In this study, we applied an extended ICH score to a large cohort of children aged 5-6 years and investigated to what extent this score was related to cardiometabolic health at age 11-12 years. The results were compared to the original ICH score. To our knowledge this was the first time that additional components were added to the original ICH score to examine if this would improve the prediction of cardiometabolic health in children.
We found that ICH score at age 5-6 was not significantly associated with CIMT at age 11-12 according to both the extended and original ICH definition. Both definitions showed associations of ICH score at age 5-6 with total cholesterol and BMI at age 11-12, while associations with systolic- and diastolic blood pressure were only found when using the extended definition. Both ICH-scores predicted subsequent overweight. When restricted to normal weight children at age 5-6 only, the extended ICH score still predicted overweight at age 11-12 whereas this association was no longer significant according to the original definition.
Interpretation of findings
STAGE 1
Ideal health factors were more common than ideal health behaviours, but most children aged 5-6 years still scored ideal on healthy weight status (87.6%), sleeping hours (81.4%) and screen time (80.4%). Extra attention is needed for a healthy diet, since only 2% of the children met the healthy diet criteria. In line with our expectations, health factors were more often ideal than health behaviours as ideal behaviours precede changes in health factors. Other studies on ICH in child populations have shown similar findings.9–12
The low prevalence of a healthy diet is in line with other studies on ICH.10–12 Another Dutch longitudinal study on the health of 8 year old Rotterdam born children found no children meeting Dutch dietary guidelines regarding intake of vegetables and sugar sweetened beverages.38 As the Dutch dietary cut-offs were fairly similar to the AHA diet guideline39, a similar percentage of ideal diet according to the AHA diet guideline can be expected. Dietary intake of both the Amsterdam and Rotterdam born children can thus be accounted as suboptimal.
In our study, the second most important ICH component was total cholesterol. We found 30.3% of 5-6 year old children with high total cholesterol concentrations. As lipoprotein concentrations change with growth and maturation (during puberty total cholesterol declines by 10-15%), it is questionable if the setpoint of cholesterol measurement at the age of 5-6 is representative for cardiovascular health. 3,40 Though in studies of children >12 years of age, the prevalence of high total cholesterol was 21.5% to 66.8%9,11,12, which indicate that our findings are similar to findings of studies that examine older children. Next, of the 559 children included in stage 2 of this study, 31.7% had high total cholesterol concentrations at age 5-6, which declined to 26.8% at age 11-12.
STAGE 2
Ideal cardiovascular health in childhood has shown to be associated with CIMT measurements in adulthood.9 Nevertheless, clustering of CVD risk factors starts in childhood and vascular changes may therefore also occur at younger ages. In our sample ICH score at age 5-6 was not related to CIMT measurements at follow-up six years later. The variation of CIMT in this group was small, which can be explained by the fact that we examined a relative young and healthy population of the same age. According to the systematic review of van den Munckhof et al.41 on CIMT in an apparently healthy adult population, 84% of CIMT can be explained by age. In contrast, several other studies in children found associations of ICH components and CIMT in childhood.14,15,37,42 The children included in these studies were mainly older at baseline (range 8 to 11) than our sample and CIMT was also measured at older ages (range 15 to 21) 14,15,37 However, the study by Geerts et al.42 found an association between prenatal exposure to smoking and CIMT measurements at the age of 5 years old. The inconsistency of findings may be explained by the way in which CIMT was measured; in our study a hand guided ultrasound was used, which has a lower sensitivity than automatic devices. In addition, we performed measurements at three different angles in one segment (common carotid artery), while other studies performed measurements at one angle in three different segments (common carotid artery, internal carotid artery, carotid bulb). Lastly, different sonographers can have different findings and all studies used different readers. This inevitably makes results difficult to compare between studies.
The extended ICH definition at age 5-6 has stronger associations with systolic- and diastolic blood pressure and BMI at age 11-12 than the original definition and remained when testing only in non-overweight individuals aged 5-6 years. This difference cannot be explained by only one of the three additional components used in the extended definition. The difference is thus due to the clustering of additional components. This might be because of the bi-directional relationship between sleep duration and physical activity and sedentary behavior.43 The implication for our study is that children with ideal levels of one of the three behaviours (sleep duration, physical activity or screen time) were more likely to have ideal levels of the other cardiovascular components, which resulted in a considerably higher ICH-level. As sleep duration, physical activity and screen time are linked, the extended definition gives a larger variance between ICH levels of those with and without ideal physical activity compared to the original ICH definition. Since there is a larger variation in ICH-level, associations of ICH-level with cardiometabolic outcomes are more likely. Furthermore, as unhealthy behaviour leads to changes in blood cholesterol concentrations which in turn leads to increased CIMT, it could be that adding more behaviours to the ICH construct will make the ICH construct more sensitive for predicting cardiovascular health.
The association between cardiometabolic outcomes and the additional components, namely: screen time, sleep behaviour and prenatal smoke exposure have been examined in previous studies.21,22,44 In children, associations between sedentary behaviour and blood pressure are inconsistent21 , although there is some evidence for a relationship between short sleep duration in children and increased blood pressure.22 Prenatal smoke exposure is not associated with an increased blood pressure in childhood according to a Portuguese study in 3-10 year old children (n ≈ 2500). However prenatal smoke exposure is associated with an increase in BMI (n = 17286).44 Sleep duration in childhood is also found to be inversely associated with BMI22, while the evidence for sedentary behaviour is inconsistent.21 It is possible that the addition of the extra components single-handedly led to the extra association. However, it might also be that the combination of more non-ideal health behaviours is the basis of reduced cardiometabolic health which is associated with higher levels of blood pressure and BMI.
Strengths and limitations
Our study has some important strengths, including the longitudinal study design and its large sample size. Measurements took place at the same timepoint, by the same researchers with the same measuring instruments. Further, extensive measurements were done with the validated FFQ at stage 1 in the study.24 The present study is the first to examine ICH in childhood with CIMT measurement in children as young as 11-12 years old.
The present study also has several limitations. First, our sample differed from the ABCD-cohort in terms of ethnicity, maternal educational level and family history of CVD. As supplementary Table 1 shows, precisely those participants are more likely to have a poor cardiovascular health. Therefore, our results probably overestimate the prevalence of ideal cardiovascular health in the ABCD-cohort. Not all children participating at age 5-6 also participated 5-6 years later, though the study cohort still consisted of 559 individuals. Information on health behaviours were reported by parents and thus all subjective. It is conceivable that parents perceive or present their child’s behaviour in a more positive manner than the actual situation is. This could lead to an overestimation of healthy behaviour. Health factors were measured during the health check and automatic oscillometric devices were used for blood pressure measurements, while the blood pressure table of the American Academy of Paediatrics was made by auscultatory measurements.35 Compared to blood pressure values measured by the auscultatory method, the oscillometric blood pressure device used in the present study found on average (SD) 4.6 (4.9) mmHg higher systolic blood pressure values and 3.3 (5.4) mmHg lower diastolic blood pressure values.32 This could have led to an overestimation of (pre-)hypertension because of higher levels of systolic blood pressure. However, a systematically overestimation of blood pressure values would not affect the present study. Many paediatric studies now use automatic devices in combination with the standard blood pressure tables45, thus findings of this study are comparable with other available research. For outcome measures, our study was limited to CIMT measurement and CVD risk factors since cardiovascular events could not be studied as clinical outcome at this age.
Implications
Our study implicates that to raise a healthy generation, there is need for dietary interventions in Dutch children. Adherence to the healthy diet guidelines among the children was extremely low, which is unfortunately since dietary habits have shown to track into adulthood.46
As the extended definition did not improve the prediction of CIMT and most cardiometabolic factors, we do not recommend the AHA to add the extra components to their definition yet. First there is need for further research.
In compliance with other studies,47,48 our study shows that children with a non-western ethnicity and lower educated mothers were more likely to have a poorer cardiovascular health, public health interventions regarding cardiovascular health should focus particularly on these at risk individuals.
Recommendations for future research
Further research is needed to gain insight into the associations between lifestyle of young children and cardiovascular health later in life. We recommend other researchers to look at ICH in children under 12 years in relation to cardiometabolic outcomes at a later age and explore the addition of sleep duration, screen time and prenatal smoke exposure to ICH on these outcomes. We also encourage other researchers to examine subpopulations (i.e. overweight children).
The ABCD-cohort will also be followed up longer and report on associations of ICH at age 5-6 and cardiometabolic health at an older age.