The effects of infant abdominal massage on the parental stress level among Chinese parents in Hong Kong – A mixed clustered RCT

Abstract

Objective

To investigate the effectiveness and the experience of infant abdominal massage on reducing the parental stress level.

Methods

A clustered RCT was conducted, 160 parents were recruited from 10 mother groups. An infant abdominal massage class was given to those in the intervention group by a certified International Association of Infant Massage instructor. The Chinese Parenting Stress Index (PSI-SF) was used to measure parental stress levels at baseline and 4 weeks after the intervention. 8 parents from the intervention group were invited for an online interview to understand the experience of parents in participating the infant abdominal massage class. Thematic analysis was applied.

Results

The parental stress level was significantly decreased among intervention group with a mean difference of -5.46 (95% CI = 0.72 to 10.2, p = .049). ITT analysis was adopted for compliance analysis. The overall compliance for the intervention was 66% and found to have a significant effect on parental stress level (p < .01). The total PSI-SF among fathers was slightly higher than that among mothers. Furthermore, no moderating factors were found to have a significant interaction effect on PSI-SF (p < .05).For qualitative data, an overarching theme was generated that infant abdominal massage brought a positive experience despite the uncertainties affecting the compliance. Four themes and nine sub-themes were identified. Which were: (1) receiving clear and informative infant abdominal massage classes; (2) The influence of uncertain conditions on compliance; (3) Improving parental satisfaction; and (4) feeling it was a worthwhile experience.

Conclusion

This study provides a cost-effective method for parents to tackle their parental stress, the findings aid in the planning or adjustment of current postpartum depression screening while also helping to lower the risk of postpartum mood disorders.

Trial registration

Clinical Trial Number (NCT05650424 || http://clinicaltrials.gov/study/NCT05650424), registered on December 1, 2022. and HKU Clinical Trails Registry (HKUCTR3008), registered on November 3, 2022.

Introduction

Parenting stress encompasses various aspects, such as how fathers and mothers perceive their parental roles, parental perceptions of the infant’s level of difficulty, and the quality of parent–child interactions especially in the first year of parenthood [1]. Transitioning to parenthood can be challenging, as it involves changes in relationships with partners, work, and concerns about the baby and their social life. These factors can trigger parenting-related stress which may affect parents’ mental health which has been found to be linked to depressive symptoms. According to the World Health Organization, approximately 10% of pregnant women and 13% of women who have just given birth experiencing a mental disorder, primarily depression [2]. In developing countries, the problem is even more prevalent, affecting 15.6% or women during pregnancy and 19.8% after childbirth [3] In China, although the problem is similar (around 10–20%), only half of them receive treatment [4]. The problem was not only involved mothers, but a meta-analysis also showed that 13% fathers experienced depressive symptoms in the 3–6 months postpartum period, and 8.4% in one-year postpartum period [5].

Coping strategies and developing resilience can be important in maintaining parents’ physical and mental well-being [6]. Thus, understanding and promoting effective parenting styles can be crucial for the well-being of both parents and children.

Studies have investigated the effect of infant massage on various areas and have shown that it can help relieving colic and release build-up stress by providing the right type of stimulation to babies’ gastrointestinal system [7]. Besides, it also gave a positive effect on improving the quality and the duration of infants’ sleep, as well as maternal sleep quality [6, 8]. Furthermore, infant massage improved parent-infant bonding by enhancing the interaction between parents and infants [9, 10]. One study found that infant massage facilitated bonding between mothers with postnatal depression and their infants [11]. While another study indicated an improvement in the level of confidence and the attitude towards childbearing experience after performing massage for their own babies [12]. Better still, most studies did not observe any evidence of harmful effects resulting from infant massage on full term babies [13, 14].

Despite these benefits, most studies related to parental stress were focused only on specific groups of non-Asian mothers (e.g. mothers with preterm babies staying in the NICU and substance abuse recovering mothers) [15, 16]. The effect of infant massage on first year parental stress among new parents in general population has not been investigated. We therefore performed a clustered randomized controlled trial together with descriptive qualitative data, to assess the effectiveness and the experience of infant abdominal massage in reducing first year parental stress level, as well as the effect on infantile colic, sleep quality, among Chinese parents in Hong Kong.

Methods

Study and design

The study employed a two-arm, parallel-group, cluster randomized, wait-list RCT, conducted between October 2021 and April 2022. A two-arm cluster RCT was conducted, with mother groups serving as clusters and parents as individual participants. Mother groups were randomized into control and intervention groups. This cluster randomized approach, aimed to prevent couples from joining different groups and avoid double counting of demographic factors, such as household income.

To enhance participant motivation, a waitlist control design was adopted, where candidates assigned to the control group were placed on a waiting list to receive the intervention after follow-up data collection [17]. Couples randomized to the waitlist group received the intervention approximately four weeks after the couples in the intervention group had completed the intervention. Parental stress level (the primary outcome), along with moderating factors and compliance (secondary outcomes), were assessed at baseline and 4 weeks post-baseline. Given that the intervention was a baby abdominal massage training program, and the instrument was a self-report questionnaire, it was not feasible to blind mother groups or trainers to their study arm. Couples in the intervention group who underwent the 4-week intervention were also subjected to qualitative interviews, which were conducted via Zoom. The cluster RCT planning adhered to the Consolidated Standards of Reporting Trials (CONSORT) guidelines.

Participants

Participants aged 18 or above, with full-term babies aged 0–1 year without medical treatment of colic were recruited from mother groups on various social media platforms such as Facebook and Instagram. Recruitment was conducted in all districts of Hong Kong. Parents with babies who had abdominal surgery or currently participating in any other massage related studies were excluded. To prevent duplication of candidates participating in multiple mother groups, preference was given to mother groups created based on the babies’ month of birth.

Online sample size calculator [18] has been used to calculate the sample size. A two independent sample t-test was used to compare the total score of Parental stress index (PSI) between subjects who received infant abdominal massage classes and those in control group. In order to detect a moderate standardized effect size of 0.5 with 80% power and at most 5% chance of committing a false positive error, we need 63 subjects per group. For the sample size, if setting n = 63 and m = 20, ICC set as 0.01. According to the equation: effective sample size = n [1+ (m-1) (ICC)], the total number of sample size were 150 participants and 8 mother groups were recruited.

We contacted 8–10 mother groups in Hong Kong either by phone or in person. Consent was obtained from these mother groups. We obtained the mother’s written consent to participate in the study before her husband was invited as well. Approximately 70–75 couples (either one of or both parents of the babies) from each group were invited to participate in the trial. Information sheets were provided to all participants, explaining the purpose and procedures of the study. Participants were informed that their participation was voluntary and that they were free to withdraw from participating at any time if desired. They were assured of the anonymity and confidentiality of the collected data.

For qualitative data, participants were purposively recruited from the intervention group assigned in the clustered RCT. Parents were approached and asked for permission 4 weeks after baseline of the study. They voluntarily joined the study, with written informed consent sought. Saturation was established after eight sets of data. Two addition recruitment has been done to ensure that inductive thematic saturation was reached. No new code or themes was generated [19].

Procedure

The intervention group consisted of mother groups who received a 30-minute infant abdominal massage training class on the same day as their first parental stress level assessment (baseline). The training was conducted by a qualified instructor from the International Association of Baby Massage (IAIM) in a meeting room of clubhouses or private multi-activity rooms. A 7-step abdominal massage procedure was demonstrated using a baby manikin and simple notes (either hardcopies or softcopies) were provided to the intervention group after the class.

To ensure consistency and maintain the standard of the abdominal massage training sessions, the same qualified IAIM instructor was responsible for all massage handouts and training sessions. Based on the results of our pilot study conducted before, we emphasized the importance of maintaining good intervention compliance among parents to obtain reliable results. Therefore, our trainer followed up with all participants after the training session by updating their progress weekly via text message or phone call.

Parents were advised to perform a 5-minute abdominal massage session per day on their babies for a total of four weeks. Parental stress level was reassessed after 4 weeks of the intervention period. We also inquired about intervention compliance, including how frequently parents massaged their babies during the intervention period (4 weeks) and whether there was any improvement in their babies’ colic problem, sleeping quality and overall parental stress after 4 weeks of intervention. Self-reported compliance reported by parents in the intervention group was used to assess the correlation between compliance and outcomes.

Outcome measurements

The primary outcome for the study was parenting stress, which was assessed using the parental stress scale of the Chinese version of the PSI-SF (4th version). The PSI-SF has been widely used in research and clinical settings to assess parenting stress levels and identify parents who may be benefit from intervention programs aimed at reducing stress and improving parent-child interactions. It consists of 3 subscales and 36 items in total (with 12 items in each subscale), including difficult child (DC), parental distress (PD), and parent-child dysfunctional interaction (P-CDI). Which either individually or combined, affect both parenting behaviors and also child outcome [20]. The PSI-SF uses a 5-point scale (from 1 = strongly agree to 5 = strongly disagree) with a total score ranging from 36 to 180. A parenting stress total raw score of ≥ 90, i.e., at or above the 90th percentile, strongly indicates significant stress in the parent-child dyad.

The Chinese version of the PSI has been validated, with an overall reliability was 0.89 [21]. PSI-SF has been found to be an appropriate tool for measuring parenting stress for Chinese mothers in Hong Kong with a reliability coefficient of 0.93 [22].

Beside the PSI-SF subscales, we also compared the moderating factors that could affect the intervention effects on the parenting stress levels. We assessed self-reported compliance and overall perceived improvement in infant colic as well as infant quality of sleep through a questionnaire given to parents in the intervention group. We compared the overall improvement with self-compliance of the intervention, both above and below 50%. This allowed us to determine the effectiveness of the intervention, as well as the level of compliance needed to achieve improvements in parenting stress and child behavior.

Additional demographic and socio-economic characteristics of the participants were assessed, including age, sex, education level, marital status, employment status, and household income.

Statistical analysis

Baseline characteristics of participants in the intervention and control groups were compared using Chi-square tests for categorical variables and independent sample t test for continuous variables.

The efficacy analysis was performed based on the intention-to-treat principle, all participants (160 parents in total) were included in the analysis which ensured that all randomized participants in both intervention group and control group were included and analyzed as randomized [13]. Moreover, we followed the CONSORT 2010 statement for reporting cluster randomized controlled trials. To identify the effect of infant abdominal massage on parental stress level, we used the generalized estimating equations (GEE) model, which can accommodate subjects with at least one measurement [23] and accounts for the extra-covariance due to clustering by mother group and couple, and repeated measurements, with adjustment to baseline value. Each estimate was accompanied by a 95% CI, and a p-value of 0.05 was considered statistically significant. Data analysis was performed using IBM SPSS Statistics for Windows (version 26.0; IBM).

Qualitative analysis

A structured interview guide was used, and data were audio- and video- recorded, transcribed verbatim, and organized using field notes with the help of NVivo 12. The thematic approach [24] was used inductively for qualitative data. An inductive approach was used to describe the phenomenon through observation based as well as summarizing the characteristics of our subjects [16]. Member checking was adopted to discuss if the meaning of participants was distorted until a consensus was reached.

Results

Participants

The participant disposition is shown in Fig. 1. Between November 2021 and March 2022, we approached 10 mother groups via social media, inviting 160 potential parents with babies less than 1-year-old to participate in the study. All of them were eligible and provided consent for study participation. We randomized the 10 mother groups into either the intervention or the wait-list control groups, with each group comprising 5 mother groups and 80 parents.

Fig. 1

CONSORT Flow diagram of the study participants

In total, 7 participants were lost to follow-up in the intervention group, and 11 participants were lost to follow up in the control group due to lost contact. By applying the intention-to-treat principle, we included all parents in the final analysis, regardless of their level of compliance or loss to follow-up.

Table 1 and 2. present the baseline characteristics of the intervention and control groups. The age range of the participants was 25–44 years old, with 43.1% aged 25–34 years and 56.9% aged 35–44 years. Of the160 participants, 91 (56.9%) were female. Most of the participants (n = 153, 95.6%) were married, with only 2 (1.3%) and 5 (3.1%) being single parents and unmarried partners, respectively. Approximately half of the participants (n = 83, 51.9%) had a postgraduate level or above educational qualification, 64 (40%) had a diploma or degree, and 13 (8.1%) had a high school or below education. In terms of employment status, 137 (85.6%) participants were employed full time or working more than 35 h per week, while 17 (10.6%) were not in the workforce or student, and 6 (3.8%) were employed part-time or working less than 35 h per week. We had only one participant who was student, we combined those who were not in the workforce or student into one group.

Table 1 Baseline characteristics of participants

Table 2 Baseline characteristics of participants (by gender of parents)

Regarding household income, we only analyzed data from mothers to avoid double-counting. From Table 6.2, one family (1.1%) earned less than HK$19,999; 16 (17.6%) earned between HK$20,000–39,999; 28 (30.8%) earned between HK$ 40,000 to 59,999 per month, and 46 (50.5%) earned more than HK$ 60,000 per month.

Of the babies, 51 (56.0%) were boys. The employment status of participants showed a statistically significant difference between the intervention and control groups. However, no statistically significant differences were found in other demographic characteristics between the two groups.

Parental stress level

At baseline, the mean total PSI-SF score in all participants was 80.7 (SD = 19.25) on the scale of 36–180, while the mean subscale scores for PD, P-CDI and DC were 30.2 (SD = 8.03), 23.6 (SD = 6.76) and 26.9 (SD = 7.45) respectively on the scale of 12–60. The mean total PSI-SF of female and male were 78.9 (18.05) and 83.6 (20.85) respectively. The mean subscale scores for PD for females and males were 29.6 (7.88) and 31.2 (8.26), respectively. The mean subscale scores for P-CDI for females and males were 22.6 (6.41) and 25.1 (7.07), respectively. The mean subscale scale scores for DC for mothers and fathers were 26.7 (7.32) and 27.4 (7.71), respectively. The results suggested that the total PSI-SF and all subscale scores were slightly higher for fathers than for mothers.

Intervention effects on PSI-SF total score and subscales

Table 3. summaries the PSI-SF total and subscale scores for both the intervention and control groups at weeks 0 and 4. The PSI-SF total score was marginal significantly decreased from 83.1(SD = 19.31) at baseline to 78.7 (SD = 20.45) at week 4 (p = .049). At week 4, the intervention group had a significantly lower PSI-SF total score compared to the control group by -5.46 (95% CI = 0.72 to 10.2, p = .049). However, there were no significant differences between the intervention and control groups for the PD and P-CDI subscales.

Table 3 Intervention effects at weeks 4 by the generalized estimating equations (GEE)

Moderating factors of the intervention effects on PSI-SF

Table 4. presents an analysis of the moderators that influence the effects of the parenting intervention on PSI-SF. The results indicated that none of the moderating factors showed a significant intervention effect on the PSI-SF (p > .05). Note that since most participants in this study were married, the moderating effects of marital status on the parenting intervention on PSI-SF could not be assessed.

Table 4 The effect of moderators on PSI-SF

Self-reported compliance and overall improvement

The purpose of compliance analysis was to find out how compliance could affect the intervention effects on parental stress level and supplement the between group comparison. In the 4-week post-baseline assessment, we asked participants about their self-reported compliance in addition to the PSI-SF questionnaire by asking subjects the frequency of performing the intervention each week. We adopted ITT analysis and the compliance of participants in the control group as zero, the overall compliance for the intervention was 66% and a statistically significant effect (p < .01) on parental stress level which shown the importance of compliance towards the intervention effects.

We also asked the participants on the improvement on babies’ colic, sleep quality of babies as well as participants’ overall parental stress. Table 5. showed the change of total PSI-SF of at pretest and posttest with consideration intervention compliance. Result revealed that more than 80% of participants with compliance higher than 50% suggested an improvement on their overall parental stress level. Furthermore, 70% of participants with compliance higher than 50% found an improvement on babies’ colic, sleep quality of babies. However, less than 30% of the participants with compliance lower than 50% found an improvement on colic and sleep quality of their babies. Only 5% of them showed an improvement on overall parental stress.

Table 5 Compliance and physical improvement on babies and overall parental stress in intervention group

Experience of parents after the intervention

We collected qualitative data through in-depth semi-structured interviews with open-ended questions (Table 6.). The interview questions focused on the evaluation of the classes given to our intervention group during phase I, experiences and understanding of abdominal massage to babies, and the characteristics of implementation to gain a deeper understanding of the successes and failures of the intervention.

Table 6 The interview guide of semi-structured interview in phase II: qualitative data

An overarching theme was generated (as shown in Table 7.) that infant abdominal massage brought a positive experience despite the uncertainties affecting the compliance. Four themes and nine sub-themes were identified based on the experience of parents receiving infant abdominal massage classes and performing infant abdominal massage on their babies. The four themes were: (1) receiving clear and informative infant abdominal massage classes; (2) The influence of uncertain conditions on compliance; (3) Improving parental satisfaction; and (4) feeling it was a worthwhile experience. The nine sub-themes were further categorized under these four themes and are presented in Table 7.

Table 7 Overarching theme, themes and subthemes

Discussion

General discussion

This study mainly focused on baby abdominal massage, and we found a marginally significant improvement in parental stress level for Chinese new parents after the intervention, as measured by the PSI-SF assessment tool. Additionally, the study evaluated how compliance with the intervention affected the overall change in parental stress. Both quantitative and qualitative data supported the notion that parents with good compliance experienced a higher improvement in their first-year parental stress level.

Furthermore, the qualitative data collected in this study showed that parents had a positive experience and provided positive feedback on the infant abdominal massage training program. Positive changes were found on both parents and babies physically and psychologically after the intervention.

Strengths, limitations and further studies

A mixed methods approach was used in this study, including a cluster randomized controlled trial and a descriptive qualitative approach, yielded comprehensive results that the positive impact of infant abdominal massage training on first year parental stress levels. The research approach adopted was a multi-method research strategy, incorporating both quantitative and qualitative data collection and analysis methods [25]. By utilizing multiple methods, a deeper and broader understanding of the effectiveness of abdominal massage on parental stress level can be achieved.

This study has important implications for clinical practice. The findings suggest that among the three PSI-SF subscales, parental distress scored the highest, indicating that parents experienced higher stress during the transition to parenthood. Healthcare providers should consider providing social supports to parents to help them with this transition, especially first-time parents. Midwives or staff working in postnatal clinics could consider learning infant abdominal massage training and encourage mothers to perform massage to improve the availability and easiness of attending classes.

In addition, this study found that the overall PSI-SF scores, including the total and subscales, were slightly higher for fathers than mothers. This highlights the importance of addressing paternal stress. However, paternal health was less frequently addressed [26]. We often focused on mother-child dyad in determining child and family health and neglected the importance of paternal health towards family well-being. In current practice in Hong Kong, there was no routine screening of paternal depression screening programme. Therefore, paternal depression routine screening program is recommended.

By providing a more comprehensive paternal depression screening and different ways of support, not only fathers’ psychological well-being will be improved; partners’ and children would also be benefited with the improvement of marital relationship and parents-children bonding.

For limitations, as the intervention is a baby abdominal massage training program, and the instrument is a self-report questionnaire, it was not feasible to blind mother groups or trainers to their study arm. Although waitlist control design was adopted, lack of blinding might affect the study results. Stress-related biomarkers, for example cortisol, may be considered as the predictor, mediator or outcome in future studies for objectively measure of parental stress [27].

This study was conducted during the period of the outbreak of COVID, a challenging period when residents in Hong Kong experienced significant changes in multiple ways. These changes may have affected the baseline parental stress levels for parents in Hong Kong due to social burdens, such as the transition from working in the office to working from home [28] and social distancing. Parents may express a higher parental stress index when compared to the level before the pandemic as well as feeling of isolated, stress and worries [29]. Financial burden such as the general economic downturn, increased spending on protective measures such as face masks and disinfectant may have also contributed to increased parental stress levels [30]. Additionally, recruitment and intervention processes were also affected by social distancing measures and fear of virus transmission, which may have impacted the results of this study. Some parents may not have been willing to bring their babies to attend classes or participate in group sessions. From the fact that this study was conducted during the period of pandemic outbreak of COVID-19. Further research is needed to determine the difference and relationship on the effect of infant abdominal massage on parental stress level with and without the factor of the outbreak of pandemic towards parental stress.

Furthermore, seasonal factors may affect the overall compliance. This study was conducted during winter period, parents were concerned about their babies catching a cold when massaging them. Providing a comfortable environment, such as providing warm blanket and heater are recommended.

Moreover, while data saturation has been reached, it is important to acknowledge that the sample size of these qualitative studies was small. A larger sample size could be considered for future research to have a deeper understanding of the experience of infant abdominal massage and the first-time parental stress. Additionally, the quality of interviews may have been impacted by unstable network connections.

Conclusion

Infant abdominal massage is a low-cost intervention with high effectiveness in reducing parental stress levels among Chinese parents in Hong Kong. The study revealed that compliance was a significant factor affecting the intervention results, and further development of the intervention should consider ways to improve compliance. Moreover, collaborative efforts from research, clinical practice, and policy should be made to develop and maintain a high quality of intervention effect for helping all parents’ physical and mental well-being, as well as family relationships.