In total, 582 women participated in the study, 281 in the intervention group and 301 in the control group.
The median duration of exclusive breastfeeding at 6 months differed significantly between the intervention group and the control group (P < .001, log-rank test).
Jiang H, Li M, Wen LM, Hu Q, Yang D, He G, Baur LA, Dibley MJ, Qian X. Effect of Short Message Service on Infant Feeding PracticeFindings From a Community-Based Study in Shanghai, China. JAMA Pediatr. 2014;168(5):471-478. doi:10.1001/jamapediatrics.2014.58
Appropriate infant feeding practices have the potential for long-term health effects. However, research findings on improving early infant feeding practices are limited. The wide use of mobile phone short message service (SMS) provides new opportunities for health promotion and services.
To assess the effect of an SMS intervention on infant feeding practices.
Design and Setting
Quasiexperimental design with follow-up measures scheduled at 4, 6, and 12 months at 4 community health centers in Shanghai, China. Two community health centers represented the intervention group, and 2 other community health centers represented the control group.
In total, 582 expectant mothers were recruited during the first trimester. Expectant mothers were eligible to participate if they owned a mobile phone, were first-time mothers, conceived a singleton fetus, were older than 20 years and less than 13 weeks’ gestation, had completed at least a compulsory junior high school education, and had no illness that limited breastfeeding after childbirth.
Mothers in the intervention group received weekly SMS messages about infant feeding from the third trimester to 12 months’ post partum.
Main Outcomes and Measures
The primary outcome was the duration of exclusive breastfeeding (EBF). Survival analysis was used to compare the duration of EBF between the intervention group and the control group.
Compared with the control group, the intervention group had a significantly longer median duration of EBF at 6 months (11.41 [95% CI, 10.25-12.57] vs 8.87 [95% CI, 7.84-9.89] weeks). The hazard ratio for stopping EBF in the intervention group was 0.80 (95% CI, 0.66-0.97). The intervention resulted in a significantly higher rate of EBF at 6 months (adjusted odds ratio, 2.67 [95% CI, 1.45-4.91]) and a significantly lower rate of the introduction of solid foods before 4 months (adjusted odds ratio, 0.27 [95% CI, 0.08-0.94]).
Conclusions and Relevance
An SMS intervention may be effective in promoting EBF, delaying the introduction of solid foods, increasing awareness of the World Health Organization breastfeeding guidelines, and improving knowledge of appropriate infant feeding practices for new mothers.
The low breastfeeding rate and inappropriate feeding practices in China are of concern.1,2 In 2008, the exclusive breastfeeding (EBF) rate in urban areas was only 16% for infants aged up to 5 months.3 Other national data suggest that one-third of infants are introduced to solid foods before age 4 months.4 Such factors may have contributed to the emerging childhood obesity epidemic in China.5,6
To date, no established model exists for improving early infant feeding practices. Early intervention studies7- 10 have included home visits, interactive group education for new parents, and web-based services. Recent Australian studies11- 13 have shown that early infant feeding interventions can be effective in prolonging breastfeeding duration and improving feeding practices, with one investigation12 also showing a reduction in body mass index at age 24 months.
Mobile phone technology is increasingly used in the health sector for delivering health care services, health promotion interventions, and disease prevention programs (also known as mobile health [mHealth]). Short message service (SMS) via mobile phones is the most widely adopted and inexpensive example of mHealth.14 It presents a new channel for information delivery, allowing access at a time and place that suits the individual.15 Short message service can be effective in promoting weight management and improving health services.16- 18 However, the effect of SMS on promoting breastfeeding or improving infant feeding practices has not been assessed to date. In China, the ownership of mobile phones is high, and most users are aged 18 to 40 years,19 including almost all expectant and new mothers. On average, 1200 messages are sent from each phone annually,20 which represents an opportunity to use SMS in promoting improved early infant feeding practices.
This article reports the 12-month results from a community-based SMS infant feeding promotion intervention study delivered to expectant first-time mothers in Shanghai, China. The mothers in the intervention group received weekly SMS plus routine maternal and child health care. The mothers in the control group received only routine care. The duration of EBF was the primary study outcome and was compared between the 2 groups.
An intervention study with a quasiexperimental design was performed between December 1, 2010, and October 30, 2012, to investigate the effectiveness of an SMS health promotion intervention on infant feeding practices. The study was approved by the institutional review board of the School of Public Health, Fudan University, Shanghai and the Human Research Ethics Committee of The University of Sydney, Sydney, Australia. Written informed consent was obtained from each participant at recruitment.
Four community health centers (CHCs) were selected from 2 districts in Shanghai. In each district, 2 CHCs with similar population size were chosen, and one was randomly assigned as the intervention site and the other as the control site. Because individuals living in the same community in Shanghai attend the same CHC, a possibility exists that they could share information. Hence, we assigned CHCs rather than individual expectant mothers to avoid potential data contamination. At approximately 12 weeks’ gestation, a pregnant woman is generally required to register and receive a pregnant woman health care card at the CHC where her household registration is held. From about 20 weeks’ gestation onward, the pregnant woman receives prenatal care at the maternity hospital where she will give birth. The mother and child are followed up by the CHC staff via 1 to 3 home visits within the first month after delivery. The CHCs are also responsible for providing child health care services from birth to age 6 years.
When expectant mothers attended the selected CHC for the first time at approximately 12 weeks’ gestation, they were approached by a maternal and child health care staff member with a letter of invitation and an information sheet. Expectant mothers were eligible to participate if they owned a mobile phone, were first-time mothers, conceived a singleton fetus, were older than 20 years and less than 13 weeks’ gestation, had completed at least compulsory junior high school education, and had no illness that limited breastfeeding after childbirth. Participating mothers were invited to complete a self-administered questionnaire containing sociodemographic questions (ie, age, education level, household registration, and family income) and questions related to their awareness of the World Health Organization (WHO) recommendations for breastfeeding.21 A total of 641 pregnant women consented to participate in the study and completed the baseline survey in the first trimester. They were followed up and checked for their eligibility again during the second trimester by the CHC staff. At the beginning of the intervention in the third trimester, 59 pregnant women were excluded because they did not meet the inclusion criteria (Figure 1). Therefore, 582 women were included in the study, 281 in the intervention group and 301 in the control group.
The intervention, one weekly text message, was performed from the beginning of the third trimester (28 weeks’ gestation) to 12 months’ post partum. The total duration of the intervention was 66 weeks. The text messages, approximately 180 to 210 characters, contained relevant breastfeeding or infant feeding advice. Messages were sent from a computer-based platform using Fetion (http://feixin.10086.cn) or FrontlineSMS (http://www.frontlinesms.com) software programs; both are available free of charge on the Internet. Messages sent and received by China Mobile users, representing 91.3% of the study participants, were free via Fetion. FrontlineSMS was used for sending messages to China Telecom and China Unicom subscribers, costing ¥0.1 (approximately US $0.02) per message sent, with no cost per message received.
The intervention messages were developed based on the WHO breastfeeding guidelines, consultation with pediatricians, community child health workers’ input, and the infant feeding literature. The messages are relevant to milestones in early child development and the specific needs of expectant mothers and new mothers. The focus of SMS messages at each stage is summarized in Table 1. In addition to weekly content-specific messages, other messages were sent inquiring whether the woman had given birth, what her breastfeeding status was, if she had returned to work, and what the timing of the introduction of solid foods was to determine the relevant types and content of messages to be sent. Participating mothers were also encouraged to communicate actively with the research team via SMS.
Mothers in the control group received the usual health care services during late pregnancy and post partum, as did the intervention group. Participating infants in both groups had routine physical checkups in the CHCs during their first year. The mean attendance rate at the routine child health checkups (at 1, 2, 4, 6, 9, and 12 months) exceeds 95%.
In this study, the primary outcome measure was the duration of EBF. Secondary outcomes included the following: (1) the rate of EBF at 6 months, (2) the duration of any breastfeeding, (3) the timing of the introduction of solid foods, (4) the breastfeeding rate at 12 months, (5) other infant feeding practices (such as drinking from a cup, taking a bottle to bed, and receiving food as a reward). Because evidence exists of widespread early introduction of complementary foods in China,22 we also measured the EBF rate at 4 months. The WHO definitions of EBF and any breastfeeding were used in this study.22
Data on EBF, breastfeeding, and the timing of the introduction of solid foods were collected and recorded by CHC child health physicians via a face-to-face interview as part of their routine work. They were unaware of the intervention allocation and at a training session were instructed to follow the standard data collection procedure. Researchers subsequently extracted data from the infants’ health checkup records at 4, 6, and 12 months. All physicians received training on how to collect information about EBF, breastfeeding, and the timing of the introduction of solid foods. For example, the mother or caregiver was asked, “Was the baby breastfed exclusively in the past 24 hours?” If the answer was no, she was further asked whether the infant had ever been exclusively breastfed and the duration of EBF. The mother was also asked, “Was the baby breastfed in the past 24 hours?” If the answer was no, she was further asked, “What was the duration of breastfeeding, including weaning?” In addition, the mother was asked, “At what age was the baby first given solids regularly?”
Data on appropriate infant feeding practices (drinking from a cup, taking a bottle to bed, and receiving food as a reward) and on the timing of the mother’s return to work were collected via a face-to-face interview at 12 months by a staff member at each CHC using a questionnaire adapted from the Healthy Beginnings Trial.11,23 For this study, we first translated the questionnaire into Chinese and adapted the questions to the Chinese context. Then, we pilot tested the questionnaire with 10 mothers in the CHCs. Based on the feedback from the pilot testing, further revisions were made before the questionnaire was used in the study.
Statistical analyses were performed using available software (SPSS for Windows, version 17.0; SPSS Inc). Proportions were compared using Pearson product moment correlation χ2 test, and Mantel-Haenszel χ2 test was used for comparing trends in proportions.
Survival analysis was used to compare the duration of EBF or any breastfeeding. Kaplan-Meier curves were used to estimate the median EBF and breastfeeding times, and differences were tested using the log-rank test. The estimated hazard ratios for stopping EBF or breastfeeding were calculated using a Cox proportional hazards regression model, in which proportional hazards assumptions were checked by plotting the cumulative hazards functions for the covariates. For each of the covariates being plotted, the lines did not cross; hence, the hazard assumptions were met. Multiple logistic regression was also applied to determine differences in infant feeding practices. Both the Cox proportional hazards regression model and multiple logistic regression controlled for awareness of the WHO breastfeeding guidelines at baseline and demographic characteristics, including age, household registration, education level, and family income, as well as when the mother returned to work after childbirth and whether she lived in rented accommodations at baseline.
In total, 582 women participated in the study, 281 in the intervention group and 301 in the control group (Figure 1). Compared with those in the control group, mothers in the intervention group were younger, were more likely to live in rental accommodations at baseline, and had a lower education level, lower family income, lower rate of return to work at 6 months, and lower awareness at baseline of the WHO breastfeeding guidelines (Table 2). At 12 months, 519 mothers (258 in the intervention group and 261 in the control group) remained in the study, for an overall retention rate of 89.2%.
The median durations of EBF at 6 months were 11.41 (95% CI, 10.25-12.57) weeks for the intervention group and 8.87 (95% CI, 7.84-9.89) weeks for the control group (P < .001, log-rank test) (Figure 2). Compared with that in the control group, the hazard ratio for stopping EBF in the intervention group was 0.80 (95% CI, 0.66-0.97).
The EBF rates at 6 months were 15.1% in the intervention group and 6.3% in the control group (Table 3). Using multiple logistic regression, the intervention group had a statistically higher rate of EBF at 6 months than the control group, with an adjusted odds ratio of 2.67 (95% CI, 1.45-4.91).
However, the median durations of any breastfeeding at 12 months were almost identical: 7.72 (95% CI, 7.26-8.19) months for the intervention group and 7.73 (95% CI, 7.28-8.18) months for the control group (P = .94, log-rank test). Furthermore, Cox proportional hazards regression analysis detected no significant difference in stopping breastfeeding between the intervention group and the control groups (hazard ratio, 1.00 [95% CI, 0.82-1.19]).
The rates of the introduction of solid foods were 1.5% before 4 months and 67.5% before 6 months for the intervention group compared with 3.8% before 4 months and 61.3% before 6 months for the control group (Table 3). Mothers in the intervention group were less likely to give solid foods to their children before 4 months than mothers in the control group (odds ratio, 0.27 [95% CI, 0.08-0.94]), but no significant difference was found between the groups for the introduction of solid foods before 6 months. No significant differences were observed at 12 months in drinking from a cup, taking a bottle to bed, and receiving food as a reward (Table 3).
In this study, the SMS intervention led to significant improvement in the duration of EBF and the rate of EBF at 6 months, as well as a significant reduction in the introduction of solid foods before 4 months. Although the effect size was small, we believe that at a population level in the Chinese context and internationally, where the rates of breastfeeding (including EBF) are still low,24 SMS use would be of public health significance. However, caution should be exercised in interpreting the findings because of the limitations of this study, especially the weak quasiexperimental study design.
The EBF rate at 6 months in the control group in our study was 6.3%, which may only partially reflect the true current EBF situation in Shanghai. This finding is similar to the reported 5.3% in Chongqing, a large city in Southwest China.25 The Chinese state council female workers protection regulation stipulates 120 to 135 days of maternity leave.26 In our study, about half of the participants (47.9%) had returned to work by 4 months and almost three-quarters (74.3%) by 6 months. This might have contributed to the dramatic decrease in the EBF rate from 4 months’ to 6 months’ post partum. The SMS intervention improved the EBF rate, which is consistent with the findings from a web-based breastfeeding education intervention that demonstrated improved EBF rate and frequency.7 However, unlike an Australian home visit intervention,11 we found no effect of SMS use on prolonging the duration of any breastfeeding. There could be several explanations, including different intervention methods and various study contexts (eg, mothers’ return to work and peer pressure).
The intervention improved the EBF rate at 6 months but not at 4 months. In China, it has been traditionally accepted that the right time for stopping EBF and introducing solid foods is approximately 4 months.22 Exclusive breastfeeding to 6 months is a new concept for Chinese mothers, and many may be unaware of the recommendation. The SMS disseminated this information to mothers in the intervention group, which might explain the effect on EBF at 6 months. Although we actively promoted the introduction of solid foods at 6 months in the SMS messages, feedback from participating mothers (not reported herein) confirmed that some CHCs still advised them to introduce solid foods at 4 months, and our results showed that 61.6% of infants were given solid foods between 4 to 6 months. This may explain the high proportion of mothers who introduced solid foods at 4 months in this study. Given this context, the intervention seems to have had some effect on reducing the introduction of solid foods before 4 months only and not before 6 months.
Encouragement of drinking from a cup and discouragement of taking a bottle to bed during infancy may help mothers and children establish an infant-led feeding model and allow them to regulate food intake by themselves.27 This in turn might be protective against the development of childhood obesity.28 Such strategies have been included in infant feeding guidelines for some countries.29 In our study, despite significant improvement in knowledge of appropriate infant feeding practices (data not shown), we found no significant improvement in appropriate infant feeding practices (not drinking from a cup, avoiding a bottle at bedtime, and not giving food as a reward), as observed in other studies.11,30 One reason could be that these are new concepts in China, and it may take some time for them to be adopted. Furthermore, the SMS promotion of appropriate infant feeding practices (regarding cup, bottle, and reward) started late in the study, commencing when the child reached age 6 months, which may explain the limited effect on behavioral changes.
To our knowledge, this is the first community-based intervention study to assess the effectiveness of promoting early infant feeding via SMS. The intervention strategy aligns with social applications of modern telecommunication technology. The message bank was established based on the WHO infant feeding guidelines, needs assessments of pregnant women and new mothers, and consultations with child health workers. Therefore, the messages were evidence based, relevant, and practical. Different messages were sent to mothers according to their child’s feeding status, which allowed the intervention to be responsive. Furthermore, we adapted the questions for most key measures at baseline and 12 months from those used in the Healthy Beginnings Trial,11,23 which will enable comparisons of different intervention approaches in various settings in the future.
Our study has several limitations. First, the study findings are limited by its quasiexperimental design. Mothers in the intervention group and the control group were not from the same CHC, which means that the care provided to participants in addition to the intervention might not be completely identical. Therefore, a potential bias could result from the design. However, we tried to minimize any bias by choosing one intervention group and one control group from each district because the counseling materials and service delivery are likely similar across the district. In addition, a disparity in some socioeconomic variables between the 2 groups might be partly due to the nonrandomized design. To overcome this, we used multiple regression analysis. Second, we did not explore the cost-effectiveness of the intervention. Third, a large proportion of the mothers were well educated, which might limit the generalizability of the findings.31 Fourth, the effect of the SMS intervention on early introduction of solid foods before 4 months needs further confirmation because the rate among the study population was low.
A mobile phone SMS intervention seems to be effective in promoting EBF and delaying the introduction of solid foods. The short median duration of EBF at 6 months in this study suggests an urgent need for breastfeeding advocacy and promotion to increase public awareness and government support and to create supportive environments (eg, by prolonging maternal leave to align with the WHO breastfeeding guidelines). Further exploration of the effectiveness of SMS interventions through large-scale cluster randomized clinical trials is needed.
Accepted for Publication: January 8, 2014.
Corresponding Author: Xu Qian, PhD, School of Public Health, Fudan University, 138 Yixueyuan Rd, Mailbox 175, Shanghai 200032, China (email@example.com).
Published Online: March 17, 2014. doi:10.1001/jamapediatrics.2014.58.
Author Contributions: Dr Jiang had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Jiang, Li, Wen, He, Dibley, Qian.
Acquisition, analysis, or interpretation of data: Jiang, Li, Wen, Hu, Yang, Baur, Qian.
Drafting of the manuscript: Jiang, Li, Wen, Baur, Qian.
Critical revision of the manuscript for important intellectual content: All authors.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was funded by a Small Research Grant from the Nestle Foundation and by grant 12GWZX0301 from the Shanghai Municipal Health Bureau.
Role of the Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank the Longhua, Caohejing, Meilong, and Xinzhuang community health centers in Shanghai, China, for their support, and we thank all the participants for their collaboration.