Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jul 23.
Published in final edited form as: Patient Educ Couns. 2014 Aug 8;97(2):256–260. doi: 10.1016/j.pec.2014.07.030

Weight-related SMS texts promoting appropriate pregnancy weight gain: A pilot study

Kathryn I Pollak 1,2, Stewart C Alexander 3,4, Gary Bennett 5, Pauline Lyna 1, Cynthia J Coffman 4,6, Alicia Bilheimer 1, David Farrell 7, Michael E Bodner 1,8, Geeta K Swamy 9, Truls Østbye 1,2
PMCID: PMC4512734  NIHMSID: NIHMS706957  PMID: 25153313

Abstract

Objectives

Excessive gestational weight gain (GWG) puts women and children at risk of obesity. We piloted an SMS-texting intervention to promote healthy GWG among overweight and obese women.

Methods

We recruited 35 women and randomized them in a 2:1 fashion to: a tailored SMS-texting intervention (Preg CHAT) vs. a generic texting intervention (Txt4baby). Preg CHAT texts provided personalized feedback based on women's intake of sweetened beverages, fruits and vegetables, fast food, daily steps taken, and weight. We abstracted women's weights from charts and surveyed women at baseline and 32 weeks gestation.

Results

Few women refused the study; many (30%) did not complete the study, however. Of those in the Preg CHAT arm, 86% responded to texts, and 80% said they would recommend this program to a friend. For women who completed the surveys (n=23), those in the Preg CHAT arm had a mean gain of 6 less pounds than women in the Txt4Baby arm (95% CI -15.9, 4.0; p=0.24).

Conclusions

This pilot study shows feasibility, acceptability, and potential efficacy of a low-intensity and disseminable intervention to help overweight and obese women reduce GWG.

Practice implications

An SMS texting program might help overweight women reduce excessive GWG.

1. Introduction

Excessive weight gain during pregnancy is a significant public health problem with as many as 57% of pregnant women gaining more weight than recommended by the Institute of Medicine [1, 2]. Excessive gestational weight gain (GWG) puts both mother and baby at risk for obesity and subsequent obesity related chronic disease [1, 3-6]. Interventions targeting GWG have only been somewhat successful, and more so for select groups[7]. For example, normal weight, higher income women, and those receiving exercise and/or nutritional counseling respond better to the interventions and are less likely to exceed GWG recommendations during pregnancy than overweight, lower income women, and those receiving no advice [8-11]. Interventions are needed to help these women, particularly those who are overweight or obese as they are at highest risk of complications.

Further, past interventions even those that are effective are not easily disseminable. Most interventions rely on face-to-face or telephone counseling or group sessions [7]. An innovative method for promoting healthy GWG that has not been tested among pregnant women is using Short Message Service (SMS) as a platform. Mobile phone use appears to be similar across all socioeconomic groups [12, 13]. In fact, some socially disadvantaged populations are more likely to text daily than their more advantaged counterparts [14]. Thus, using an SMS intervention could have strong and low-cost impact and help hard-to-reach women if found effective. Although SMS interventions have been tested and found effective for weight management interventions among non-pregnant populations[15-19], none has attempted to promote healthy GWG via SMS texting.

The purpose of this pilot study was to assess the feasibility, acceptability, and preliminary efficacy of a SMS based intervention to help pregnant overweight and obese women gain an appropriate amount of weight.

2. Methods

2.1. Design

In 2012, we recruited participants from two prenatal clinics. We randomized eligible women in a 2:1 fashion to either a tailored SMS intervention arm (Preg CHAT text) or a generic texting intervention arm (Text4baby). The study was approved by our university's Institutional Review Board.

2.2. Participant recruitment

Study staff reviewed electronic obstetric medical records weekly to identify potentially eligible women scheduled for a prenatal visit. Clinic staff approached such women and asked whether they would be willing to talk to a research staff member who would explain the study. If women were interested, study staff obtained verbal consent to determine eligibility using a brief screener survey. Eligibility criteria included: age of 18 years or older, English-speaking, registered for prenatal care at participating clinics, pre-pregnancy BMI of 25-40, 12-21 weeks pregnant, and having a cell phone with an unlimited texting plan for the next five months. Exclusion criteria were: pre-existing diabetes, limited mobility or inability to walk, impaired cognition or mental health with inability to provide consent.

If eligible, study staff obtained written consent and conducted a baseline survey. Then, staff randomized eligible patients, stratified by BMI category (overweight vs. obese)., Study staff administered two follow-up surveys at approximately 22 and 32 weeks gestation. Participants received $10 for each survey. Finally, staff abstracted all gestational weights from clinic charts at the end of women's pregnancy.

2.3. Measures

The surveys included questions about pregnancy history, weight history, stress, worry, self-efficacy, motivation and outcome expectations for gaining adequate weight, and food intake using the PRIME screen[20]. The PRIME screen contains 18 questions that target fruits, vegetables, whole and low-fat dairy products, whole grains, fish and red meat (as well as other foods that are major contributors to the intake of saturated and trans fat), their weekly consumption of fast food French, 2001 #6700} and current physical activity using the validated Pregnancy Physical Activity Questionnaire. (PPAQ, validated 24-hour recall of 32 activities)[21]. Both follow-up surveys also had additional process questions to assess usability and acceptability of the text messages. At each survey point, study staff weighed women using a calibrated scale (height assessed at baseline only). For the primary outcome, staff abstracted clinical weights between baseline and delivery of the baby.

2.4. Interventions

We sent SMS text messages to women randomized to the Preg CHAT text arm for approximately 16 weeks. Text messages were based on counseling protocols developed by the research team from previous studies targeting GWG. These messages were based on Social Cognitive Theory[22] and designed to increase self-efficacy, improve outcome expectations, address barriers, and promote self-monitoring. We targeted four weight-related behaviors: 1) increase their daily walking to 10,000 steps, 2) avoiding sweetened drinks, 3) eating at least five fruits and vegetables per day, and 4) eliminating fast food intake. We also asked women to text us their weight weekly. For the first 10 weeks, women were asked to start slowly with just two goals rather than being overwhelmed with four: walking 10,000 steps and avoiding sweetened drinks. At the time of the 22-week survey, we added two more goals (four goals total): eating at least five fruits and vegetables per day and eliminating fast food intake.

We texted women three days per week. On Mondays and Wednesdays women received two messages related to their targeted goals. The first text each day asked women to report how well they met their goals from the previous day (“Good morning. Please send me yesterday's: # steps, # sugary drinks. Separate with a comma (ex: 6123,1).”) We composed these messages ahead of time and included them in a message library. Research assistants then manually sent the messages out at the designated times. After women responded, we sent another text message with feedback on how well they were meeting each goal and a tip on how to achieve those goals. On Fridays, women received two texts (similar to the ones listed above) related to their goals and an additional text requesting their weight as well as a text asking the women to confirm they receiving their weekly texts. Based on the weekly texts, we tracked their weight gain. Monthly, women received a text message reminding them of the guidelines for gaining appropriate weight.

If women did not respond to any text, we sent out a separate text reminding them to respond (“We haven't heard from u regarding ur goals. Pls respond to my txts as it will help u keep on track. Thx”). Also, at the end of each month, we sent women a text message stating the percentage of texts to which they had responded (“Thx for responding to XXX% of the txts we sent this week! For the program to work, pls keep it up. Great job!”). To encourage adherence, women who responded to at least 80% of the texts were entered into a raffle for a $25 gift card.

For women randomized to the control arm, study staff helped them sign up for “Text4Baby,” a texting program sending three text messages a week, timed to a woman's due date, with general pregnancy information [23]. Text4baby is a free mobile information service developed by the National Healthy Mothers, Healthy Babies Coalition, designed to promote maternal and child health. A sample text message is “Brush & floss your teeth each day. When you're pregnant, your gums can swell and bleed. See a dentist to check. Call 800-311-2229 to find a dentist” (only a few of these texts were related to healthy eating or physical activity).

2.5. Analysis

This exploratory pilot study assessed preliminary feasibility, acceptability, and efficacy. Given the small sample size, the purpose of the analyses was to assess means and variations in outcome measures with a view to formal sample size calculations in a future study. For assessing feasibility, a priori, we considered the intervention feasible if 1) we could recruit 30 pregnant women into the pilot study in five months, 2) at least 80% of the 20 women in the Preg CHAT text arm would report reading some or all of the text messages, and 3) women in the Preg CHAT text would respond to 75% of the text messages upon receiving them. For assessing acceptability, we asked all women how useful the intervention was (1=“Not at all useful” to 5=“Extremely useful”), whether the intervention helped them gain the adequate amount of weight during pregnancy (1=“Did not help at all” to 5=“Helped a lot”), and whether they would recommend the program to a friend (1=“Definitely would not recommend” to 5=“Definitely would recommend”). To be deemed acceptable, at least 75% of women in the intervention arm would have to rate each item a “3”, “4” or a “5.”

For assessing efficacy, we conducted an intent-to-treat analysis with women analyzed as part of the arm they were assigned to regardless of treatment adherence. As a sensitivity analysis we conducted analyses including only women who completed the interventions. We fitted a repeated measures linear mixed-effects regression model with a spatial covariance structure to analyze the effect of the Preg CHAT text intervention, compared to Txt4Baby, on GWG as there was unequal spacing between weight measurements as well as measurements occurring at differing weeks of gestation across women [24]. The final model included the following fixed effects: self-reported pre-pregnancy weight, a linear term for week of gestation and a treatment arm by week gestation term. In preliminary models, we found that a quadratic term for week of gestation was not needed. From the final model, we estimated mean weights at 16 weeks and 40 weeks gestation and the difference in weight gain between the Preg CHAT text intervention and the Txt4Baby intervention at 40 weeks gestation. We also plotted the mean trajectories from this model for the observed trajectories for the intervention completers (n=23). For the nutrition score and physical activity outcomes of moderate and light physical activity, we fit linear mixed models using an unstructured covariance for the repeated measures over time[24] that included fixed effects for gestation time indicator variables of 22 and 32 weeks and treatment by gestation time indicator variables. Based on assessment of residuals, moderate and physical activity outcome variables were square root transformed for final linear mixed models. We also plotted the estimated means from these models. All analyses were performed by using SAS software (SAS Institute, Cary, North Carolina).

3. Results

See Table 1 for demographic characteristics of the participants. In terms of feasibility, we were able to recruit 35 women in 3 months from only two prenatal clinics. The initial refusal rate among women was low (11%). Of the 35 women, 23 were randomized to the Preg CHAT text arm and 12 to the Txt4Baby arm. Two women miscarried at 17 and 18 weeks of gestation and were excluded from all analyses (one in each arm). Furthermore, 10 women withdrew from the study; however, their abstracted weight data was used in the ITT analysis. Reasons women gave for withdrawing were: problem wearing pedometer (n=3), unreliable phone (n=2), program not for them (n=3), medical reason (n=1) and inconsistency with their schedule (n=1). In total, 23 participants completed the study interventions, 14 in the Preg CHAT text arm and 9 in the Txt4Baby arm. Women who completed the final survey (n=23) did not differ substantially from the total sample of women (n=33) in BMI (28.4 vs. 28.9) or self-reported baseline weight (175.1 vs. 177.2 lbs).

Table 1. Sample characteristics (n=33).

Demographic Preg CHAT Text (N=22) Txt4baby (N=11)
Age 29 (5) 32 (2)
White (n=19) 53% 47%
>HS (n=29) 66% 34%
Married (n=27) 59% 41%
Weeks gestation 16 (3) 17 (3)
Pre-pregnancy weight 180 (29) 172 (28)
Pre-pregnancy BMI 29 (4) 28 (5)
Open in a new tab

Most women (86%) in the Preg CHAT text arm reported reading their texts; 86% of women in the Preg CHAT text arm responded to their texts. Women in the Preg CHAT text arm rated the program as helpful in attaining the recommended weight (64% rated 5 or higher on a 7-point Likert item); no women in the Text4Baby arm (0%) rated the program as helpful for weight. Most (71%) of women in the Preg CHAT text intervention arm would recommend the program to a friend.

Women's mean weeks of gestation at baseline was approximately 16 weeks, and the mean number of weight measurements was 11 (with a minimum 3, maximum 20) following baseline for estimating weight trajectories from baseline (see Figure 1 that shows individual weight trajectories for intervention completers only as well as estimated mean trajectories from the linear mixed-effect models by intervention arm). In the intent-to-treat analysis (n=33), we found that participants in the Preg CHAT text arm had a mean gain of two less pounds than participants in the Txt4Baby arm (95% CI -9.8, 5.7; p=0.61) at 40 weeks gestation (Figure 1). Estimated mean weight at 40 weeks gestation was 210 lbs in the Txt4Baby arm and 208 lbs in the Preg CHAT txt arm. When we evaluated weight gain for women who completed the interventions (n=23), we found that women in the Preg CHAT text arm had an estimated mean gain of six less pounds than participants in the Txt4Baby arm (95% CI -15.9, 4.0; p=0.24) at 40 weeks gestation (Figure 2). Estimated mean weight at 40 weeks gestation was 214 lbs in the Txt4Baby arm and 208 lbs in the Preg CHAT txt arm.

Figure 1.

Figure 1

Open in a new tab

Black lines are estimated mean trajectories from repeated measures linear mixed models for intent-to-treat analysis (n=33, solid Preg CHAT text arm; dashed Txt4Baby arm). Light grey lines are observed weight trajectories (solid Preg CHAT text arm; dashed Txt4Baby arm). Solid grey vertical line is mean weeks of gestation at baseline (approximately 16 weeks).

Figure 2.

Figure 2

Open in a new tab

Black lines are estimated mean trajectories from repeated measures linear mixed models for completers only (n=23, solid Preg CHAT text arm; dashed Txt4Baby arm). Light grey lines are observed weight trajectories ((solid Preg CHAT text arm; dashed Txt4Baby arm). Solid grey vertical line is mean weeks of gestation at baseline (approxiamately 16 weeks).

Based on the PPAQ, baseline mean (SD) activity levels for light and moderate levels were 95.0 (49.7) met-hours/week and 80.8 (78.8) met-hours/week, respectively. The baseline (SD) mean Nutrition Score was 5.6 (6.9), indicating “fair” eating habits. At baseline no woman had an excellent rating for eating habits (score 32-38), 12% (n=4) had a good rating (score 15-31), 67% (n=22) had a fair rating (score 1-14) and 21% (n=7) had a poor rating (score 0 or less). We found no differential treatment effect for the physical activity level outcomes at 32 weeks gestation (moderate: 95%CI (-3.5, 0.3), p=0.71; light: 95%CI (-2.6,0.4), p=0.08) or the nutrition score (95% CI (-1.1, 6.9), p=0.15) (see Figure 2 for estimated mean trajectories from the linear mixed models). Arm differences in models on these outcomes for those that completed the surveys were similar.

4. Discussion and Conclusion

4.1. Discussion

This was the first SMS-based study to attempt to change gestational weight gain among overweight and obese women. Similar to trials in non-pregnant populations [15-17, 19], we found significant and positive effects for women who were prompted to report their weight-related goals several times a week. What differentiates our work from all others is that pregnant women in this trial were expected to gain weight, just not gain excessive weight, whereas all other trials were attempting to help people lose weight. Thus, we had to craft the messages artfully to acknowledge that some weight gain was expected, and that healthful weight gain was optimal. Plus, pregnant women are weighed frequently when they attend their many obstetric visits, which also is different from patients in general who might leave months or a year in between their doctor visits.

Even though we sent messages manually because it was a pilot study, in a larger trial, like in other SMS texting interventions, we would automate delivery of messages. A challenge with many SMS texting interventions is the inability to assess treatment penetration. Although we cannot tell if women read our text messages that were meant to boost self-efficacy and outcome expectations while also overcome barriers, we do know whether women texted back information on their goals. Thus, like in other SMS texting projects, the two-way aspect of the texting might have helped women feel accountable to respond with reports of how they did with their goals.

This pilot data indicates feasibility, acceptability, and preliminary efficacy for the SMS texting intervention targeting pregnant women. We were able to recruit women with a relatively low refusal rate. Women rated the intervention helpful and effective enough to recommend it to a friend. We had high attrition, however. Lessons from this pilot will inform the development of a refined intervention in a larger trial. First, women struggled to wear two pedometers, which made a few drop out of the program. Pregnant women's bodies and clothing are not conducive to wearing apparatuses on their hips. Some women dropped out because they did not have consistent phone coverage or their phone could not read our text messages. Also, some women did not own a scale with which they could weigh themselves at home. To reduce attrition in a larger efficacy trial, phones and scales should be supplied to women. This study has several limitations including its small sample and lack of long term follow-up into the postpartum period.

4.2. Conclusion

Overall, this was the first SMS texting pilot promoting healthy GWG among overweight and obese pregnant women. Similar to our work in other populations, this pilot study shows that SMS texting is a promising vehicle for behavior change among pregnant women. It is also widely disseminable and acceptable within the target population. The program needs to be tested in a large scale randomized controlled trial.

4.3. Practice Implications

An SMS-texting intervention might be an innovative way to help overweight and obese women reduce GWG.

Figure 3.

Figure 3

Open in a new tab

Estimated trajectories for moderate physical activity (square-root transformed), light physical activity (square-root transformed) and Nutrition Score from linear mixed models. Solid line is PregCHAT intervention, dashed line is Txt4Baby arm.

Acknowledgments

This project was funded by internal funds. Data was collected from February through October 2012 and analyzed in June 2013.

Footnotes

Conflicts of interest: The authors have no conflicts of interest to disclose.

References

  • 1.Vesco KK, Sharma AJ, Dietz PM, Rizzo JH, Callaghan WM, England L, et al. Newborn size among obese women with weight gain outside the 2009 Institute of Medicine recommendation. Obstet Gynecol. 2011;117:812–8. doi: 10.1097/AOG.0b013e3182113ae4. [DOI] [PubMed] [Google Scholar]
  • 2.Simas TA, Liao X, Garrison A, Sullivan GM, Howard AE, Hardy JR. Impact of updated Institute of Medicine guidelines on prepregnancy body mass index categorization, gestational weight gain recommendations, and needed counseling. J Womens Health (Larchmt) 2011;20:837–44. doi: 10.1089/jwh.2010.2429. [DOI] [PubMed] [Google Scholar]
  • 3.Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. Jama. 1999;282:1523–9. doi: 10.1001/jama.282.16.1523. [DOI] [PubMed] [Google Scholar]
  • 4.van den Brandt PA, Spiegelman D, Yaun SS, Adami HO, Beeson L, Folsom AR, et al. Pooled analysis of prospective cohort studies on height, weight, and breast cancer risk. Am J Epidemiol. 2000;152:514–27. doi: 10.1093/aje/152.6.514. [DOI] [PubMed] [Google Scholar]
  • 5.Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med. 2003;348:1625–38. doi: 10.1056/NEJMoa021423. [DOI] [PubMed] [Google Scholar]
  • 6.Josefson J. Excessive pregnancy weight gain raises the risk of having a fat baby. Endocrine Society's 93rd Annual Meeting; Boston. 2011. [Google Scholar]
  • 7.Muktabhant B, Lumbiganon P, Ngamjarus C, Dowswell T. Interventions for preventing excessive weight gain during pregnancy. Cochrane Database Syst Rev. 2012;4:Cd007145. doi: 10.1002/14651858.CD007145.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Phelan S, Phipps MG, Abrams B, Darroch F, Schaffner A, Wing RR. Randomized trial of a behavioral intervention to prevent excessive gestational weight gain: the Fit for Delivery Study. Am J Clin Nutr. 2011;93:772–9. doi: 10.3945/ajcn.110.005306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Polley BA, Wing RR, Sims CJ. Randomized controlled trial to prevent excessive weight gain in pregnant women. Int J Obes Relat Metab Disord. 2002;26:1494–502. doi: 10.1038/sj.ijo.0802130. [DOI] [PubMed] [Google Scholar]
  • 10.Asbee SM, Jenkins TR, Butler JR, White J, Elliot M, Rutledge A. Preventing excessive weight gain during pregnancy through dietary and lifestyle counseling: a randomized controlled trial. Obstet Gynecol. 2009;113:305–12. doi: 10.1097/AOG.0b013e318195baef. [DOI] [PubMed] [Google Scholar]
  • 11.Olson CM, Strawderman MS, Reed RG. Efficacy of an intervention to prevent excessive gestational weight gain. Am J Obstet Gynecol. 2004;191:530–6. doi: 10.1016/j.ajog.2004.01.027. [DOI] [PubMed] [Google Scholar]
  • 12.Koivusilta LK, Lintonen TP, Rimpela AH. Orientations in adolescent use of information and communication technology: a digital divide by sociodemographic background, educational career, and health. Scand J Public Health. 2007;35:95–103. doi: 10.1080/14034940600868721. [DOI] [PubMed] [Google Scholar]
  • 13.Smith A. Pew Internet and American Life Project. Washington, D.C.: Pew Research Center; 2011. Americans and their cell phones: mobile devices help people solve problems and stave off boredom, but create some new challenges and annoyances. [Google Scholar]
  • 14.Smith A. Washington, D.C.: Pew Research Center's Internet & American Life Project 2010; 2012. Mobile Access. [Google Scholar]
  • 15.Haapala I, Barengo NC, Biggs S, Surakka L, Manninen P. Weight loss by mobile phone: a 1-year effectiveness study. Public Health Nutr. 2009;12:2382–91. doi: 10.1017/S1368980009005230. [DOI] [PubMed] [Google Scholar]
  • 16.Gerber BS, Stolley MR, Thompson AL, Sharp LK, Fitzgibbon ML. Mobile phone text messaging to promote healthy behaviors and weight loss maintenance: a feasibility study. Health Informatics J. 2009;15:17–25. doi: 10.1177/1460458208099865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Morak J, Schindler K, Goerzer E, Kastner P, Toplak H, Ludvik B, et al. A pilot study of mobile phone-based therapy for obese patients. J Telemed Telecare. 2008;14:147–9. doi: 10.1258/jtt.2008.003015. [DOI] [PubMed] [Google Scholar]
  • 18.Patrick K, Raab F, Adams MA, Dillon L, Zabinski M, Rock CL, et al. A text message-based intervention for weight loss: randomized controlled trial. J Med Internet Res. 2009;11:e1. doi: 10.2196/jmir.1100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Shaw R, Bosworth H. Short message service (SMS) text messaging as an intervention medium for weight loss: A literature review. Health Informatics J. 2012;18:235–50. doi: 10.1177/1460458212442422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Rifas-Shiman SL, Willett WC, Lobb R, Kotch J, Dart C, Gillman MW. PrimeScreen, a brief dietary screening tool: reproducibility and comparability with both a longer food frequency questionnaire and biomarkers. Public Health Nutr. 2001;4:249–54. doi: 10.1079/phn200061. [DOI] [PubMed] [Google Scholar]
  • 21.Chasan-Taber L, Schmidt MD, Roberts DE, Hosmer D, Markenson G, Freedson PS. Development and validation of a Pregnancy Physical Activity Questionnaire. Med Sci Sports Exerc. 2004;36:1750–60. doi: 10.1249/01.mss.0000142303.49306.0d. [DOI] [PubMed] [Google Scholar]
  • 22.Bandura A. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall, Inc; 1986. [Google Scholar]
  • 23.Whittaker R, Matoff-Stepp S, Meehan J, Kendrick J, Jordan E, Stange P, et al. Text4baby: Development and Implementation of a National Text Messaging Health Information Service. Am J Public Health. 2012;102:2207–13. doi: 10.2105/AJPH.2012.300736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Verbeke G, Molenberghs G. Linear Mixed Models for Longitudinal Data. New York: Springer-Verlag; 2000. [Google Scholar]

RESOURCES