Sleep and obesity in children at different obesity risks : patterns, associations and early intervention

Abstract

Background: Childhood obesity is a global health concern with a range of adverse outcomes. Short sleep has been consistently linked to childhood obesity. However, associations between other sleep characteristics and obesity are less studied in children. Some multi-component obesity intervention studies har started included sleep as a target. Nevertheless, whether sleep can be improved and further benefit obesity prevention are still uncertain. Moreover, parental obesity is one of the major risk factors for child obesity, which probably due to both shared genetic factors and obesity-related environment and behaviours within families. Understanding whether and how sleep is involved in the obesity transfer within families has important implications for developing better interventions.

Aims: The overall aim of this thesis was to explore the role of sleep in the development of obesity among children at high and low obesity risks, determined by their parental weight, as well as to assess the effect of a long-term, low-intensive, family-based multi-component intervention on both preventing obesity and promoting sleep.

Materials: All studies included in this thesis are embedded in the Early Stockholm Obesity Prevention Project (Early STOPP), which was a long-term (5-year), low-intensive, family-based obesity prevention project conducted in Stockholm County. In Early STOPP, 181 children with two overweight or at least one obese parent (defined as at high obesity risk) were recruited and randomly allocated to the intervention (n=66) and control (n=115) groups. Meanwhile, a group of children (n=57) with normal-weight parents were recruited randomly as a reference group (defined as at low obesity risk). The baseline data was collected when the child was 1-year-old and an annual follow-up was conducted until the child was 6-year-old.

Methods: Study I is a cross-sectional study of Early STOPP baseline data, where child and parental sleep patterns were compared between high and low obesity risk groups, respectively. In Study II, an explorative study, changes of child sleep patterns from age 1 to 2 years were compared between risk groups. Furthermore, child weekday-weekend sleep variations at age 2 years and associated family factors were explored. Study III is a longitudinal study, where the development of child sleep patterns from age 2 to 6 years were studied and the association between child sleep and obesity was explored. Study IV is a randomised controlled trial (RCT) examining the effect of a long-term, low-intensive, family-based intervention on preventing obesity in children at high obesity risk, as well as effects on secondary behaviour outcomes, including child sleep, physical activity and eating behaviour.

Results: In Study I and II, some differences were observed in sleep patterns between children at different obesity risks. Compared to children at low obesity risk, children at high obesity risk had slightly longer sleep onset latency at both age 1 and 2 years, as well as greater weekday-weekend sleep variations in sleep schedules at age 2 years. Moreover, children in the high obesity risk group were more likely to experience unfavourable sleep characteristics than their peers in the low-risk group during the first two years of life, including more transient prolonged sleep onset latency and low sleep efficiency. In Study III, when child sleep was assessed using actigraphy, no difference was observed in either child sleep variables or the prevalence of unfavourable sleep characteristics from age 2 to 6 years. In Study III, from child 2 to 6 years, higher short sleep duration score was associated with a greater increase in BMI z-score (0.12, 95% confidence interval [CI]: 0.01-0.25) across ages. Independent of sleep duration, higher late sleep score was associated with greater increases in both BMI z-score (0.16, 95% CI: 0.05-0.27) and waist circumference (0.60 cm, 95% CI: 0.23-0.98) across ages. Moreover, a significant combined effect was detected between late sleep and family obesity risk on child weight gain, as children at high obesity risk and having habitual late sleep had the greatest increase in both BMI z-score (0.93, 95% CI: 0.40-1.45) and waist circumference (3.45 cm, 95% CI: 1.78-5.12) from age 2 to 6 years. In Study IV, during the follow-up period, no main intervention effect was identified on either primary outcome (BMI z-score) or other weight-related outcomes (weight and BMI) across ages. At age 6 years, there were 36 children with overweight and obesity, with 12 (23%) in the intervention group and 24 (26%) in the control group. The intervention was not significantly effective for reducing the risk of developing overweight and obesity. For the secondary outcomes, no significant intervention effect was detected in child eating behaviour, total sleep duration or average physical activity across ages.

Conclusion: Frequent exposures to short sleep and late sleep were independently associated with more increases in weight measures in children from age 2 to 6 years. Although the development of sleep was similar in children at different obesity risks, determined by parental weight, a combined effect between child late sleep and high family obesity risk on more weight gain was observed. Moreover, a five-year, low-intensive, family-based, multicomponent obesity intervention was not more effective than routine health care in either promoting behaviours or preventing obesity among children with overweight and obese parents.