Letters

Treating children with sleep disorders

BMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7250.1667 (Published 17 June 2000) Cite this as: BMJ 2000;320:1667

Children with breathing difficulties are being overlooked

  1. Mary Fay, general paediatrician (mfay{at}enteract.com)
  1. 2303 Glenview Avenue, Park Ridge, IL 60068
  2. 4294 Shihmen Drive, Antioch, TN 37013, USA
  3. Department of General Practice, University of Adelaide, South Australia
  4. Adelaide Women's and Children's Hospital Adelaide, South Australia
  5. Department of Psychology, University of Hertfordshire, Hatfield AL10 9AB
  6. University of Oxford Section of Child and Adolescent Psychiatry Park Hospital for Children, Oxford OX3 7LQ
  7. Child and Family Psychiatric Service, Sue Nicholls Centre, Manor House, Aylesbury HP20 1EG
  8. Florence Nightingale House, Aylesbury HP21 8AL

    EDITOR—I was disappointed to see that once again children with sleep disorders are being lumped into a homogeneous group of children with “behavioural” problems, then “studied” without using polysomnography. I think it's presumptuous and dangerous to think that every child who has trouble sleeping has a behavioural problem. Few physicians are aware that breathing difficulty can cause night waking and bedtime resistance, and it is because of studies like that of Ramchandani et al.1

    Doctors don't look for sleep disorders properly, don't know much about them, and are told repeatedly that behavioural treatments are the appropriate treatment. This is wrong unless we can guarantee that the children have behavioural problems. A study looking at obese children using polysomnography diagnosed unsuspected obstructive sleep apnoea in 75% of the children studied—meaning the physicians examining the children picked up only 1 in 4 cases. It therefore seems a big leap of faith not to be doing polysomnography more often. Some of the children waking at night could have upper airways resistance syndrome, which can also cause sleep problems. Most doctors can barely take a proper history for sleep disorders, let alone diagnose a subtle case of upper airways resistance syndrome causing repeated night waking or bedtime resistance. Studies like that of Ramchandani et al certainly don't help because they take the emphasis away from finding a “real physiological” problem and on to stress behavioural modification.

    Before we continue to send physicians out to behaviourally modify breathing difficulties or other subtle causes of sleep disorders that can mimic behavioural problems, we need to do polysomnography on these children.

    References

    1. 1.

    Night waking is natural behaviour

    1. Zan Buckner, breast feeding counsellor
    1. 2303 Glenview Avenue, Park Ridge, IL 60068
    2. 4294 Shihmen Drive, Antioch, TN 37013, USA
    3. Department of General Practice, University of Adelaide, South Australia
    4. Adelaide Women's and Children's Hospital Adelaide, South Australia
    5. Department of Psychology, University of Hertfordshire, Hatfield AL10 9AB
    6. University of Oxford Section of Child and Adolescent Psychiatry Park Hospital for Children, Oxford OX3 7LQ
    7. Child and Family Psychiatric Service, Sue Nicholls Centre, Manor House, Aylesbury HP20 1EG
    8. Florence Nightingale House, Aylesbury HP21 8AL

      EDITOR—The basic assumption of the study of Ramchandani et al that very young children should settle down to sleep away from their parents and not be seen or heard from again until morning is flawed.1 In many non-Western societies this type of sleep pattern would be seen as an aberration.

      Anthropologist Carol Wortham of Emory University in Atlanta uncovered a variety of sleep patterns in 10 traditional hunter-gatherer societies, none of which coincide with our society's current ideal of sleeping alone without waking alone for the entire night.2 Perhaps some amount of night waking in children and their desire for comforting by parents are both natural and healthy. Our difficulty in combating this behaviour may arise because we are struggling against nature and basic human biology. We must be careful not to confuse that which is desirable for human health with that which is merely a current preference and expectation of our society.

      References

      1. 1.
      2. 2.

      Early intervention increases sleep times in young babies

      1. Brian Symon, senior lecturer (bsymon{at}medicine.adelaide.edu.au),
      2. John Marley, professor of general practice,
      3. James Martin, director of respiratory medicine
      1. 2303 Glenview Avenue, Park Ridge, IL 60068
      2. 4294 Shihmen Drive, Antioch, TN 37013, USA
      3. Department of General Practice, University of Adelaide, South Australia
      4. Adelaide Women's and Children's Hospital Adelaide, South Australia
      5. Department of Psychology, University of Hertfordshire, Hatfield AL10 9AB
      6. University of Oxford Section of Child and Adolescent Psychiatry Park Hospital for Children, Oxford OX3 7LQ
      7. Child and Family Psychiatric Service, Sue Nicholls Centre, Manor House, Aylesbury HP20 1EG
      8. Florence Nightingale House, Aylesbury HP21 8AL

        EDITOR—Ramchandani et al's systematic review of treatments for established infant sleep problems shows that behavioural interventions improve sleep patterns.1 The authors reference the frequency of sleep disorders as 20% in children aged 13 years, the tendency for these problems to be persistent, and their associations with behavioural difficulties and family disharmony. Other authors quote the frequency of sleep disturbance to be as high as 35%.2 Armstrong reported that 60% of parents were getting up at night when their child was 12 months old, and half of them were rising three times a night or more.3

        We tested the impact of a short behavioural intervention in normal neonates in a randomised trial (submitted for publication). In all, 269 families with normal newborn singletons were recruited within the first three weeks of life and were randomly allocated to intervention or control groups. Families in the intervention group received a single 45 minute tutorial from a research nurse. The tutorial emphasised that in children who were healthy and gaining weight satisfactorily sleep achievement should be regarded as a “learnt skill” influenced by environmental factors or “cues of sleep.” Parents were encouraged to avoid fatigue in their infants by not overhandling and to establish cues of sleep independent of the parent. Outcomes were measured through a sleep diary at 6 weeks of age.

        All aspects of sleep performance were significantly improved in the intervention group. Total sleep increased by almost nine hours a week at 6 weeks (P0.0001). The proportion of infants sleeping 15 hours a day was 61% in the intervention group and 28% among controls (P0.0001). 78% of sleep cycles among infants in the intervention group included over eight hours of night sleep, compared with 61% of cycles among controls (P 0.0001).

        Although the technique emphasised minimising parental handling at sleep time, there was no difference in the amount of crying between the groups. Follow up at 3 months of age showed that improvements were maintained. Our findings therefore support, and add to, those reported in the systematic review.

        Although Ramchandani et al did a well designed systematic review, our unpublished work was not identified, and it would be unreasonable to expect it to have been. This experience therefore supports the calls for a register of planned and ongoing randomised trials,4 and it also shows the importance of regularly updating systematic reviews through the Cochrane Collaboration as new evidence becomes available.

        References

        1. 1.
        2. 2.
        3. 3.
        4. 4.

        Treatment of child sleeping problems and the quality of trials are important

        1. Dieter Wolke, research professor of psychology (D.F.H.Wolke{at}herts.ac.uk)
        1. 2303 Glenview Avenue, Park Ridge, IL 60068
        2. 4294 Shihmen Drive, Antioch, TN 37013, USA
        3. Department of General Practice, University of Adelaide, South Australia
        4. Adelaide Women's and Children's Hospital Adelaide, South Australia
        5. Department of Psychology, University of Hertfordshire, Hatfield AL10 9AB
        6. University of Oxford Section of Child and Adolescent Psychiatry Park Hospital for Children, Oxford OX3 7LQ
        7. Child and Family Psychiatric Service, Sue Nicholls Centre, Manor House, Aylesbury HP20 1EG
        8. Florence Nightingale House, Aylesbury HP21 8AL

          EDITOR—Ramchandani et al's finding that behavioural treatments are effective in treating sleep problems will inform practice but omitted three important issues.1

          Firstly, two of the nine identified treatment studies included infants as young as 1 month. Infants are not born with the ability to sleep through the night, but sleep patterns are entrained by environmental cues.2 Waking at night in the early months of life is an adaptive behaviour to secure frequent feeding and to allow for rapid weight gain and brain growth. Sleeping through the night in the early weeks of life and a lack of waking for feeds is likely to put vulnerable infants (such as preterm infants or poor feeders) at risk of poor appetite development and suboptimal growth.3 Behavioural treatments for night waking should thus not start before 6 months of age.

          Secondly, infants who are breast fed do sleep less during 24 hours 4 and are more likely to wake at night than bottle fed infants.3 Parents should be informed that breast feeding does not impede infants from learning to sleep through the night. Early routines, avoidance of feeding the infant into “sleep submission,” and the use of focal feeds can, however, help to reduce the stress of night time breast feeding.

          Thirdly, sleeping problems often coexist with excessive crying or feeding difficulties.2 Parents with infants with multiple difficulties are more likely to seek help. Studies are still required to establish that behavioural interventions are effective in infants with multiple problems.

          Finally, quality criteria developed for drug trials were used to judge the quality of the behavioural treatments.1 These include double blinding to guard against performance bias and randomisation to exclude selection bias. Double blinding is not possible in behavioural studies. Therapists have to know what they are doing and why. Randomisation in small samples (less than 30) is often not effective as the effect sizes may become inflated because of outliers and there may be pretreatment differences. Quality ratings are there to judge the internal validity of a trial—that is, the confidence that the cause of any observed differences is due to the treatment conditions. The increasing use of “off the shelf” criteria for judging the quality of behavioural treatments is of doubtful validity. There are specific methods to guard against performance bias in behavioural trials such as the investigation of generalised placebo effects. Checking and controlling for a range of pretreatment differences may be as important as randomisation itself.5

          References

          1. 1.
          2. 2.
          3. 3.
          4. 4.
          5. 5.

          Authors' reply

          1. Luci Wiggs, research psychologist,
          2. Gregory Stores, professor,
          3. Paul Ramchandani, specialist registrar,
          4. Vicky Webb, general practitioner
          1. 2303 Glenview Avenue, Park Ridge, IL 60068
          2. 4294 Shihmen Drive, Antioch, TN 37013, USA
          3. Department of General Practice, University of Adelaide, South Australia
          4. Adelaide Women's and Children's Hospital Adelaide, South Australia
          5. Department of Psychology, University of Hertfordshire, Hatfield AL10 9AB
          6. University of Oxford Section of Child and Adolescent Psychiatry Park Hospital for Children, Oxford OX3 7LQ
          7. Child and Family Psychiatric Service, Sue Nicholls Centre, Manor House, Aylesbury HP20 1EG
          8. Florence Nightingale House, Aylesbury HP21 8AL

            EDITOR—We are aware of Fay's general point that it is the underlying sleep disorder rather than the presenting complaint at which treatment should be directed. She suggests that polysomnography is indicated for all children presenting with a sleep problem. We consider it more important that careful clinical inquiry is made to identify children with wide ranging physical conditions or other sleep disorders associated with sleep disturbance rather than to focus on the few conditions that can be identified by polysomnography. Furthermore, the appropriate diagnostic criteria for paediatric sleep related breathing disorders are being frequently refined (adult criteria, which have been used in the past, are not appropriate), and polysomnography is just one of the clinical investigations that may be required to diagnose these conditions.1

            All but one of the studies in our review (one study did not explicitly state screening procedures) included a broad assessment of the children's overall sleep pattern and medical history, including careful screening for symptoms suggestive of other sleep disorders and medical conditions. In the absence of any clinical signs of sleep related breathing disorders, and with clear indication of behavioural aspects maintaining the child's sleep problem, it would be difficult to justify (both to parents and to healthcare providers) the need for potentially unhelpful polysomnographic assessment.

            Fay suggests that waking at night and bedtime resistance may themselves be a clinical sign of a sleep related breathing disorders, arguing that a large proportion of children with obstructive sleep apnoea also have such problems. This link is contentious and disputed,1 and, even if it is accepted it does not follow that a large proportion of children with night waking and bedtime resistance also have obstructive sleep apnoea. The prevalence figures make it more likely that children with obstructive sleep apnoea form a small subset of those with behavioural sleeplessness. Research suggests that a detailed sleep history would alert clinicians to which patients are likely to fall into this group.1 A phased approach to both assessment and treatment is needed to ensure maximum diagnostic accuracy and to minimise unnecessary procedures.2

            Buckner is right that assumptions about children's sleep are culturally determined. However, the reality that parents and healthcare professionals face within a Western society is that it is often considered desirable for children to sleep independently. In this context continuing sleep disturbance in a young child has many adverse effects on both the child and his or her parents or carers. Given this, seeking the best ways of helping such parents or carers find solutions to their difficulty seems a reasonable pursuit.

            We agree with Symon et al that a register of planned and ongoing trials, and the updating of systematic reviews through the Cochrane Collaboration are important. We have had discussions with the Cochrane Collaboration to that end. Nevertheless, their study would not have been included in our review as we focused on children with an established sleep problem. Another systematic review has dealt with studies similar to theirs.3

            We agree with the first three points made by Wolke. His fourth, about the choice of quality criteria, identifies a dilemma facing anyone conducting a systematic review: should they describe each study and its strengths and weaknesses in detail or use a quality rating scale as a form of shorthand? Within a journal it is rarely possible to describe each study in detail, and it can also make for difficult reading. The use of quality criteria makes comparison of trials easier to understand, but at the cost of inevitable loss of accuracy. We would welcome the development of quality criteria more appropriate for psychological treatment trials. Until that time use of well standardised quality ratings 4 represents the best compromise.

            References

            1. 1.
            2. 2.
            3. 3.
            4. 4.
            View Abstract

            Sign in

            Log in through your institution

            Free trial

            Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
            Sign up for a free trial

            Subscribe