Trial design and participants
Our study was conducted as a parallel randomised clinical trial with intervention, control, and placebo groups. Reporting of this study has been done in accordance with the Consolidation Standards of Reporting Trials (CONSORT) statement. It was funded by Iran University of Medical Sciences and conducted in July 2021. The study’s protocol was registered in the Clinical Trial Registration Center on 14 March 2021. Code: IRCT20180427039436N9. Our study population included breast feeders and premature infants whose gestational age was less than 32 weeks who were hospitalized in the neonatal intensive care unit of Akbarabadi Hospital (referral and educational hospital) in Tehran, Iran.
Maternal inclusion criteria
Participants were included if they had a desire to exclusively breastfeed and pump human milk using an electronic pump from the third day following birth. Included participants were also required to be literate and at least 18 years of age.
Infant inclusion criteria
Infants were required to have been born preterm (at less than 32 weeks of gestation), and admitted to the intensive care unit, where such infants are typically fed by orogastric or nasogastric tube.
Maternal exclusion criteria
Participants were excluded if they were engaged in smoking, alcohol, and/or drug use, suffering from infectious diseases transmitted through human milk feeding (e.g., AIDS), had active pulmonary tuberculosis, had a history of infertility, were taking anticoagulants which are contraindicated with the use of Anise [32], had a history of cancer estrogen-dependent diseases (e.g., uterine cancer, breast cancer, endometrial hyperplasia) and/or had mastectomy, nipple fractures, breast abscesses and/or infectious mastitis.
Infant's exclusion criteria
Infants were excluded if they had congenital anomalies (e.g., cleft lip and palate), and/or were born as part of a multiple birth.
Withdrawal criteria
Participants were able to withdraw at any time without giving a reason. Whilst the supplementation of iron and multivitamins are recommended from the 16th week of pregnancy until three months after giving birth, our withdrawal criteria included cases in which there was use of any additional galactagogues (human milk inducers/enhancers). Participants were similarly withdrawn if they were consuming less than 10% of Pimpinella Anisum herbal tea, along with cases of allergic reaction and intolerance.
Data collection/recruitment
Convenience sampling was used to recruit eligible participants over a period of 6 months. Those who met the eligibility criteria were given participant information and those who subsequently agreed to take part in the study were invited to complete informed consent forms and enrolled.
Data collection was carried out between July 2021 and November 2021. Eligible participants were assigned to three groups: (1) intervention, (2) placebo and (3) control using the block randomization method (https://www.sealedenvelope.com) and followed for a period of 7 days. To determine the sequence of participants’ allocation based on the block randomization method, it is necessary to ascertain the total sample size, the number of groups and the number of group repetitions in each block (considered equal). In the current study, the size of each block was twice the number of groups (six groups in each block). The inclusion of a control group meant that it was not possible to assure blindness. Therefore, a randomization code was generated for each participant using a sealed envelope website. An epidemiologist, who was not part of the study assimilated the randomization list. For allocation concealment, the assignment list remained with the epidemiologist.
Human milk volume was recorded as one of the main outcomes of the study by participants in a daily information recorder form. The infant's weight was also measured and recorded by the lead researcher. Statistical analysis was performed by a statistician who also did not know the content provided for the study groups and their allocation.
Description of intervention
Participants in the intervention group received Pimpinella Anisum herbal tea (2 g of dried Anise plant plus 1 g of black tea). Those in the placebo group received black tea containing 3 g of dried black tea only. In both groups, tea was ingested 3 times [32,33,34] a day for a week at breakfast, lunch, and dinnertime. The intervention period was limited to 7 days, because typically after this time the preterm infant is fed human milk directly (rather than being tube fed), and thus the recording of milk production/volume becomes impossible. Uniformity and same-color packaging for herbal compounds was considered in both intervention and placebo groups. Both teas were prepared by the School of Persian Medicine in Iran University of Medical Sciences under the supervision of a specialized consultant of medicinal plants (PhD of Pharmacognosy) (A.SH.). Participants placed a tea bag in the cup, added 150 ml of boiling water, and sipped the tea after 10 min of infusion. Those in the control group did not receive any herbal or chemical composition of galactagogue. At the end of the intervention period, those in the control group were offered Pimpinella Anisum herbal tea also.
Participants in the intervention and placebo groups had no information about the type of herbal tea they were ingesting. To increase accuracy in relation to the intervention and placebo groups, the participants, the researcher, and statistical analyst did not know anything about the type of herbal tea ingested by the either the intervention or placebo groups.
Participants in all groups used the Medela electronic machine to pump their milk at least 6 times a day (3–4 hourly upon waking at approximately 8am) for a period of 7 days. Excess milk was stored in the milk bank of Akbarabadi Hospital.
Assessment of trial variables
The variables of this study were measured as follows:
Demographics and baseline characteristics
Demographics and baseline data were related to both parents and the infant. Maternal data was collected in relation to age, body mass index, education level, employment status, whether the pregnancy had been planned (or not), number of previous births, mode of birth, chronic diseases, use of drugs during pregnancy and steroid use, as well as consumption of iron and other supplements. Paternal data collected included occupation type, level of education, and socio-economic status. Data collected in relation to the infant included birth weight, height, head circumference, gestational age at birth, neonatal age upon entry to the study, observed sex at birth, cause of hospitalization and type of serum injected into the infant (e.g., Dextrose water) along with the infant feeding method chosen. Maternal food consumption known to increase human milk production such as lettuce, basil leaves, dill, fenugreek, carrot juice, spinach, sesame, and fennel was also monitored. Data relating to the infant was extracted from medical records by the researcher. Maternal and paternal data was recorded and self-reported by the parents.
Daily information recorder form
The daily information form included space for data entry in relation to the frequency of pump use, time spent on pump, volume of milk produced at each pumping, time spent skin-to-skin with the baby, frequency of liquids and tea based on the number of glasses consumed during the day, and the amount of pumped milk at the end of each day. All participants pumping human milk in the study completed this form daily for 7 days.
Infant weight recorder form
The infant's weight was measured by a German Seca scale on day 0 (before intervention), day 3 and day 7 of the intervention period. Measurements were taken by the researcher during the morning shift prior to the first human milk feed and after the infant’s first bowel and urinary outputs. Values were recorded in the infant weight recorder form.
Sample size
To determine the required sample size at 95% confidence level and 80% test power, assuming that the effect of Pimpinella Anisum on human milk volume in the intervention group should be 25 unit (d = 25) for the effect to be considered statistically significant compared with the control group, the sample size in each group was estimated to be at least 38 people using the sample size formula. The total number of participants in this study was therefore 145 individuals.
Based on the study by Turkyılmaz et al., (2011) the standard deviation of milk volume was estimated to be 53.5 and 12.9 in the intervention and control groups, respectively. Also, with regards to the weight of neonate, the sample size of 15 people in each group was determined [34], allowing for a drop-out rate of 20%.
$$n=\frac{({{z}_{1- {\alpha }\!\left/ \! {2}\right.}+{z}_{1-\beta })}^{2}\times ({s}_{1}^{2}+{s}_{2}^{2})}{{d}^{2}}$$
$$n=\frac{({1.96+0.84)}^{2}\times ({53.5}^{2}+{12.9}^{2})}{{25}^{2}}\approx 38$$
Ethics
The protocol of the present study was approved by the Ethics Committee of Iran University of Medical Sciences with the ethics code: IR.IUMS.REC.1399.1417. All participants were fully informed about the objectives and process of the study and informed written consent was obtained from them. The information obtained during the study process remained completely confidential. All participants were informed that they could withdraw from the study at any time. No fees were paid for the study and all services were completely free. Only one code was assigned to the study, and all study data were entered into data systems anonymously. All participants were given the researcher's phone number at the beginning of the study, as participants were asked to inform the researcher of any side effects.
Statistical analysis
One-way analysis of variance (ANOVA) was used to statistically analyse the data and to compare the groups in terms of quantitative variables such as maternal age, Body Mass Index (BMI), infants' gestational age at birth and age upon entry to the study, infant’s weight, height and head circumference, and maternal consumption of fluids and tea (glass/day). Fisher's exact test was used to compare the groups in terms of variables such as economic status, mode of birth, paternal education level, and type of serum injected into the infant (e.g., Dextrose water). Chi-squared test was used to compare the group in terms of maternal and paternal occupation, maternal education and number of previous births, infant’s observed sex at birth, whether pregnancy was planned or unplanned, chronic diseases, any drug use during pregnancy, maternal use of steroids, supplements and drugs, cause of infant hospitalization and chosen infant feeding method.
One-way ANOVA was used to compare the frequency of pumping, time spent on pump, time spent skin-to-skin with the infant, frequency of liquids and tea based on the number of glasses consumed during the day, amount of pumped milk at the end of each day, and infant weight in the groups on the first to seventh days after the start of the intervention. Analysis of variance with repeated measures was used for the comparison of the mean frequency of pump use, time spent on pump, time spent skin-to-skin with the infant, frequency of liquids and tea based on the number of glasses consumed during the day, amount of pumped milk at the end of each day, and infant weight over time (first to seventh days after the start of the interventions). Data were analysed by SPSS software version 22 (IBM Corp, US). In all tests, a p-value of < 0.05 was considered significant.