Introduction
Neonatal hyperbilirubinemia is one of the most common conditions requiring hospitalisation in the newborn period. Approximately 60% of term and 80% of preterm neonates develop clinical jaundice in the first week of life [1]. Although physiological jaundice is generally self-limiting, elevated unconjugated bilirubin carries the risk of neurotoxicity, and severe hyperbilirubinemia is responsible for approximately 70% of neonatal morbidity and 10% of neonatal mortality globally [2]. Bilirubin-induced neurological dysfunction may progress to kernicterus, causing permanent choreoathetoid cerebral palsy and sensorineural hearing loss [2].
Phototherapy remains the primary treatment for non-haemolytic neonatal jaundice; however, it is associated with adverse effects including dehydration, hyperthermia, erythematous rash, retinal damage, and disruption of mother-infant bonding [2]. Exchange transfusion, reserved for severe or refractory cases, carries procedural morbidity and is not universally available [2]. These limitations have stimulated interest in adjunctive, non-invasive strategies capable of reducing bilirubin levels or diminishing phototherapy requirements.
Neonatal massage has been practised in India and other Asian cultures for centuries. Its established benefits include improved weight gain, thermoregulation, and neurobehavioural outcomes [3]. More recently, evidence has emerged suggesting that massage accelerates meconium passage and reduces enterohepatic bilirubin circulation, thereby lowering serum bilirubin levels [4-6]. Coconut oil, widely used in southern India, is non-occlusive, non-irritant, rich in lauric acid, and possesses antibacterial properties, making it a suitable medium for neonatal massage [7].
Despite these promising attributes, studies evaluating oil massage specifically on neonatal bilirubin levels remain limited and heterogeneous in methodology. The present prospective interventional study was designed to determine whether coconut oil massage, initiated from day 2 of life, reduces bilirubin levels, decreases the need for and duration of phototherapy, and promotes stool passage in healthy term neonates.
Materials and Methods
A prospective, randomised, interventional study was conducted in the postnatal wards of a tertiary care hospital in Ahmedabad, India (B.J. Medical College and Civil Hospital) from 1 January 2019 to 31 July 2020. Ethical approval was obtained from the Institutional Ethics Committee and written informed consent was obtained from parents or guardians of all enrolled neonates.
Inclusion Criteria
Eligible neonates were healthy term infants (gestational age ≥37 weeks), appropriate for gestational age (birth weight 2500–3800 g), delivered by lower-segment caesarean section, kept with their mothers in the postnatal ward, and exclusively breastfed from day 1.
Exclusion Criteria
Neonates were excluded if they were preterm, post-term, or small for gestational age; had received any prelacteal feed or formula; required NICU admission; had pathological hyperbilirubinemia, ABO/Rh or other blood group incompatibility, or a major congenital anomaly.
One hundred neonates meeting inclusion criteria were randomised by simple randomisation into Group A (Massage Group, n=50) and Group B (Control Group, n=50). From day 2 of life until day 7, neonates in Group A received coconut oil massage three times daily at 0800, 1500, and 2100 hours, in addition to routine postnatal care. Control neonates received routine care only. Massage was performed by trained healthcare personnel using 2 mL of commercially available edible coconut oil per session in a room maintained at 26–28°C.
The massage protocol followed Field et al. (1986) and International Association of Infant Massage (IAIM) guidelines, consisting of a 10-minute session of three phases: (i) supine tactile stimulation of face, chest, abdomen, and limbs using 10 gentle strokes per area; (ii) prone stimulation of neck, back, and buttocks; and (iii) 5 minutes of kinaesthetic stimulation comprising passive flexion and extension of limbs. A skin sensitivity test was performed on the palm or sole before the first session. Massage was temporarily interrupted if the neonate cried excessively, urinated, or passed stool. For neonates commencing phototherapy, massage was performed during scheduled phototherapy breaks.
Outcome measures
Transcutaneous bilirubin (TCB) was measured daily from day 2 to day 7 using a calibrated BILIPROBE device (AVI Healthcare) at the forehead and manubrium sterni; TCB was not recorded after phototherapy commencement. Total serum bilirubin (TSB), direct and indirect fractions, were measured on days 3, 5, and 7 from peripheral venous samples analysed on an ERBA XL-640 automated chemistry analyser. Serum bilirubin values were plotted on the Bhutani hour-specific nomogram to determine the threshold for phototherapy initiation. Additional outcomes included: proportion of neonates requiring phototherapy; total phototherapy duration (hours); day of phototherapy initiation (as a proxy for peak hyperbilirubinemia); daily stool frequency; duration of meconium passage; and daily breastfeeding frequency.
Statistical analysis
Data were entered into Microsoft Excel and analysed using the online statistical package at www.medcalc.org. Continuous variables are expressed as mean ± standard deviation. Between-group comparisons were performed using the unpaired t-test for continuous variables and the chi-square test for proportions. A two-tailed p-value <0.05 was considered statistically significant.
Results
| Characteristic | Massage Group (n=50) | Control Group (n=50) | P value |
| Sex | 0.50 | ||
| Male, n (%) | 28 (56) | 31 (62) | |
| Female, n (%) | 22 (44) | 19 (38) | |
| Male: Female ratio | 1.27 | 1.63 | |
| Birth weight | 0.39 | ||
| 2500–2999 g, n (%) | 28 (56) | 31 (62) | |
| 3000–3499 g, n (%) | 18 (36) | 16 (32) | |
| 3500–3800 g, n (%) | 4 (8) | 3 (6) | |
| Mean birth weight (g), mean ± SD | 2890.2 ± 321.2 | 2945.9 ± 318.7 | |
| Maternal characteristics | |||
| Mean maternal age (years), mean ± SD | 25.04 ± 4.20 | 25.66 ± 4.77 | 0.49 |
| Mean birth order, mean ± SD | 1.74 ± 0.89 | 1.64 ± 0.80 | 0.56 |
| Educational status, n (%) | 0.50 | ||
| Illiterate | 7 (14) | 6 (12) | |
| Up to high school | 30 (60) | 34 (68) | |
| College degree | 13 (26) | 10 (20) | |
| Socioeconomic status, n (%) | 0.50 | ||
| Middle class | 38 (76) | 35 (70) | |
| Lower class | 12 (24) | 15 (30) | |
| Residence, n (%) | 0.50 | ||
| Rural | 8 (16) | 7 (14) | |
| Urban | 42 (84) | 43 (86) | |
| Occupation, n (%) | 0.50 | ||
| Housewife | 33 (66) | 40 (80) | |
| Farmer | 7 (14) | 4 (8) | |
| Employee | 10 (20) | 6 (12) | |
| Mean breastfeeding frequency (Number of times per day) | 8.25 ±1.70 | 7.99 ±1.59 | 0.44 |
The two groups were comparable in all baseline characteristics. Male neonates predominated in both groups (56% in Group A, 62% in Group B; p=0.5). Mean birth weight was 2890.2±321.2 g in Group A and 2945.9±318.7 g in Group B (p=0.39). Mean maternal age was 25.04±4.20 years in Group A and 25.66±4.77 years in Group B (p=0.49). Most mothers resided in urban areas, belonged to the middle socioeconomic stratum, were educated to high-school level, and were housewives. Mean daily breastfeeding frequency was 8.25±1.70 in Group A and 7.99±1.59 in Group B (p=0.44), confirming the absence of a confounding feeding effect.
| Day | Group A (mg/dL) | Group B (mg/dL) | p-value |
| Day 2 | 6.92 ± 2.84 (n=50) | 7.69 ± 2.42 (n=50) | 0.141 |
| Day 3 | 11.81 ± 3.83 (n=50) | 13.21 ± 2.28 (n=41) | 0.040* |
| Day 4 | 9.63 ± 4.80 (n=38) | 12.11 ± 4.78 (n=26) | 0.046* |
| Day 5 | 9.90 ± 3.23 (n=34) | 12.41 ± 6.33 (n=26) | 0.049* |
| Day 6 | 8.41 ± 5.90 (n=34) | 8.69 ± 6.03 (n=26) | 0.851 |
| Day 7 | 6.88 ± 6.85 (n=34) | 7.45 ± 6.41 (n=26) | 0.742 |
| Cumulative Mean | 8.93 ± 0.85 | 10.26 ± 1.19 | 0.049* |
* p<0.05 (statistically significant). NS = not significant. n = number of neonates assessed (those not yet commenced on phototherapy).
Mean TCB values in both groups were similar on day 2 (p=0.141). From day 3 to day 5, mean TCB was significantly lower in the massage group (p<0.05 for each day). Differences on days 6 and 7 were not statistically significant. The cumulative mean TCB over days 2–7 was 8.93±0.85 mg/dL in Group A versus 10.26±1.19 mg/dL in Group B (p=0.049). Both groups followed a similar pattern of rising TCB to a peak on day 3 and gradual decline thereafter, with the massage group bilirubin nomogram lying consistently below that of the control group during days 3–5.
| Day | Group A (mg/dL) | Group B (mg/dL) | p-value |
| Day 3 | 11.10 ± 4.44 | 12.81 ± 3.68 | 0.037* |
| Day 5 | 13.02 ± 2.75 | 14.32 ± 3.65 | 0.047* |
| Day 7 | 10.72 ± 2.73 | 11.31 ± 2.52 | 0.260 |
| Cumulative Mean | 11.61 ± 1.23 | 12.81 ± 1.50 | <0.0001* |
* p<0.05 (statistically significant).
Mean TSB was significantly lower in the massage group on days 3 and 5 (p=0.037 and p=0.047 respectively). Day 7 values did not differ significantly (p=0.26). The cumulative mean TSB was 11.61±1.23 mg/dL in Group A versus 12.81±1.50 mg/dL in Group B (p<0.0001).
| Parameter | Group A (Massage) | Group B (Control) | p-value |
| Neonates requiring phototherapy | 16 (32%) | 24 (48%) | |
| Mean phototherapy duration (hours) | 33.57 ± 8.39 | 40.33 ± 7.32 | 0.01* |
| Mean hour of life at phototherapy initiation | 93.23±3.73 hour of life | 73.44±3.51 hour of life | <0.0001 |
* p<0.05 (statistically significant).
Phototherapy was required in 32% (n=16) of neonates in Group A compared with 48% (n=24) in Group B. Among neonates receiving phototherapy, mean duration was significantly shorter in the massage group (33.57±8.39 hours vs 40.33±7.32 hours; p=0.01). The mean hour of life at phototherapy initiation (as a proxy for peak hyperbilirubinemia) was later in the massage group (93.23±3.73 hours) compared with the control group (73.44±3.51 hours; p<0.0001), indicating that significant jaundice requiring intervention peaked later in the massage group.
| Day | Group A (mean/day) | Group B (mean/day) | p-value |
| Day 1 | 1.64 ± 0.88 | 1.54 ± 0.81 | 0.56 |
| Day 2 | 2.30 ± 1.27 | 2.42 ± 1.09 | 0.61 |
| Day 3 | 3.92 ± 1.08 | 3.60 ± 0.88 | 0.11 |
| Day 4 | 5.66 ± 1.70 | 4.22 ± 1.34 | <0.0001* |
| Day 5 | 6.34 ± 1.61 | 5.56 ± 0.97 | <0.0001* |
| Day 6 | 7.38 ± 1.70 | 6.48 ± 1.47 | 0.005* |
| Day 7 | 8.76 ± 3.01 | 7.72 ± 2.65 | 0.07 |
| Cumulative Mean | 5.14 ± 0.44 | 4.51 ± 0.61 | 0.045* |
* p<0.05 (statistically significant).
Daily stool frequency was comparable between groups on days 1, 2, and 3 (p>0.05). From day 4-6, stool frequency was significantly higher in the massage group on days 4, 5, and 6 (p<0.0001, p<0.0001, and p=0.005, respectively). The cumulative mean stool frequency was 5.14±0.44 per day in Group A versus 4.51±0.61 in Group B (p=0.045). This difference persisted when neonates receiving phototherapy were excluded from analysis (5.40±2.86 vs 4.73±2.38; p=0.03), confirming that the effect is independent of phototherapy. Meconium passage beyond 72 hours occurred in 28% of control neonates versus 12% of massage neonates (chi-square=4.0; p=0.045).
Discussion
This study demonstrates that coconut oil massage initiated on day 2 of life and continued three times daily for six days significantly reduces both TCB and TSB in healthy term neonates, decreases the proportion requiring phototherapy, shortens phototherapy duration, and accelerates meconium clearance. These findings are consistent with and extend the growing body of evidence supporting neonatal oil massage as an adjunct in the management of physiological hyperbilirubinemia.
The proposed biological mechanism centres on enhanced vagal tone. Massage stimulates the parasympathetic nervous system, increasing vagal activity as evidenced by increased high-frequency heart rate variability [4]. This promotes gastric motility, facilitates intestinal peristalsis, and accelerates the transit and excretion of bilirubin-laden stool. Delayed meconium passage is an established risk factor for neonatal jaundice because meconium contains conjugated bilirubin that is deconjugated by intestinal beta-glucuronidase and reabsorbed into the enterohepatic circulation [5]. By promoting earlier and more frequent stooling, massage reduces this reabsorption, lowering the total bilirubin burden. Additionally, massage enhances peripheral lymphatic drainage and blood flow, facilitating systemic bilirubin clearance [6].
Our TCB findings align closely with those of Babei et al.,[8] who observed significantly lower TCB in a massage group versus controls on days 3–5 (p<0.05), and Moghdam et al.,[9] who reported significantly reduced TCB on days 3 and 4 in preterm neonates receiving massage (p<0.003). The significantly lower serum bilirubin in our massage group on days 3 and 5 is consistent with El-Magd et al.,[10] Lin et al.,[11] and Ahmed et al.,[12] all of whom reported statistically significant reductions in TSB on day 3 in massage groups (p<0.05). The convergence of day-7 TSB values in our study likely reflects physiological resolution of jaundice in both groups and mirrors the pattern seen in Babei et al.
The reduction in phototherapy requirement from 48% to 32% and the shortening of mean phototherapy duration by nearly 7 hours in the massage group are clinically meaningful. Phototherapy carries its own burden-dehydration, temperature instability, separation from the mother, and potential gonadal and retinal damage-so any intervention that reduces its use or duration is of clinical value, particularly in resource-limited settings. The finding that phototherapy was initiated significantly later in the massage group (93.23 vs 73.44 hours; p<0.0001) further suggests that massage dampens the early bilirubin rise, delaying the crossing of phototherapy thresholds.
The comparable breastfeeding frequency between groups (p=0.44) confirms that the observed bilirubin differences are attributable to the massage intervention rather than differential feeding. Similarly, the persistence of the stool frequency advantage in non-phototherapy neonates demonstrates that phototherapy itself (which can cause loose stools) is not the sole driver of increased stooling in the massage group.
Coconut oil was chosen for this study because of its favourable skin profile: it is non-occlusive, well-tolerated by neonatal skin, antibacterial (by virtue of lauric acid content), readily available, and inexpensive. No adverse skin reactions were observed in any neonate in this study, consistent with previous safety data [7]. Current evidence advises against olive oil and sunflower oil in neonates under four weeks of age due to potential disruption of the skin barrier, further supporting the choice of coconut oil [13].
Limitations of this study include its single-centre design and restriction to caesarean-section deliveries, which limits generalisability to vaginally delivered neonates. The study did not include a group receiving massage without oil, precluding separation of tactile stimulation effects from oil-specific effects. TCB measurement was discontinued upon phototherapy initiation, reducing the sample size at later time points. Future multicentre randomised controlled trials comparing different massage oils, massage frequencies, and gestational age groups would further consolidate the evidence base.
Conclusion
Coconut oil massage, initiated from day 2 of life and administered three times daily, significantly reduces transcutaneous and serum bilirubin levels during the critical first five days of life, lowers the proportion of neonates requiring phototherapy, shortens phototherapy duration, and accelerates meconium clearance through enhanced gut peristalsis. Being simple, non-invasive, inexpensive, and devoid of adverse effects, coconut oil massage can be recommended as a routine adjunct in the care of healthy term neonates to mitigate the burden of physiological hyperbilirubinemia.
Declaration
Ethical Clearance
The study was approved by the Institutional Ethics Committee of B J Medical College and Civil Hospital, Ahmedabad, Gujarat, India. Written informed consent was obtained from the parents or legal guardians of all enrolled neonates prior to participation in the study.
Conflict of Interest Declaration
None declared
Funding / Financial Support
Nil
Contributors
Jigar Jain: Conceptualization, study design, methodology, data analysis and interpretation, manuscript drafting, and critical revision of the manuscript.
Twisha Vaishnav: Data collection, literature review, data curation, and manuscript preparation.
Rutva Virpara: Patient recruitment, data collection, data entry, and literature review.
Keerthi Jayachandran: Data collection, statistical analysis, interpretation of results, manuscript drafting, correspondence with the journal, and overall coordination of the study.
Aditya Kushwah: Methodology, data interpretation, critical revision of the manuscript, and intellectual input.
Jolly Vaishnav: Conceptualization, supervision of the study, study design, critical revision of the manuscript, and overall academic oversight.
Arif Vohra: Methodology, supervision, interpretation of findings, critical review of the manuscript, and approval of the final version.
Acknowledgements
The authors thank the nursing staff of the postnatal ward, the biochemistry laboratory team at B.J. Medical College and Civil Hospital, Ahmedabad, and all families who participated in this study.