The current study aimed to examine the changes following a newly developed sleep-hygiene intervention in relation to sleep hygiene practices, sleep quality, and daytime symptoms in children and adolescents. We found significant improvements in sleep hygiene, sleep quality as well as daytime symptoms such as a reduction in daytime sleepiness and sedentary/light activity, following the intervention. We reported statistically significant improvements in sleep hygiene practices post-intervention which earlier studies have also demonstrated
[20, 24, 40]. The improvements in sleep practices (as assessed by the ASHS) reported in the current study, albeit significant, were small. However, although the improvements in sleep practices reported are slight, the statistically significant increase might indeed reflect real-life benefits as seen in the improvements in sleep quality reported by both parents and participants, as well as the decrease in daytime sleepiness.
The improvements in sleep hygiene and sleep quality reported in the current study are also consistent with previous studies that have shown similar findings in participants ranging from infants to adults
[21, 24, 40–42]. However given both participants and parents were aware of the treatment, a study of sleep hygiene intervention with a concurrent control group with parents blind to treatment would be required to provide definitive answers in regard to efficacy. Baseline data on sleep quality and daytime sleepiness indicated that the cohort of enrolled children had poor sleep quality and elevated index scores for daytime sleepiness. With regard to sleep quality, baseline mean (SD) PSQI scores of 7.47 (2.43) were greater than five, indicating poor sleep quality
 while baseline mean (SD) SDSC scores of 53.4 (9.0) were greater than 39 which was also an indication of poor sleep quality
. These scores were very similar to those reported in slightly older students with poor sleep quality
[20, 43] and baseline SDSC index scores were also very similar to those of patients with known sleep disorders such as insomnia and respiratory disturbances
. Baseline daytime sleepiness scores were also indicative of daytime sleepiness
[34, 38] and similar to those reported in children of a similar age with OSA and epilepsy
. The baseline mean sleep duration was within recommended sleep times for this age group
[1–3] which might be why the mean increase in total sleep time (after the intervention) was only five minutes.
A significant decrease in daytime sleepiness (PDSS), a daytime symptom of poor sleep quality, also suggests further evidence of improved sleep quality. Mean scores for PSQI, SDSC, and PDSS improved from scores indicative of sleep dysfunction or above control group means, to acceptable or control group mean scores. There was a trend (albeit non-significant) for energy expenditure during sleep to decrease post-intervention which could indicate this as a marker for the improved sleep quality reported subjectively. Furthermore, a significant decrease in participants’ sedentary/light activity behaviour during the day was also observed which could be due to the decrease in sleepiness during the day as observed in PDSS scores.
Encouragingly, we found a slight but significant reduction in BMI z-scores 20 weeks post-intervention. This seems to suggest that although significant increases in absolute weight and height were also observed, the reduction in BMI z-scores reflects a slower weight gain of our participants (compared to peers of the same age and sex) rather than an increase in height that led to improved BMI z-scores. The recruited sample of children, although not overweight, had BMI z-scores almost one SD above the population mean that reduced by −0.13 (CI −0.20 to 0.05, p = 0.001) to a mean (SD) BMI z-score of 0.66 (1.19) 20 weeks post-intervention. An earlier study, adopting a diet and physical activity behavioural intervention in children
 reported a slightly smaller reduction in BMI z-scores, as did a similar study in obese adolescents
. Compared to the two above mentioned studies, the reduction in BMI z-scores reported in the current study are important given the fact that our sample of children were not overweight/obese. It is possible improvements in sleep might have led to the decrease in BMI z-scores seen through the changes in the regulation of appetite hormones leptin and ghrelin
[47–49] and/or due to changes in fat metabolism
. However caution must be applied in linking this BMI reduction to improved sleep as rules around food restriction before bedtime were part of the sleep hygiene practice regime to encourage sleep and thus more research is needed to corroborate our findings around sleep and weight. A larger, behavioural sleep study investigating the health outcomes of obese adults after a behavioural sleep extension intervention of sleep deprived (< 6.5 hours) participants is currently underway
. It employs a non-pharmacological sleep only intervention (no modifications to energy intake or expenditure). Interim analysis of PSQI scores also revealed a greater reduction in PSQI scores in the sleep intervention group compared to the control group at the first year of follow-up
. These trends provide indications that sleep may be an effective adjunct treatment strategy for weight management.
This study was not without limitations. First, our sample was a small community-recruited sample, which limits generalizability. For example, as enrolment to the study was voluntary and dependent on the participant/family approaching us, this cohort of participants could have been especially motivated. Secondly, this was a before and after intervention study and did not include a control group, thus any improvements seen could be due to the placebo effect which might have been amplified by the participants’ motivation for change. Thirdly, although the measurement of sleep quality by participant using the PSQI is written in a language easily understood by this age group, the validity and reliability of the instrument in this young age group has yet to be demonstrated. Finally, it is likely that the self-selected recruitment of parents will have produced a sample of children with high levels of sleep problems. This was an intended effect of the recruitment strategies used in order to demonstrate that the pilot was acceptable for this population and so could be used in a larger trial. In the absence of a control group, however, regression towards the mean cannot be ruled out as at least partially explaining the changes observed during the study. The full study was completed over a school year, thus we cannot exclude seasonal changes being an advantage or disadvantage to sleep changes in some. A strength of the study was the retention rate at 100% with every enrolled participant completing the 20 week F.E.R.R.E.T sleep hygiene programme. This is probably atypical, but the high retention may have reflected the age-appropriateness of the intervention, perhaps aided by the multiple researcher contact time points and/or by the individuals being more motivated to complete because they volunteered and presented with self-identified sleep problems.