Massage intervention for promoting mental and physical health in infants aged under six months
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The aims of this review are as follows:
a) To assess whether infant massage intervention is effective in promoting infant mental health including infant‐adult interaction in population samples;
b) To assess the role of infant massage intervention in the promotion of infant physical health in population samples.
Background
The context
Since Bowlby (Bowlby 1969; Bowlby 1988) first outlined the crucial nature of attachment for the foundation of mental health, a wealth of other research (Stern 1998; Sroufe 1996; Steele 1996) has demonstrated the importance of early infant relationships. Children who are responded to sensitively, with gentle touch, eye contact and infant directed speech, are more likely to form secure attachments and to have a solid foundation for positive mental health (Tronick 1982). Environmental stimulation, including touch, has been shown to be a factor in the myelination of neurons and proliferation of nerve cell connections (Carlson 1997). The advent of new techniques for imaging the functioning of the brain has indicated that early interactions influence the evolution of brain structures responsible for the individual's socio‐emotional functioning for the rest of the lifespan (Schore 1994).
It has been recognised that it is the quality of the parent‐infant relationship, in particular, that will create the conditions for establishing healthy patterns of functioning (Stein 1991; Murray 1996). Murray has demonstrated that the more the parent is sensitive towards identifying the infant's signals and cues at two months the better the outcomes for cognitive and emotional development (Murray 1992). The quality of this interaction relies to a large extent on the parent's ability to read and respond appropriately to the infant's emotional state (Kropp 1987; Zeanah 2000). In healthy relationships the main carer is aware of and responding to the infant's affective cues, and vice versa, and lack of this synchrony may be the foundation of an infant's behavioural and physiological difficulties (Gianino 1988). Dysregulation has been linked with sleep and feeding disorders (Benoit 2000; Benoit 2000). Gunnar et al reported that infants who gave clear stress signals and had responsive care givers were likely to have an effective dyadic stress regulatory system by six months whereas repeated unregulated stressful experiences may promote anxiety difficulties and dispositions (Gunnar 1993). Other research shows a clear relationship between poor maternal ‐ infant relationships and outcomes such as emotional and cognitive deficits (Cogill 1986) and a range of mental health problems (Fonagy 1997).
This suggests a need for practical interventions aimed at improving parent infant interactions by enabling carers to respond to the infant's emotional state and read signals and cues. Infant massage is one such intervention that is currently being used to promote infant‐carer relationships.
Infant massage
Infant massage is defined as "systematic tactile stimulation by human hands" (Vickers 2004).Traditionally infant massage has been practiced in many areas of the world especially in the African and Asian continents, indigenous South Pacific cultures and the Soviet Union (Field 2000). A survey of 332 primary caretakers of neonates in Bangladesh found that 96% engaged in massage of the infant's whole body between one and three times daily (Darnstadt 2002). Recently infant massage has gained popularity in western cultures, particularly as an intervention with high‐risk infants in neo‐natal intensive care units (NICU), where babies experience a stressful environment and deprivation of tactile stimulation. Increasingly, infant massage groups are also being set up in the community for lower risk babies and their primary care givers. The anecdotal claims for the effects of infant massage are wide and include physiological benefits for sleep, respiration, elimination and the reduction of colic and wind (Field 2000). Other claims indicate that infant massage promotes positive parent infant interaction, particularly by encouraging parents to recognise infant cues (Onozawa 2001; Pardew 1996) and a reduction in infant stress measured by reduced cortisol levels (Field 1996a).
The evidence
The majority of controlled trials of the effectiveness of infant massage have been conducted with high‐risk groups, in particular, low birth weight and pre‐term infants. A brief summary of research from the different groups follows.
Low birth weight and pre‐term infants
Ireland and Olson reviewed literature on massage and therapeutic touch in children, looking at seven studies of massage therapy in preterm infants who did not have other co‐morbid conditions and found beneficial effects on weight gain, activity level, and hospital stay, but attempted no quantitative analysis (Ireland 2000). A Cochrane systematic review (Vickers 2004) of infant massage for promoting growth and development of pre‐term/and or low birth infants concluded that the evidence that massage is of benefit for developmental outcomes is weak and does not warrant wider use of pre‐term infant massage. Where massage is currently provided by nurses, it is suggested that consideration be given as to whether this is a cost‐effective use of time. The evidence indicated that infant massage decreased the length of hospital stay by 4.6 days but the reviewers were concerned that the meta‐analysis of this outcome was strongly influenced by one study with a small sample size. The review recommends that future research should assess the effects of massage interventions on clinical outcome measures, such as medical complications or length of stay, and on process‐of‐care outcomes, such as care‐giver or parental satisfaction.
Infants of depressed mothers
Studies of the effects of infant massage carried out by depressed mothers suggested that post intervention, the babies fell asleep faster and for longer, and that mothers in the massage group interacted more positively in recorded face‐to‐face interactions (Field 1996b; Onozawa 2001). Studies of depressed mothers using infant massage have recorded a consistent and marked improvement in both infants' and mothers' mood and behaviour (Onozawa 2001) and Field's (Field 1996b) study indicated a reduction in infant stress measured by lower salivary cortisol levels.
Infants in the general population
In cultures where infant massage is not a routine part of care, increasing numbers of health staff and independent individuals are undertaking training to teach infant massage techniques within the community. Investigations into the benefits of infant massage within the general population of 'low risk' infants in the general population suggest some evidence of effectiveness. For example, Field found fewer stress behaviours in massaged infants together with lower cortisol levels (Field 1996a). The relationship of fathers with their infants has been the subject of much debate, and Scholz found that fathers who were taught to massage their first born children showed greater involvement with their infants, who greeted them with more eye contact, smiling and vocalisation (Scholz 1992).
Objectives
The aims of this review are as follows:
a) To assess whether infant massage intervention is effective in promoting infant mental health including infant‐adult interaction in population samples;
b) To assess the role of infant massage intervention in the promotion of infant physical health in population samples.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials. Studies will be included if the investigation included randomisation to either infant massage or a control group that received no intervention.
Types of participants
Babies under the age of six months, massaged by parents or other carers, will be eligible for inclusion. Studies of pre‐term and low birth weight babies receiving massage within a hospital setting will be excluded.
Types of interventions
Infant massage is defined in this review as systematic tactile stimulation by human hands. Studies will be included if they evaluate the effectiveness of infant massage, irrespective of the training programme or cultural practice underpinning the intervention. This will include studies where the technique of infant massage has been specifically taught to parents and/or staff, and evaluations of infant massage where it is used as a routine cultural practice. Multi‐modal interventions, of which massage is a part, will be included provided evidence of the benefits of massage as a separate intervention can be elicited.
Types of outcome measures
To be eligible for inclusion in the review, studies must include at least one standardized instrument that measures the effect of infant massage on either infant mental health (for example, the CARE index to measure infant‐adult interaction) or on physical health (for example, growth monitoring).
Information on any parent satisfaction (important to maintenance of the intervention); adverse effects and cost data will also be sought and reported, if identified.
Search methods for identification of studies
Relevant studies will be identified through the Cochrane Central Register of Controlled Trials (CENTRAL) and electronic searches of bibliographic databases, including:
Medline
PsycINFO
CINAHL
Embase
Dissertation Abstracts
AMED (Alternative and Complementary Medicine)
LILACS
The National Research Register
ClinicalTrials.gov
Cochrane Neonatal Review Group specialised register
The search terms will be adapted for use in different databases. No methodological terms have been included to ensure that all relevant papers are retrieved.
1.Infant massage or baby massage
2.Touch or tactile stimulation with infant or baby
3.#1 and #2
There will be no language restriction and relevant papers will be translated where necessary.
References lists of articles identified through database searches, and bibliographies of systematic and non‐systematic review articles will be examined to identify further relevant studies. Known experts will be contacted by letter or e‐mail to request details of any on‐going or unpublished trials and searches of conference proceedings will be undertaken.
Data collection and analysis
Selection of trials
Titles and abstracts of trials identified through searches of electronic databases will be independently reviewed by two reviewers to determine whether they meet the inclusion criteria (AU and JB). Abstracts that do not meet the inclusion criteria will be rejected. Two independent reviewers (AU and JB) will assess full copies of papers that appear to meet the inclusion criteria. Uncertainties concerning the appropriateness of studies for inclusion in the review will be resolved through consultation with a third reviewer (SSB) .
Quality Assessment
Critical appraisal of the included studies will be carried out by two reviewers (AU and JB). Disagreement will be resolved by consensus.
Trials will be assigned a quality category based on allocation concealment because there is a relationship between the potential for bias and the security of allocation concealment (Alderson 2004). Categories are described in the Cochrane Collaboration Handbook and are as follows:
(A) indicates adequate concealment of the allocation (for example, by telephone randomisation, or use of consecutively numbered, sealed, opaque envelopes);
(B) indicates uncertainty about whether the allocation was adequately concealed (for example, where the method of concealment is not known);
(C) indicates that the allocation was definitely not adequately concealed (for example, open random number lists or quasi‐randomisation such as alternate days, odd/even date of birth, or hospital number).
Quasi‐randomised trials will be allocated to category C (high risk of bias).
The following aspects of the study will also be appraised: the numbers in each group, number of children lost to follow‐up, the method of dealing with attrition/drop outs, and whether there was any assessment of distribution of confounders.
Where insufficient information about any aspect of study quality is available clarification will be sought from the trial authors. Trials in all categories will be included in the review, and a sensitivity analysis will be performed to determine impact.
Blinding of providers and patients will not be used as criteria to assess internal validity of included trials because it is rarely possible to blind to behavioural interventions. However, blinding of outcome assessors will be assessed.
Data Management
Data will be extracted independently by two reviewers and entered into RevMan 4.2.7. Where data are not available in the published trial reports, authors will be contacted to supply missing information.
Tests of homogeneity
An assessment will be made of the extent to which there are variations in the methods, population, intervention or outcome. Consistency of results will be assessed visually and by examining I2 (Higgins 2002), a quantity which describes the approximate proportion of variation in point estimates that is due to heterogeneity rather than sampling error. We will supplement this with a test of homogeneity to determine the strength of evidence that the heterogeneity is genuine. The possible reasons for heterogeneity will be explored by scrutinising the studies and where appropriate performing sub‐group analyses.
Analysis of missing data
Missing data and drop‐outs will be assessed for each included study and the reviews will report the number of participants who are included in the final analysis as a proportion of all participants in each study. Reasons for the missing data will be provided in the narrative summary, and the extent to which the results of the review could be altered by the missing data will be assessed. Best‐case and worst‐case scenario analyses will be attempted to assess the effect of the missing data. Assessment will be made of the extent to which studies have conformed to an intention‐to‐treat analysis and the effect of this on the result.
Synthesis of the data
In the absence of heterogeneity, a quantitative synthesis of the data will be undertaken using a fixed effect model. Where there is significant clinical heterogeneity, the results of individual studies will be presented graphically but not combined in a meta‐ analysis. Where there is sufficient clinical homogeneity to justify combining the results in a meta‐analysis, statistical tests of homogeneity will be applied. In the event of significant unexplained heterogeneity a random effects model will be used for the purpose of data synthesis, and caution will be exercised in the interpretation of the results.
Dichotomous data
For dichotomous data, relative risk will be estimated based on the fixed effect model. The number needed to treat (NNT) and the number needed to harm (NNH) for relevant clinical baseline risks will be calculated where appropriate.
Continuous data
Continuous outcomes will be analysed if the mean and standard deviation of endpoint measures are presented. Where baseline or pre‐treatment means are available, these will be examined to determine similarities between groups. For the meta‐analysis of continuous outcomes, weighted mean differences (WMDs) between groups will be estimated. In the case of continuous outcome measures, if data are reported on different and incompatible scales which cannot be adjusted to a uniform scale, data will be analyzed using the standardised mean difference (SMD) rather than the WMD.
Sensitivity Analysis
Sensitivity analysis will be used assess the robustness of the overall findings by examining the impact of study quality, e.g. lack of allocation concealment or high rates of loss to follow‐up, the impact of missing data or the possibility of one or more large studies dominating the results. If appropriate, subgroup analysis will be undertaken for clinically different interventions. Sub group analyses will be pre‐specified as far as possible to avoid data‐dredging.
Assessment of bias
Where appropriate, the possibility that the study selection was affected by bias will also be assessed in the data analysis, using funnel plots to investigate any relationship between effect size and study precision (closely related to sample size) (Egger 1997). Such a relationship could be due to publication or related biases or due to systematic differences between small and large studies. If a relationship is identified, clinical diversity of the studies will be further examined as a possible explanation.
