The Baby Bites Text Messaging Project with randomized controlled trial: texting to improve infant feeding practices

{"Background": "Rapid weight gain in infancy is associated with childhood obesity, yet most pediatric obesity prevention interventions target older children. Pediatric primary care offers frequent contact with families during the first year, but barriers to effective communication exist. Mobile health (mHealth), including text messaging, is a low-cost approach that may influence early feeding decisions, though few rigorous studies have examined its impact on infant feeding practices and weight outcomes. This study aimed to explore current parental attitudes toward feeding and early obesity, then use those findings to develop and pilot-test a theory-informed text messaging intervention.", "Methods": "This mixed-methods study used a sequential exploratory design. In the qualitative phase (spring 2018), 15 parents (10 mothers, 5 fathers; mean age 26.1\u00b17.3 years; 40% Black, 33% Hispanic/Latina, 20% White) of infants <1 month old were recruited from an urban, low-income pediatric clinic in Fort Worth, Texas. Semi-structured interviews explored attitudes toward breastfeeding, formula feeding, introduction of solids, causes of early obesity, and preferences for receiving feeding information. Interviews were transcribed and analyzed using thematic analysis. Findings informed selection of the Health Belief Model (HBM) and development of gain-framed text messages targeting perceived benefits, barriers, self-efficacy, and cues to action. For the pilot RCT (fall 2018\u2013spring 2019), 38 parents of newborns (17 control, 21 intervention) were enrolled. Inclusion criteria: parent of healthy term infant aged 3\u201330 days, \u226518 years, English-speaking, with mobile phone and internet access. Randomization used a computer-generated schedule with balanced distribution for infant sex. The intervention group received feeding-related text messages 4 times/week for 4 weeks, decreasing to twice/week by week 5, then once/week from week 8 through 12 months. The control group received safety text messages on the same schedule plus usual care handouts. Weight and length were extracted from electronic health records at baseline (0\u20132 weeks), 2\u20134 months, 6\u20139 months, and 12 months. Weight-for-length percentiles and z-scores were calculated using WHO Anthropometric Calculator. Feeding practices were assessed via online surveys (adapted from NHANES) and 24-hour dietary recalls. Due to COVID-19, no 12-month anthropometric data were available. Statistical analysis used linear mixed models and repeated measures ANOVA.", "Key Results": "Qualitative findings revealed two overarching themes: parental attitudes about feeding/parenting and parental preferences about feeding information. Parents viewed breastfeeding positively but identified barriers including discomfort breastfeeding in public, pain, latching issues, and time constraints. Most identified 6 months as the ideal age to introduce solids and preferred fruits and vegetables as first foods. Parents attributed early obesity to early introduction of solids, unhealthy foods provided by family, lack of healthy food access, and overfeeding. They expressed desire for practical feeding information from healthcare providers and positive attitudes toward receiving it via text. For the RCT, 38 participants were analyzed (intervention n=21, control n=17). The intervention group had lower mean birth weight (3.11\u00b10.30 kg vs 3.43\u00b10.57 kg, P=0.03). There was a significant main effect of time on weight-for-length percentile (F(2,43.03)=3.26, P=0.04) and a significant main effect of group (F(1,34.47)=6.90, P=0.01), with the control group having higher weight-for-length percentiles across all time points. The time-by-group interaction was not significant (P=0.18). Breastfeeding persistence did not differ between groups at 6\u20139 months (\u03c7\u00b2[1]=1.45, P=0.23) or 12 months (\u03c7\u00b2[1]=0.24, P=0.63). Mean time breastfeeding was 4.99\u00b14.1 months in the intervention group vs 4.53\u00b14.4 months in controls. Age of introduction of solids was similar (6.5 months intervention vs 6.3 months control). Statistical analysis of sugar-sweetened beverage offering was not possible due to low response rates. Survey response rates were low throughout, and 6 participants were lost to follow-up due to phone number changes or loss of cell service.", "Clinical Implications": "This study found no significant improvements in weight outcomes or feeding practices from a text messaging intervention, despite careful theory-based development informed by parent input. The lack of effect may be partly explained by higher-than-average baseline breastfeeding rates (over 90%), making improvement difficult, and lower breastfeeding persistence at 6 and 12 months compared to national averages\u2014particularly in this low-income, predominantly non-White population where return-to-work pressures are common. The study was underpowered due to low survey response rates and loss of 12-month data from COVID-19. The findings suggest that text messaging alone, without social support elements, may be insufficient to change feeding practices. Future mHealth interventions should incorporate social support, flexible communication methods to address phone service interruptions, and in-person data collection for surveys. Interventions beginning during pregnancy and extending through age 2 may be more effective, as risk for obesity is present before infancy. Despite null findings, text messaging remains a potentially cost-effective method to extend provider communication, particularly for high-risk populations, but requires further research with better adherence strategies."}