Long-term sleep and cardiovascular outcomes in children with sleep disordered breathing: a four year follow-up
2017-05-18T03:07:15Z (GMT) by
There is much more to snoring than harsh noisy breathing and a sleep-deprived bed-partner. Snoring is the cardinal symptom of sleep disordered breathing (SDB) – a common condition caused by an upper airway obstruction during sleep. It exhibits a spectrum of severity with primary snoring (PS) at the mildest end of the spectrum and obstructive sleep apnoea (OSA) at the most severe end. Although it was only reported in the medical literature in the late 1970s, childhood SDB is now recognised as a common condition with the major cause being the presence of large tonsils and adenoids. All severities of SDB have been associated with adverse sequelae including behavioural problems, neurocognitive deficits, poor quality of life and cardiovascular abnormalities. Our group has previously shown that primary school children (aged 7-12 y) with all SDB severities exhibit elevated blood pressure and abnormal blood pressure and heart rate control compared to their non-snoring counterparts. This is concerning as we know that high blood pressure in children can predict hypertension in adulthood. In addition, adults with SDB are at increased risk of developing hypertension as well as adverse cardiovascular events such as myocardial infarction and stroke. The first-line treatment for childhood SDB is the surgical removal of the tonsils and adenoids known as adenotonsillectomy (T&A) and it is most commonly reserved for children with moderate to severe OSA. Whilst a number of studies have assessed the outcome of T&A on SDB, few have assessed the impact of treatment on cardiovascular outcomes such as blood pressure and blood pressure control. Existing studies are short-term (<1 y follow-up) and generally focus on children with more severe OSA. Cardiovascular outcomes in children who are not treated, such as those with mild SDB, are currently unknown. Hence the aim of this PhD project was to assess the long-term sleep, respiratory and cardiovascular outcomes of SDB in children with all severities of SDB over a long follow-up period of four years. In addition to these aims, we also had the unique opportunity of validating a new surrogate measure of blood pressure, pulse transit time. To achieve these aims, we followed up our previously studied cohort of primary school-aged children (now aged 11-16 y). Children had a repeat polysomnography (PSG) four years after their initial diagnosis. In addition to the normal PSG recording, they had their blood pressure measured continuously and non-invasively during the night using a FinometerTM, a device which operates using a small blood pressure cuff on the finger. Sleep, respiratory and cardiovascular outcomes were compared between baseline and follow-up, and between three groups – non-snoring control children, children with resolved SDB (an obstructive apnoea hypopnoea index (OAHI) ≤1 event/h and no habitual snoring) and children with unresolved SDB (an OAHI >1 event/h and/or the presence of habitual snoring). In Chapter 3, we assessed sleep and respiratory outcomes in our three groups at four years follow-up. The non-snoring control children had no change in any sleep and respiratory parameters from baseline to follow-up. We identified that 46% of children had a complete resolution of SDB at follow-up, however, over half (54%) had persisting or unresolved SDB. Whilst most of the children in the unresolved group had PS, 18% showed worsening of OSA or newly developed OSA. Both children with resolved and unresolved SDB had a significant improvement in respiratory parameters from baseline to follow-up, in particular the OAHI. Children in the unresolved group showed worsened sleep parameters and had no change in their self-reported sleepiness. Children in the resolved group, however, demonstrated a significant improvement in their sleep as well as their self-reported sleepiness. An improvement in SDB severity was also accompanied by an improvement in blood pressure at follow-up. In Chapter 4, we demonstrated that although there was no difference over time in blood pressure in the control group, both children with resolved and unresolved SDB showed a significant fall in blood pressure (by 5-15 mmHg), from baseline to follow-up. The fall in blood pressure was associated with an improvement in nocturnal oxygen saturation, and blood pressure in both SDB groups was now no different to control children at follow-up. Our following study also assessed blood pressure control. At baseline, children with OSA had decreased baroreflex sensitivity (BRS) and all severities of SDB showed increased blood pressure variability (BPV) compared to non-snoring control children, both indicative of abnormal blood pressure control. In Chapter 5, we found a significant decrease in BPV in both children with resolved and unresolved SDB from baseline to follow-up, but no change in BRS in any of the groups. However, a sub-analysis revealed that children originally diagnosed with OSA at baseline who showed either an improvement to PS or complete resolution of SDB had a significant increase in BRS at follow-up. The improvement in both BPV and BRS from baseline to follow-up in these children was associated with the improvement in OAHI. We continued our investigation of cardiovascular control in our cohort by assessing changes in heart rate control from baseline to follow-up using heart rate variability (HRV). Our baseline study revealed reduced HRV in children with all severities of SDB. In Chapter 6, we found a significant reduction in HRV in all groups (control, resolved and unresolved groups). This decrease may reflect an age-related phenomenon in these children. There was no difference in HRV between the three groups at follow-up. Finally, this PhD project also gave us the opportunity to validate pulse transit time (PTT) as a surrogate measure of blood pressure. Whilst studies in adults show a significant correlation between PTT and blood pressure, studies in children have been limited to wake, and have only compared PTT to clinic blood pressure values (not a continuous measure of blood pressure). In Chapter 7, we demonstrated a strong inverse and linear relationship between PTT and continuously measured blood pressure in children during sleep. Furthermore, our linear mixed model analysis indicated that our PTT-predicted blood pressure model very closely tracked actual blood pressure over time. Thus we propose that PTT is an easy, valid and non-invasive surrogate measure of blood pressure change suitable for use in children and feasible in a clinical setting. Taken together, the findings from the studies presented in this PhD suggest that an improvement in SDB severity is associated with reduced blood pressure and heart rate, as well as improved control of both blood pressure and heart rate. Our findings suggest that SDB-associated cardiovascular abnormalities in childhood are potentially reversible, and that even an improvement (as opposed to complete resolution) in SDB can have a positive impact on the cardiovascular system. Thus this series of studies highlights the importance of continued monitoring of these children to ensure there is either a resolution or at least an improvement in their SDB severity, as this may avoid the development and progression of deleterious cardiovascular outcomes in the future.
Awards: Winner of the Mollie Holman Doctoral Medal for Excellence, Faculty of Medicine, Nursing and Health Sciences, 2013.