Balance ability of 7 and 10 year old children in the population: Results from a large UK birth cohort study

Abstract

Objective

The literature contains many reports of balance function in children, but these are often on atypical samples taken from hospital-based clinics and may not be generalisable to the population as a whole. The purpose of the present study is to describe balance test results from a large UK-based birth cohort study.

Methods

Data from the Avon Longitudinal Study of Parents and Children (ALSPAC) were analysed. A total of 5402 children completed the heel-to-toe walking test at age 7 years. At age 10 years, 6915 children underwent clinical tests of balance including beam-walking, standing heel-to-toe on a beam and standing on one leg. A proportion of the children returned to the clinic for retesting within 3 months allowing test–retest agreement to be measured.

Results

Frequency distributions for each of the balance tests are given. Correlations between measures of dynamic balance at ages 7 and 10 years were weak. The static balance of 10 year old children was found to be poorer with eyes closed than with eyes open, and poorer in boys than in girls for all measures. Balance on one leg was poorer than heel-to-toe balance on a beam. A significant learning effect was found when first and second attempts of the tests were compared. Measures of static and dynamic balance appeared independent. Consistent with previous reports in the literature, test–retest reliability was found to be low.

Conclusions

This study provides information about the balance ability of children aged 7 and 10 years and provides clinicians with reference data for balance tests commonly used in the paediatric clinic.

Introduction

The term balance describes the body's ability to keep its centre of gravity over its base of support [1] and it is this ability that underpins many motor skills such as the ability to sit, stand and walk.

The assessment of balance in children is not straightforward. A variety of different methods have been used ranging from simple tests such as single-leg stance and standing on foam [2], [3], [4], to standardised tests such as the Paediatric Clinical Test of Sensory Interaction and Balance [5] or the balance subtest of the Bruininks–Oseretsky Test of Motor Proficiency [6], [7], [8], [9]. Force-platforms and Computerised Dynamic Posturography (CDP) have also been used to assess both static and dynamic balance [10], [11], [12], [13], [14], [15]. However, concerns have been expressed about the validity and reliability of all assessment methods in terms of test–retest reliability [2], [16], [17], [18], [19]. Further, the cost of some assessment methods, alongside these concerns, has limited their use [20].

Although these standardised measures are readily available, in reality it remains the case that many clinicians continue to evaluate a child's balance ability subjectively, using their clinical experience and acumen rather than by comparison of results with reference data. Tests such as standing on one leg, standing heel-to-toe and walking across a beam are commonly used in the paediatric balance clinic with the clinician offering a subjective opinion about the child's balance ability. Basing a clinical judgement on subjective opinion is very likely to give misleading results, especially if one bears in mind that even standardised tests have been shown to have a large amount of variability [19]. The provision of reference data is therefore needed for these commonly used clinical tests to minimise subjectivity in interpretation and so that appropriate management decisions, such as referral for vestibular rehabilitation [21] or occupational therapy can be made.

Previous studies which have described balance function in children have claimed to provide normative data in that they have selected children for their normal prenatal, birth and medical history (and therefore without features which might increase their risk of balance problems) [5], [14], [15]. However, such studies are not normative in the sense of describing ‘what is usual’ [22]. Balance problems are likely to be less prevalent in children selected for their normal medical history and therefore this data cannot be considered as representative of the general population. Although a commonly used standardised test, the Bruininks–Oseretsky Test of Motor Proficiency (BOT-2) provides normative data for composite scores [23], these data are not available for individual test items. Other studies reporting balance function in children have been based on children already attending hospital clinics [6], [7], [9], [24]. Such atypical samples do not enable inferences about the general population to be made.

There have been 2 previous reports of balance ability in the paediatric population, both based on the 1970 British Birth Survey [25], [26]. Of greatest relevance to the present study is that by Haslum and Miles [26] who looked at the balance ability of 10 year old children in the context of exploring the relationship between dyslexia and motor performance. A total of 7542 children completed tests of single leg stance and 7778 children were tested walking backwards. A total of 595 children (7.9%) failed the single leg stance test and 599 children (7.7%) failed the walking backwards test. Unfortunately, Haslum and Miles [26] report neither frequency distributions for each of the tests, nor give measures of average balance ability (with a measure of the spread of the data). Hence, their data cannot be used for reference in the clinical setting. In addition, balance was only assessed with eyes open which both precludes an investigation of the relative contributions of the vestibular and visual systems and lowers the sensitivity of the test. Test–retest reliability was also not considered, and individual test scores rather than composite measures were used leaving questions around test validity and the possibility of measurement error.

The Avon Longitudinal Study of Parents and Children (ALSPAC) is an unselected sample of children drawn from the general population and therefore results should be applicable to the general paediatric population. As part of the assessment of the ALSPAC cohort, measures of the children's balance were performed at ages 7 and 10 years. The balance measures used by ALSPAC have commonality both with some of the commonly used standardised measures as well as being similar to some of the more subjective measures used by many clinicians. The ALSPAC data therefore has the potential of being able to provide reference data for use in the paediatric balance clinic. In addition, the frequency distributions of balance outcome measures in ALSPAC could be used for calibration of future studies to determine the prevalence of balance problems in other populations. In this study, the balance ability of children in ALSPAC at ages 7 and 10 years is described and correlations between parental report and clinical measures of balance are estimated.