A literature review of evidence for primary prevention of overweight and obesity in healthy weight children and adolescents: A report produced by a working group of the Danish Council on Health and Disease Prevention.

Primary prevention targets development of overweight in individuals with healthy weight and is a great challenge. This paper summarizes the main findings of a working group of the Danish Council on Health and Disease Prevention that reviewed the literature on primary prevention of overweight and obesity among children and adolescents. The results were presented in a Danish report, in which a 2019 Cochrane review on childhood obesity prevention was complemented by searches in PubMed to include all relevant subsequent studies published from January 2018 until March 2020. In this paper, the review was updated until June 2023. Numerous childhood overweight prevention interventions have been developed during the past decades, primarily targeting diet and/or physical activity. Several of these interventions showed positive effects on diet and physical activity level but did not show effects on risk of developing overweight. The evidence foundation is inconsistent as four out of five interventions did not show positive effects. Previously observed intervention effects may not reflect excessive weight gain prevention among children with healthy weight but rather bodyweight reduction among those with overweight or obesity. We do not have sufficient knowledge about how to prevent children with healthy weight from developing overweight, and creative solutions are urgently needed.

Evidence for the protein leverage hypothesis in preschool children prone to obesity.

BACKGROUND & AIMS: The protein leverage hypothesis (PLH) proposed that strict regulation of protein intake drives energy overconsumption and obesity when diets are diluted by fat and/or carbohydrates. Evidence about the PLH has been found in adults, while studies in children are limited. Thus, we aimed to test the PLH by assessing the role of dietary protein on macronutrients, energy intake, and obesity risk using data from preschool children followed for 1.3 years. METHODS: 553 preschool children aged 2-6 years from the 'Healthy Start' project were included. EXPOSURES: The proportion of energy intake from protein, fat, and carbohydrates collected from a 4-day dietary record. OUTCOMES: Energy intake, BMI z-score, fat mass (FM) %, waist- (WHtR) and hip-height ratio (HHtR). Power function analysis was used to test the leverage of protein on energy intake. Mixture models were used to explore interactive associations of macronutrient composition on all these outcomes, with results visualized as response surfaces on the nutritional geometry. RESULTS: Evidence for the PLH was confirmed in preschool children. The distribution of protein intake (% of MJ, IQR: 3.2) varied substantially less than for carbohydrate (IQR: 5.7) or fat (IQR: 6.3) intakes, suggesting protein intake is most tightly regulated. Absolute energy intake varied inversely with dietary percentage energy from protein (L = -0.14, 95% CI: -0.25, -0.04). Compared to children with high fat or carbohydrate intakes, children with high dietary protein intake (>20% of MJ) had a greater decrease in WHtR and HHtR over the 1.3-year follow-up, offering evidence for the PLH in prospective analysis. But no association was observed between macronutrient distribution and changes in BMI z-score or FM%. CONCLUSIONS: In this study in preschool children, protein intake was the most tightly regulated macronutrient, and energy intake was an inverse function of dietary protein concentration, indicating the evidence for protein leverage. Increases in WHtR and HHtR were principally associated with the dietary protein dilution, supporting the PLH. These findings highlight the importance of protein in children's diets, which seems to have significant implications for childhood obesity risk and overall health.

Attained body mass index among children attending rural outdoor or urban conventional kindergartens.

Objective: This study aimed to examine whether children in rural outdoor kindergartens had attained a lower body mass index z-score (BMIz) and were at lower risk of overweight after school entrance compared to children in urban conventional kindergartens. Methods: This is a longitudinal observational study of 1,544 children from outdoor kindergartens and 1,640 from conventional kindergartens. The mean age at kindergarten enrolment was 3.5 years (SD: 0.9) in the outdoor kindergartens and 3.6 years (SD: 1.0) in the conventional kindergartens. Anthropometry was measured after school entry by school health nurses when the children were 6 to 8 years old. Attained BMIz was included as the primary outcome. The risk of attaining overweight (including obesity) was included as a secondary outcome. Register-based information was available on potential confounding factors. Linear and logistic regression models were used to assess group differences in outcome measures. Results: Our basic models, with information on outcome, kindergarten type, and birth weight showed a borderline statistically significantly lower attained BMIz (-0.07 [95% CI: -0.14, 0.00], P = 0.060) and a lower risk of overweight (adjusted risk ratio: 0.83 [95% CI: 0.72, 0.97], P = 0.016) among children attending outdoor kindergartens. However, when adjusting for sociodemographic factors and parental BMI, there was no evidence of differences in attained BMIz (P = 0.153) or overweight (P = 0.967). Conclusion: When considering confounding factors, our findings indicate no differences in attained BMIz or risk of overweight after school entry among children attending rural outdoor kindergartens compared to those attending urban conventional kindergartens.

Familial associations in hair cortisol concentration: A cross-sectional analysis based on the Healthy Start study.

A few studies have shown a direct association between maternal and child hair cortisol concentrations (HCC), but the potential correlations within paternal-child and maternal-paternal dyads are not clear from the current evidence. Thus, we aimed to thoroughly examine associations between family members HCC. We conducted a cross-sectional study based on 159 children and their parents (159 mothers and 159 fathers) participating in the Danish Healthy Start Study. Information was available on HCC from both children, mothers and fathers, as well as on several sociodemographic factors. First, Pearson's correlation coefficients were applied to assess crude correlations between family members' HCC. Secondly, analysis of covariance, adjusted for covariates, was applied to estimate child mean HCC in quartiles of maternal and paternal HCC, and mean paternal HCC in quartiles of maternal HCC. Our results showed direct associations between HCC of all family members. We found statistically robust correlations between maternal and child HCC (r = 0.33; P < 0.001), paternal and child HCC (r = 0.37; P < 0.001) and between maternal and paternal HCC (r = 0.31; P < 0.001). Similar results were found when adjusting for covariates in analyses of covariance. Our data provides evidence of associations between family members' HCC. However, we were unable to determine the extent to which these associations were due to shared genetics, assortative mating or environmental factors.

Associations between objective measures of physical activity, sleep and stress levels among preschool children.

BACKGROUND: Cortisol is often used as a biological marker for stress. When measured in urine or serum, representing a short-term measurement of the hormone, it has been associated with unfavorable sleep characteristics and both low and high physical activity levels. However, cortisol in hair represents a long-term stress measure and has been suggested as a promising new marker for chronic stress. Therefore, we aimed to examine the association between objectively measured sleep, physical activity and hair cortisol levels in preschool children. METHODS: In order to obtain objective measures of physical activity and sleep habits, 54 children aged 2-6 years wore an ActiGraph for 5 consecutive days and nights. For chronic stress measurements of each child, hair was cut from the back of the head close to the scalp for analysis of cortisol levels. Associations between measured sleep quality and quantity and level of physical activity and hair cortisol levels were estimated using linear regression analysis, presented as ß. Results were adjusted for sex, age and BMI z-score. RESULTS: We found no significant association between log-transformed cortisol (pg/mg) and sleep duration (hours) (ß = - 0.0016, p = 0.99), sleep efficiency (ß = - 3.1, p = 0.18), sleep latency (ß = 0.015, p = 0.16) or physical activity level (100 counts per min) (ß = 0.014, p = 0.22). However, sleep latency (min) was directly associated with physical activity (counts per min) levels (ß = 35.2, p = 0.02), while sleep duration (hours) (ß = - 142.1, p = 0.55) and sleep efficiency (%) (ß = - 4087, p = 0.26) showed no significant associations. CONCLUSIONS: In our study, a high physical activity level was associated with poorer sleep habits. Neither sleep quality nor physical activity were related to long term cortisol exposure. These results are among the first to study associations between objectively measured sleep, physical activity and chronic cortisol levels among preschool children. More and larger studies are therefore needed.

Child behaviour and subsequent changes in body weight, composition and shape.

OBJECTIVE: Studies have found an association between child behavioural problems and overweight, but the existing evidence for this relationship is inconsistent, and results from longitudinal studies are sparse. Thus, we examined the association between behavioural problems and subsequent changes in body mass index (BMI) and anthropometry over a follow-up period of 1.3 years among children aged 2-6 years. DESIGN: The study was based on a total of 345 children from The Healthy Start Study; all children were healthy weight but predisposed to develop overweight. The Danish version of the Strengths and Difficulties Questionnaire (SDQ), classified as SDQ Total Difficulties (SDQ-TD) and SDQ Prosocial Behaviour (SDQ-PSB), was used to assess child behaviour. Linear regression analyses were used to examine associations between SDQ scores and subsequent change in BMI z-score, body fat percentage, waist circumference and waist-hip ratio, while taking possible confounding factors into account. RESULTS: We found an association between SDQ-PSB and subsequent change in BMI z-score (ß: 0.040 [95% CI: 0.010; 0.071, p = 0.009]). However, there was no evidence of an association between SDQ-PSB and measures of body composition or body shape. CONCLUSIONS: Among 2 to 6 years old children predisposed to overweight, the association between SDQ-scores and weight gain is either absent or marginal. The SDQ-PSB score may be associated with subsequent increases in BMI z-score, but this association does not seem driven by an increased relative fat accumulation.

Association between hair cortisol concentration and dietary intake among normal weight preschool children predisposed to overweight and obesity.

BACKGROUND: The association between chronically elevated cortisol, as measured by hair cortisol concentration (HCC), and dietary intake among children has generally not been explored. Moreover, it is unknown whether there is an association between parental HCC and dietary intake among their children. OBJECTIVE: To examine associations between HCC and dietary intake among children, and to explore the association between parental HCC and dietary intake among their children. METHODS: We conducted a cross-sectional study based on 296 children predisposed to overweight and obesity who participated in the Healthy Start study. Multiple Linear regression analyses were conducted to assess the association between HCC and total energy intake, macronutrients, fruit and vegetables, added sugar, sugar-sweetened beverages (SSB), and a diet quality index (DQI). RESULTS: Among the children, we found that higher HCC was associated with a lower consumption of dietary fat (ß: -0.7 g/day [95% CI: -1.3, -0.0] per 100 pg/mg HCC). We found no statistically significant association between HCC and intake of total energy, protein, carbohydrate, fruit and vegetables, added sugar, SSB or DQI. We found no association between parental HCC and intake of total energy, added sugar, selected food groups or DQI among their children. However, stratified analyses showed that paternal HCC was associated with a borderline significant lower total energy intake and significantly lower protein intake, but only among daughters (adjusted ß: -42 kcal/day [95% CI: -85, 0] and -2.6 g/day [95% CI: -4.4, -0.8] per 100 pg/mg HCC, respectively). CONCLUSION: Among children, chronic stress as measured by HCC may be associated with a lower fat consumption, and paternal HCC may be associated with a lower intake of energy and protein among their daughters. However, the associations observed were weak, and any clinical relevance of these findings remains questionable.

Effects of the Healthy Start randomized intervention on dietary intake among obesity-prone normal-weight children.

OBJECTIVE: The study aimed to evaluate the impact of a 15-month intervention on dietary intake conducted among obesity-prone normal-weight pre-school children. DESIGN: Information on dietary intake was obtained using a 4 d diet record. A diet quality index was adapted to assess how well children's diet complied with the Danish national guidelines. Linear regression per protocol and intention-to-treat analyses of differences in intakes of energy, macronutrients, fruit, vegetables, fish, sugar-sweetened beverages and diet quality index between the two groups were conducted. SETTING: The Healthy Start study was conducted during 2009-2011, focusing on changing diet, physical activity, sleep and stress management to prevent excessive weight gain among Danish children. SUBJECTS: From a population of 635 Danish pre-school children, who had a high birth weight (≥4000 g), high maternal pre-pregnancy BMI (≥28·0 kg/m2) or low maternal educational level (<10 years of schooling), 285 children completed the intervention and had complete information on dietary intake. RESULTS: Children in the intervention group had a lower energy intake after the 15-month intervention (group means: 5·29 v. 5·59 MJ, P=0·02) compared with the control group. We observed lower intakes of carbohydrates and added sugar in the intervention group compared with the control group after the intervention (P=0·002, P=0·01). CONCLUSIONS: The intervention resulted in a lower energy intake, particularly from carbohydrates and added sugar after 15 months of intervention, suggesting that dietary intake can be changed in a healthier direction in children predisposed to obesity.

Hair dyeing, hair washing and hair cortisol concentrations among women from the healthy start study.

BACKGROUND: Hair cortisol concentration (HCC) has been suggested as a promising marker for chronic stress. However, studies investigating the influence of hair dyeing and hair washing frequency on HCC have shown inconsistent results. OBJECTIVE: To examine associations between HCC and hair dyeing status or weekly hair washing frequency among women. METHODS: This cross-sectional study was based on data from 266 mothers participating in the Healthy Start intervention study. HCC was measured in the proximal end of the hair (1-2cm closest to the scalp) while hair dyeing status, frequency of hair washing and covariates were reported by the women. Linear regression analyses were applied to assess the associations between HCC and hair dyeing or weekly frequency of hair washing. RESULTS: No statistically significant difference (p=0.91) in HCC was found between women who dyed hair (adjusted mean: 137pg/mg [95% CI: 122,153]) and women with natural hair color (adjusted mean: 139pg/mg [95% CI: 123,155]). Frequency of hair washing was not associated with HCC (ß: -3.7 [95% CI: -9.0, 1.5; P=0.20]). CONCLUSIONS: This study of 266 Danish women provides no evidence in support of an association between HCC and hair dyeing status or hair washing frequency.

Association between Hair Cortisol Concentration and Adiposity Measures among Children and Parents from the "Healthy Start" Study.

BACKGROUND: Previous studies have suggested a direct association between hair cortisol concentration (HCC) and Body Mass Index (BMI), as well as other adiposity measures. However, these studies have mostly been conducted among adult populations. OBJECTIVE: To examine the association between HCC and different measures of adiposity among a selected group of children predisposed to obesity and their parents. METHODS: We conducted a cross-sectional study based on 363 children and their parents (301 mothers and 231 fathers) participating in the "Healthy Start" study. Linear regression analysis was used to investigate associations between HCC and adiposity measures while taking into account possible confounding factors. Analyses were performed examining the association between HCC and BMI, fat mass and fat free mass index Z-scores, as well as waist circumference and waist-hip ratio among the children. Likewise, the association between HCC and BMI among the parents was explored. Finally, we examined the association between parental HCC and children's adiposity measures. RESULTS: HCC was directly associated with a higher BMI among the fathers (0.49 kg/m2 [95% CI: 0.09, 0.90, P = 0.02] per 100 pg/mg) and the mothers (0.93 kg/m2 [95% CI: 0.24, 1.61, P = 0.01] per 100 pg/mg). We found no clear evidence of an association between HCC and adiposity measures among children. However, a high maternal HCC was associated with a high fat mass index and low fat free mass index z-score in the offspring (0.14 SD [95% CI: 0.02, 0.26, P = 0.02] and -0.17 SD [95% CI: -0.30, -0.05, P = 0.01] per 100 pg/mg, respectively). CONCLUSIONS: Our study found no evidence of an association between HCC and measures of adiposity among children predisposed to obesity. However, HCC may be directly associated with BMI among men and women, and maternal HCC may be related to a higher fat mass and a lower fat free mass among their children.

Interactions between genetic variants associated with adiposity traits and soft drinks in relation to longitudinal changes in body weight and waist circumference.

BACKGROUND: Intake of sugar-sweetened beverages is associated with obesity, and this association may be modified by a genetic predisposition to obesity. OBJECTIVE: We examined the interactions between a molecular genetic predisposition to various aspects of obesity and the consumption of soft drinks, which are a major part of sugar-sweetened beverages, in relation to changes in adiposity measures. DESIGN: A total of 4765 individuals were included in the study. On the basis of 50 obesity-associated single nucleotide polymorphisms that are associated with body mass index (BMI), waist circumference (WC), or the waist-to-hip ratio adjusted for BMI (WHRBMI), the following 4 genetic predisposition scores (GRSs) were constructed: a complete genetic predisposition score including all 50 single nucleotide polymorphisms (GRSComplete), a genetic predisposition score including BMI-associated single nucleotide polymorphisms (GRSBMI), a genetic predisposition score including waist circumference-associated single nucleotide polymorphisms (GRSWC), and a genetic predisposition score including the waist-to-hip ratio adjusted for BMI-associated single nucleotide polymorphisms (GRSWHR). Associations between soft drink intake and the annual change (Δ) in body weight (BW), WC, or waist circumference adjusted for BMI (WCBMI) and possible interactions with the GRSs were examined with the use of linear regression analyses and meta-analyses. RESULTS: For each soft drink serving per day, soft drink consumption was significantly associated with a higher ΔBW of 0.07 kg/y (95% CI: 0.01, 0.13 kg/y; P = 0.020) but not with the ΔWC or ΔWCBMI In analyses of the ΔBW, we showed an interaction only with the GRSWC (per risk allele for each soft drink serving per day: -0.06 kg/y; 95% CI: -0.10, -0.02 kg/y; P = 0.006). In analyses of the ΔWC, we showed interactions only with the GRSBMI and GRSComplete [per risk allele for each soft drink serving per day: 0.05 cm/y (95% CI: 0.02, 0.09 cm/y; P = 0.001) and 0.05 cm/y (95% CI: 0.02, 0.07 cm/y; P = 0.001), respectively]. Nearly identical results were observed in analyses of the ΔWCBMI CONCLUSIONS: A genetic predisposition to a high WC may attenuate the association between soft drink intake and BW gain. A genetic predisposition to high BMI as well as a genetic predisposition to high BMI, WC, and WHRBMI combined may strengthen the association between soft drink intake and WC gain. However, the public health impact may be limited.

Child behavioural problems and body size among 2-6 year old children predisposed to overweight. results from the "healthy start" study.

OBJECTIVE: Psychological adversities among young children may be associated with childhood overweight and obesity. We examined if an increased level of child behavioural problems was associated with body size among a selected group of 2-6 year old children, who were all predisposed to develop overweight. METHODS: Cross-sectional analyses were conducted using baseline data from the "Healthy Start" intervention study. A total of 3058 children were invited to participate, and data from 583 children who were all predisposed for obesity was analyzed. The Danish version of the Strengths and Difficulties Questionnaire (SDQ) was used to assess child stress by the SDQ Total Difficulties (SDQ-TD) score and the Prosocial Behavior (PSB) score. Height and weight were measured, and BMI z-scores were calculated. RESULTS: A direct, but non-significant linear trend was found between SDQ-TD score and BMI z-score (ß = 0.021, p = 0.11). Having an SDQ-TD score above the 90(th) percentile was associated with BMI z-score (ß = 0.36, p = 0.05). PSB score was not associated with BMI z-score. Analyses were adjusted for parental socioeconomic status, parental BMI, family structure, dietary factors, physical activity, and family stress level. CONCLUSION: The results suggested a threshold effect between SDQ-TD score and BMI z-score, where BMI z-score was associated with childhood behavioural problems only for those with the highest scores of SDQ-TD. No significant association between PSB score and BMI z-score was found.