and HOMA insulin resistance indexes were significantly associated with the increment in z scores for height and weight from birth to age 4 y. Compared with children with fasting insulin concentrations below the group mean + SD, the children with concentrations above that value were smaller as newborns and had larger increments in growth z scores from birth to age 4 y.

Conclusion: In healthy Swedish 4-y-olds from well-educated families, low fat intake was related to high body mass index. Upward weight and height percentile crossings were related to insulin resistance, especially in girls.

Key Words: Body mass index • BMI • fat • HOMA indexes • insulin • preschool children

INTRODUCTION  
Different factors such as genetic predisposition, low physical activity, and altered dietary habits have been suggested to be involved in the dramatically increased prevalence of overweight and obesity, which also affects young children (1, 2). Factors that may affect the prevalence of overweight early in life must be identified. Low birth weight is associated with a higher risk of type 2 diabetes, overweight, and coronary heart disease later in life, and epidemiologic studies have also shown that the growth rate during infancy and childhood can predict later development of type 2 diabetes (3). An early adiposity rebound and upward percentile crossing of weight and body mass index (BMI) have been identified as risk factors for subsequent obesity (4-6). Insulin resistance differs between ethnicities and sexes; it is more common in girls than in boys (7-9). Dyslipidemia is well documented in older children and adults who are overweight (10, 11), and associations have been found between serum fasting insulin concentrations and serum lipids in 2–3-y-old Hispanic children (12). These results indicate that factors related to the metabolic syndrome may be identified early in life.

Even though energy intake has been widely viewed as a crucial determinant of overweight, the findings are contradictory (13, 14). The quality of macronutrients has attracted less attention, and concerns have recently been raised regarding recommendations related to low-fat, high-carbohydrate diets (15). A high beverage intake has, especially in older children, been associated with high BMI (16, 17).

To analyze the effects of dietary intake on the health of preschool children, we combined anthropometry with analysis of metabolic markers in serum and studied correlations with energy and macronutrient intakes. This study was a part of a larger investigation of nutrition, bone mass, health, and lifestyle in 4-y-old children in an urban community in western Sweden.

SUBJECTS AND METHODS  
Subjects
Parents received initial information about the study together with their invitation to the routine 4-y check-up at the Pediatric Health Care Center. Two hundred thirty children from different socioeconomic districts were invited to participate in a study on nutrition and health. The parents of 182 children accepted and answered questionnaires about socioeconomics, health data, and food choice, and 95 of these parents also both kept a 7-d food record and permitted their children to undergo blood testing. These 95 children constitute the study group reported on here. Extensive analysis of other data will be reported elsewhere.

The study was performed according to the Helsinki Declaration and was approved by the ethics committee of Göteborg University, Sweden. Informed consent was obtained from all parents. None of the children was forced to undergo vein puncture.

Design
The parents who agreed to participate received further information at the visit to the Pediatric Health Care Center, where they completed the questionnaires and agreed to keep a 7-d food record. They also were given local anesthetic plasters (EMLA; Astra-Zeneca, Södertälje, Sweden) to be applied to the child’s skin 1 h before coming in for the blood tests, which were performed in the morning after the children had fasted overnight.

Anthropometry
Height and weight were measured with the use of calibrated standard equipment. BMI (kg/m²) and SD score of the anthropometric data [z score = (attained value – mean standard value for age)/SD of value for age (18, 19)] were calculated. Each child’s z score was calculated by using the closest mean standard value of the age defined as follows: 51 mo old was defined as being 4 y old, 52–57 mo old as 4.5 y old, and 59 mo as 5 y old. The International Obesity Task Force (IOTF) definition was used to determine the respective cutoffs for overweight and obesity (20).

Food records
To achieve good compliance and quality in the 7-d record-keeping, the process was divided into 2 periods, one lasting from Saturday to Tuesday and the other from Wednesday to Friday of the following week. Splitting the recording in this way decreased the drop out to 14% in a study of 153 individuals, compared with 37% in a pilot study of 60 individuals (data not shown). Many parents in the pilot study had given the large work load involved in recording on 7 consecutive days as the reason for dropping out.

Serving sizes were given in household measures and were based on standardized pictures (Matmallen offprint; Swedish Food Authority, Uppsala, Sweden). The food records were coded and computerized for analysis by using the computer nutrient software package MATs (Rudans Lättdata, Västerås, Sweden) with the Swedish Food Administration’s food database, which includes 1500 foods and dishes. Around 20 food items often consumed by the children were added by using nutrient data given by the manufacturers.

Blood testing and analysis
Blood testing was undertaken after the children had fasted overnight and was facilitated by the analgesic effect of the local anesthetic plasters. Serum concentrations of cholesterol, triacylglycerol, and plasma glucose were analyzed by photometric methods according to clinical routines (111491458, 2004;11730711, 2003; and 1876899, 2000, respectively; Roche Diagnostics Scandinavia AB, Stockholm, Sweden). Serum insulin was assessed by using the Insulin Ultrasensitive ELISA kit (Mercodia AB, Uppsala, Sweden); the intraassay and interassay CVs were 5.3% and 3.9%, respectively (n = 6). Homeostatic model assessment (HOMA) ß-cell function and HOMA insulin resistance (HOMA-IR) indexes were calculated according to Matthews et al (21). The HOMA ß-cell function index has been found to be higher in obese than in nonobese children and to be significantly related to both first- and second-phases of insulin secretion in hyperglycemic clamps in children and adolescents (22).

Statistical analysis
Means (±SDs) were used to describe the data unless indicated otherwise. The STATVIEW 5 software program was used for statistical analysis. Unpaired Student’s t tests were used when comparing groups. Linear regression and Pearson’s correlation coefficients were used as measures of relations between 2 variables. P values < 0.05 were regarded as statistically significant.

RESULTS  
The children’s mean age was 51 ± 2 mo at the time of the examination, and 42 of the 95 children were girls. Forty-five percent of the parents were <34 y old, and 8% of them had been born abroad. Six percent had an elementary school education and 47% and 47% had a secondary or a university education, respectively.

Anthropometric data are shown in Table 1. On the group level, birth weights and lengths corresponded with the reference values. On the individual level, 86% and 89% of the girls and boys, respectively, were within ± 2 SDs of the reference values (18). Significant increments in z scores for both weight and height were seen between birth and 4 y of age. According to the IOTF definition, 20% of the children were overweight (n = 16; 7 girls) or obese (n = 3; one girl). The children’s mean energy intake was significantly higher than that stipulated in the Nordic Nutrition Recommendations (NNR) (23) for both girls and boys: 5.8 ± 0.79 and 6.6 ± 0.78 MJ, respectively (P < 0.0001; Table 2). On the group level, intake expressed as percentages of energy from protein, fat, and carbohydrates corresponded with the guidelines, but the percentage of energy from sucrose was above the NNR level, and the percentage of energy from polyunsaturated fatty acids, both n–6 and n–3 fatty acids, was below NNR levels. The percentage of energy from sucrose and that from carbohydrates were closely related (r = 0.59, P < 0.0001). A strong negative association was found between the percentage of energy from fat and that from carbohydrates (r = –0.91, P < 0.0001), as well as between the percentage of energy from fat and that from sucrose (r = –0.36, P = 0.0004). The percentage of energy from carbohydrates was negatively associated with protein intake (r = –0.38, P = 0.0002), and the association between the percentage of energy from sucrose and protein was r = –0.61 (P < 0.0001).

View this table:
TABLE 1. Anthropometric characteristics of 95 healthy 4-y-olds¹

 

View this table:
TABLE 2. Energy and macronutrient intakes based on the 7-d food records of 95 healthy 4-y-olds¹

 
The children’s energy intake was associated with both body weight and height, r = 0.30 (P = 0.003) and r = 0.32 (P = 0.002), respectively, but no association was found with BMI. The percentage of energy intake from carbohydrates was associated with both body weight (r = 0.27, P = 0.009) and BMI (r = 0.29, P = 0.004), and the percentage of energy from fat was negatively associated with both body weight (r = –0.29, P = 0.005) and BMI (Figure 1).

View larger version (22K):
FIGURE 1.. Fat intake as a percentage of energy in relation to BMI on the basis of 7-d food records from 95 healthy 4-y-olds (42 girls). Linear regression analysis: r = –0.32, P = 0.001.

 
No significant difference was found in absolute energy intake between normal-weight and overweight or obese children, but the normal-weight children had higher energy and macronutrient intakes, expressed as g/kg body wt, than did the overweight and obese children (Table 3). The overweight and obese children had a lower percentage of energy intake as fat.

View this table:
TABLE 3. Energy and macronutrient intakes in normal-weight (BMI < 25) and overweight or obese (BMI 25) healthy 4-y-olds¹

 
Metabolic markers and serum lipids are depicted in Table 4. The girls had significantly higher values than did the boys for all variables except fasting blood glucose and serum cholesterol. Fasting blood glucose was <3.5 mmol/L in 2 children (1 girl), and HOMA ß-cell function could not be calculated in these girls, neither of whom had symptoms of hypoglycemia. Fasting serum insulin concentrations and blood glucose were significantly associated in boys (r = 0.38, P = 0.008) but not in girls.

View this table:
TABLE 4. Metabolic markers in 95 healthy 4-y-olds¹

 
The girls had higher serum triacylglycerol concentrations than did the boys (Table 4). The mean serum triacylglycerol concentration was 0.76 ± 0.24 mmol/L in the 3 obese children, compared with 0.68 ± 0.21 mmol/L in the normal-weight and overweight children (P = 0.50). No associations were found between dietary intake or anthropometric data and serum concentrations of triacylglycerol and cholesterol. No association was found between serum lipids and glucose, insulin, or HOMA indexes.

HOMA ß-cell function was negatively associated with energy intake (r = –0.33, P = 0.002). In girls, an association was also found between HOMA ß-cell function and the percentage of energy intake as carbohydrates (r = 0.33, P = 0.04), and negative associations were found with fat intake, expressed as g (r = –0.40, P = 0.01), g/kg body wt (r = –0.41, P = 0.01), and % of energy (r = –0.37, P = 0.02). No association was found between HOMA ß-cell function and macronutrients in boys.

In girls, but not in boys, serum fasting insulin concentrations and HOMA-IR were associated with z scores for birth length, r = 0.36 in both cases (P = 0.02 and P = 0.03, respectively), and HOMA-IR was associated with birth length (r = 0.36, P = 0.03). In girls, but not in boys, serum fasting insulin concentration and HOMA-IR were also associated with the change in z score from birth to 4 y for height, r = 0.42 in both cases (P = 0.007 and P = 0.009, respectively); corresponding associations for weight were r = 0.37 (P = 0.02) in both cases. No associations between serum fasting insulin concentration or HOMA-IR and birth weight, z score for birth weight, weight or height at age 4 y, z score for weight or height at age 4 y, or BMI were found in the children.

Further analysis of fasting serum insulin concentrations and anthropometry is depicted in Table 5. Subjects were grouped according to serum insulin concentration above or below the group mean + SD. There were no significant differences between the girls and the boys in a regression model with both sex and insulin concentration and their interaction term. Because no significant sex-related differences or interactions were found, significance testing included all children. The results further support a relation between small size at birth as well as upward centile crossing and increased fasting serum insulin concentrations at this early age.

View this table:
TABLE 5. Comparison of anthropometric data in healthy 4-y-olds with fasting serum (fS) insulin concentrations above or below the group mean + SD

 

DISCUSSION  
In this urban Swedish population with well-educated parents, healthy 4-y-olds had a significantly higher mean energy intake than is currently recommended (23), and 20% were overweight or obese according to the IOTF definitions. Interestingly, a high percentage of energy intake as fat was associated with lower BMI. The normal-weight children had a higher intake of macronutrients, expressed as g/kg, and a higher percentage of energy intake from fat than did the overweight and obese children, and we could not find an association between total energy intake and BMI, which corroborates other findings (24, 25). The girls had higher values for the metabolic markers than did the boys, similar to earlier reports (9, 26), which indicates higher pressure on glucose-insulin homeostasis. We could not confirm earlier findings of a relation between weight at age 5 and insulin resistance in girls (27), but we found a correlation with the change in z score from birth to age 4 y for both weight and height. Both serum fasting insulin concentration and HOMA-IR were associated with increased z scores for weight and height from birth to age 4 y, indicating that an upward centile crossing was associated with insulin resistance (6). Our results agree with those in 4-y-old Indian children with low birth weight (28). The data imply that very early factors might influence growth, indicating that fetal or neonatal programming must be considered (29). In animal studies, dietarily induced insulin resistance has been shown to precede other manifestations of the metabolic syndrome (30), although obesity is usually considered to precede insulin resistance in humans.

A high percentage of energy intake as fat was also associated with lower insulin production, measured as HOMA ß-cell function, in girls. The strong inverse relation found between intakes of carbohydrates and fat as a percentage of energy raises the question of which is more important for HOMA ß-cell function. These dietary associations and the relation between high insulin concentrations (>1 SD of the group mean) at age 4 y and low birth weight and low birth length, as well as the increment of z scores from birth to age 4 y, support the original observations by Barker and others (31, 32). Carbohydrates might stimulate insulin production, and high insulin concentrations might stimulate appetite and growth and thus be associated with a higher risk of subsequent disease such as obesity and diabetes (4, 5). Only a few reports on insulin sensitivity at young ages have been published, the results of which concur with our findings, indicating that insulin resistance in healthy children at this early age is related to weight gain (28, 33).

The serum triacylglycerol concentration was higher in girls than in boys but was not higher in overweight children than in normal-weight children. Serum lipid concentrations corresponded with concentrations previously reported in Swedish 9-y-olds (34). No relation was found between insulin resistance and serum lipids, as was previously reported in 2–3-y-olds (12). This difference might be explained by differences in ethnicity.

Twenty years ago, mean energy intake in the same age group was 5–10% higher than in this study, ie, 6.4 MJ in girls and 6.9 MJ in boys, but the prevalence of overweight was only 5% compared with 20% in our study group (35). Low socioeconomic status has been shown to be a risk factor for obesity, but many parents in our study had university educations and high socioeconomic status (data not shown), and other factors must therefore be considered (24, 36). Our study suggests that sources of energy might be stronger predictors of BMI than is total energy intake. During the past 2 decades, decreased fat intake has been reported, but overweight has increased remarkably during the same period (37). Fat results in greater satiety than do carbohydrates (38, 39). Our results showed that low fat intake was associated with higher carbohydrate intake, which was also observed by Payne et al (40). Many studies, mainly in older children, have reported a relation between soft drink consumption and overweight, which was not confirmed in our study of 4-y-olds (to be published), but the strong relation between sucrose and carbohydrates in this study indicated an unfavorable profile (16, 17). Because fat was not replaced with complex carbohydrates, a higher fat intake seemed to be more favorable at this age, which is supported by the observed negative associations between fat intake and BMI in all children and between fat intake and HOMA index and insulin concentrations in girls. The negative association between protein and sucrose intake suggests that a higher protein intake might also be more favorable, thus contributing to a higher degree of satiety. The results suggest that a high-carbohydrate diet would not be recommended at this age and that it might be more important to focus on the sources of energy in predicting overweight. Also, the quality of fat must be considered, because different fatty acids can influence gene expression of lipolytic and glycolytic enzyme systems (41), and animal studies have shown that perinatal exposure to different fatty acids can program for different development of body weight and insulin production (42, 43).

In summary, 20% of the healthy 4-y-olds in our study group of children of well-educated parents were overweight or obese. No relation was found between BMI and energy intake, but a negative association was found between the percentage of energy intake as fat and BMI as well as between the percentage of energy from fat and that from carbohydrates, which suggests that it might be the sources of energy rather than total energy intake that predisposes to the development of overweight at this age. The associations between fasting serum insulin concentrations and the insulin resistance index and the change in z scores for height and weight from birth to age 4 y in girls suggest that early upward centile crossing, as an indication of early adiposity rebound, might be primarily linked to metabolic changes. We are currently prospectively following our cohort of healthy children at the age of 8 y.

ACKNOWLEDGMENTS  
We thank Ellen Karlge Nilsson for assisting with data collection, Käthe Strandner and Kerstin Herlitz for taking all the blood samples without frightening the children, and Associate Professor Max Petzold for statistical advice. We gratefully acknowledge Ragnhild Arvidssson Lenner for stimulating discussion and the children and parents for participating in the study.

MG participated in the design and execution of all parts of the study, including data collection, analysis, and interpretation and writing the manuscript. VP performed the insulin analyses. BS was the principal investigator; developed the underlying hypothesis; was responsible for the initial concept of the study design; supervised the execution of the study, the data collection, and the data analysis; and had the main responsibility for the manuscript. None of the authors had any personal or financial conflicts of interest.

REFERENCES  

1. Ogden CL, Troiano RP, Briefel RR, Kuczmarski R, Flegal KM, Johnson CL. Prevalence of overweight among preschool children in the United States, 1971 through 1994. Pediatrics1997; 99 :e1 –7.
2. Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public-health crisis, common sense cure. Lancet2002; 360 :473 –82.
3. Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, Law C. Being big or growing fast: systematic review of size and growth in infancy and later obesity. BMJ2005; 331 :929 –34.
4. Rolland-Cachera MF. Rate of growth in early life: a predictor of later health? Adv Exp Med Biol2005; 569 :35 –9.
5. Eriksson JG, Forsén T, Tuomilehto J, Osmond C, Barker DJP. Early adiposity rebound in childhood and risk of type 2 diabetes in adult life. Diabetologia2003; 46 :190 –4.
6. Cole TJ. Children grow and horses race: is the adiposity rebound a critical period for later obesity? BMC Pediatr2004; 4 :6 .
7. Ehtisham S, Crabtree N, Clark P, Shaw N, Barrett T. Ethnic differences in insulin resistance and body composition in United Kingdom adolescents. J Clin Endocrinol Metab2005; 90 :3962 –9.
8. Goran MI, Cruz ML, Bergman RN, Watanabee R. Insulin resistance and associated compensatory responses in African-American and Hispanic children. Diabetes Care2002; 25 :2184 –90.
9. Murphy MM, Metcalf BS, Voss L, et al. Girls at five are intrinsically more insulin resistant than boys: the programming hypothesis revisited—the Early Bird study (Early Bird 6). Pediatrics2004; 113 :82 –6.
10. Reinehr T, Andler W, Denzer C, Siegried W, Mayer H, Wabitsch M. Cardiovascular risk factors in overweight German children and adolescents: relation to gender, age and degree of overweight. Nutr Metab Cardiovasc Dis2005; 15 :181 –7.
11. Boyd GS, Koeinigsberg J, Falkner B, Gidding S, Hassink S. Effect of obesity and high blood pressure on plasma lipid levels in children and adolescents. Pediatrics2005; 116 :442 –6.
12. Shea S, Aymong E, Zybert P, Berglund L, et al. Fasting plasma insulin modulates lipid levels and particle size in 2- to 3-year-old children. Obes Res2003; 11 :709 –21.
13. Speakman JR. Obesity: the integrated roles of environment and genetics. J Nutr2004; 134 :2090S –105S.
14. Livesey G. Low-glycaemic diets and health: implications for obesity. Proc Nutr Soc2005; 64 :105 –13.
15. Weinberg SL. The diet-heart hypothesis: a critique. J Am Coll Cardiol2004; 43 :731 –3.
16. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational study. Lancet2001; 357 :505 –8.
17. Harnack L, Stang J, Story M. Soft drink consumption among US children and adolescents: nutritional consequences. J Am Diet Ass1999; 99 :436 –41.
18. Karlberg P, Taranger J, Engström I, Lichtenstein H, Svennberg-Redegren I. The somatic development of children in a Swedish urban community. A prospective longitudinal study. Acta Paediatr Scand Suppl1976; 258 :1 –148.
19. Karlberg J, Luo ZC, Albertsson-Wikland K. Body mass index reference values (mean and SD) for Swedish children. Acta Paediatr2001; 90 :1427 –34.
20. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ2000; 320 :1240 –3.
21. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia1985; 28 :412 –9.
22. Gungor N, Saad R, Janosky J, Arslanian S. Validation of surrogate estimates of insulin sensitivity and insulin secretion in children and adolescents. J Pediatr2004; 144 :47 –55.
23. Nordic Council of Ministers. Nordic nutrition recommendations 2004 integrating nutrition and physical activity. 4th ed. Copenhagen, Denmark: Nordic Council of Ministers,2004 :13 .
24. Newby PK, Peterson KE, Berkey CS, Leppert J, Willett WC, Colditz GA. Dietary composition and weight change among low-income preschool children. Arch Pediatr Adolesc Med2003; 157 :759 –64.
25. Bogaert N, Steinbeck KS, Baur LA, Brock K, Bermingham MA. Food, activity and family-environmental vs biochemical predictors of weight gain in children. Eur J Clin Nutr2003; 57 :1242 –9.
26. Garces C, Cano B, Granizo JJ, et al. Insulin and HOMA in Spanish prepubertal children: relationship with lipid profile. Clin Biochem2005; 38 :920 –4.
27. Wilkin TJ, Metcalf BS, Voss LD, et al. The relative contributions of birth weight, weight change and current weight to insulin resistance in contemporary 5-year-olds: the EarlyBird Study. Diabetes2002; 51 :3468 –72.
28. Yajnik CS, Fall CH, Vaidya U, et al. Fetal growth and glucose metabolism in four-year-old Indian children. Diabet Med1995; 12 :330 –6.
29. Barker DJP, Eriksson JG, Forsén T, Osmond C. Fetal origins of adult disease: strength of effects and biological basis. Int J Epidemiol2002; 31 :1235 –9.
30. Barnard RJ, Roberts CK, Varon SM, Berger JJ. Diet-induced insulin resistance precedes other aspects of the metabolic syndrome. J Appl Physiol1998; 84 :1311 –5.
31. Barker DJP. Fetal origins of coronary heart disease. BMJ1995; 311 :171 –4.
32. Ericsson JG, Forsén T, Tuomilehto J, Jaddoe VWV, Osmond C, Barker DJP. Effects of size at birth and childhood growth on the insulin resistance syndrome in elderly individuals. Diabetologia2002; 45 :342 –8.
33. Mericq V, Ong KK, Bazaes R, et al. Longitudinal changes in insulin sensitivity and secretion from birth to age three years in small- and appropriate-for-gestational-age children. Diabetologia2005; 48 :2609 –14.
34. Wennlöf AH, Yngve A, Nilsson TK, Sjöström M. Serum lipids, glucose and insulin levels in healthy school children aged 9 and 15 years from central Sweden: reference values in relation to biological, social and lifestyle factors. Scand J Clin Lab Invest2005; 65 :65 –76.
35. Hagman U, Bruce Å, Persson LÅ, Samuelson G, Sjölin S. Food habits and nutrient intake in childhood in relation to health and socio-economic conditions. A Swedish multicentre study 1980–81. Acta Paediatr Scand Suppl1986; 328 :1 –56.
36. Prentice AM. Overeating: the health risks. Obes Res2001; 9 :234S –8S.
37. Nicklas TA, Baronowski T, Cullen KW, Berenson G. Eating patterns, dietary quality and obesity. J Am Coll Nutr2001; 6 :599 –608.
38. Cox JE, Tyler WJ, Randich A, et al. Suppression of food intake, body weight, and body fat by jejunal fatty acid infusions. Am J Physiol Regul Integr Comp Physiol2000; 278 :R604 –10.
39. Phifer CB, Berthoud H-R. Duodenal nutrient infusions differentially affect sham feeding and Fos expression in rat brain stem. Am J Physiol Regul Integr Comp Physiol1998; 274 :R1725 –33.
40. Payne JA, Belton NR. Nutrient intake and growth in preschool children I. Comparison of energy intake and sources with growth. J Hum Nutr Diet1992; 5 :287 –98.
41. Clarke SD, Gasperikova D, Nelson C, Lapillonne A, Heird WC. Fatty acid regulation of gene expression. A genomic explanation for the benefits of the Mediterranean diet. Ann N Y Acad Sci2002; 967 :283 –98.
42. Korotkova M, Gabrielsson BG, Holmäng A, Larsson BM, Hansson LÅ, Strandvik B. Gender-related long-term effects in adult rats by perinatal dietary ratio of n-6/n-3 fatty acids. Am J Physiol Regul Integr Comp Physiol2005; 288 :R575 –9.
43. Korotkova M, Ohlsson C, Gabrielsson B, Hanson LÅ, Strandvik B. Perinatal essential fatty acid deficiency influences body weight and bone parameters in adult male rats. Biochim Biophys Acta2005; 1686 :248 –54.