Gestational diabetes mellitus (GDM) is “glucose intolerance first recognized during pregnancy”1 and is associated with up to a 70% risk of type 2 diabetes in later life.2 A family history of type 2 diabetes, maternal adiposity, and excess gestational weight gain seem to confer the greatest risk.3,4 Despite pregnancy being considered a “teachable moment,” overweight and obese women often find it difficult to change dietary and physical activity behaviors during pregnancy.5 Recently published studies describing combined nutrition and exercise interventions have reported only some success in reducing GDM.2,6–9 The reason for this could be attributed to the lack of emphasis on psychologic theories in their methodologies that are imperative for changing lifestyle behaviors in an already potentially unmotivated group. Therefore, it is now recommended that lifestyle interventions should incorporate behavior change theories into study design and methodology.10
Mobile health or “mHealth” technologies are becoming commonplace to assist in the management of chronic illnesses and support behavior change.11 Considering that more than 70% of Irish pregnant women use smartphones,12 mobile health offers an opportunity to improve health behaviors in pregnancy. This article describes a multifaceted “healthy lifestyle package” for overweight and obese pregnant women, consisting of personalized low glycemic index nutritional and physical activity advice underpinned by behavior change theories with reinforcement through a specifically designed smartphone application. We hypothesized that this lifestyle intervention would reduce the incidence of GDM per the International Association of Diabetes and Pregnancy Study Group diagnostic criteria.13
MATERIALS AND METHODS
This was a single-center randomized controlled trial with ethical approval and maternal written consent conducted at the National Maternity Hospital, Dublin, Ireland. Recruitment ran from March 2013 to February 2016. The final delivery occurred in August 2016. The trial steering committee met bimonthly. An independent data monitor reviewed recruitment and safety data after 250 patients had been enrolled.
Details of the full study protocol, including eligibility criteria, recruitment and enrollment, and data collection, have been previously published.14 Briefly, singleton pregnant women between 10 and 15 weeks of gestation with body mass indexes (BMIs, calculated as weight (kg)/[height (m)]2) between 25.0 and 39.9 and in possession of a smartphone were recruited at their first antenatal visit. Participants returned for their first study visit within 2 weeks for randomization to either the intervention or control group. Randomization was performed using a computer-generated sequence in a ratio of one to one. The biostatistician prepared sequentially numbered, sealed opaque envelopes, which were opened at the first study visit. Randomized participants were stratified by BMI to ensure equal numbers of overweight and obese women in each group. As a result of the nature of the intervention, neither participants nor researchers were blinded to the intervention or outcomes.
Women allocated to the control group received standard antenatal care, which in Ireland does not consist of any uniform advice on diet, exercise, or weight gain in pregnancy. Participants allocated to the intervention group received standard antenatal care plus a “Healthy Lifestyle Package.” The “Healthy Lifestyle Package” began with a single face-to-face education session conducted individually or in pairs. This education session was delivered at the first study visit and centered on targeted nutrition and physical activity advice. The nutritional component of the intervention focused on healthy eating in pregnancy. Participants were encouraged to swap high glycemic index foods for low glycemic index alternatives and were informed about healthy carbohydrate portions. The recommended diet was approximately eucaloric to their typical diet.15 The exercise component of the intervention focused on promoting the benefits and safety of physical activity in pregnancy. Women were advised to exercise per the American College of Obstetricians and Gynecologists' guidance,16 that is 30 minutes of moderate exercise 5–7 days per week, divided into two 15-minute or three 10-minute periods to maximize metabolic benefit.17 The information received at this education session was reinforced through the following delivery channels: a smartphone application, emails every 2 weeks (sent by the research team), and two follow-up face-to-face hospital visits at 28 and 34 weeks of gestation. The content of the emails were standardized to a specific theme on a 2-week basis with some discourse between the researchers and participants where individuals had specific questions. The specifically designed smartphone application was downloaded during the education session from the iTunes or Google player website free of charge and consisted of three components: a comprehensive database of low glycemic index recipes, an exercise advice section, and a homepage comprising daily nutritional and exercise tips and an encouraging thought of the day. Focus groups conducted before study inception guided the content and design of the application (see Appendix 1, available online at http://links.lww.com/AOG/B90). The education session and its delivery channels were informed by two main behavior change theories: control theory and social cognitive theory.18,19 Goal setting for dietary and exercise targets was individualized and based on the “SMART” (Specific, Measurable, Achievable, Relevant, and Time-specific) goals principle.18
Both study groups had a formal antenatal research consultation at baseline and 28 and 34 weeks of gestation. At baseline and the 28-week visit, all patients had the following anthropometric measurements and blood samples collected: weight, height, BMI, midupper arm circumference, and body composition (using Impedimed SFB7 Bio-electrical Impedance Analysis). Maternal weight was recorded at each antenatal consultation. The last measured weight taken before delivery was recorded from medical charts and used to compute total gestational weight gain. Fasting blood samples were collected for measurement of glucose, insulin, C-peptide, and lipids and at 28 weeks of gestation, these blood samples were followed by a 2-hour oral glucose tolerance test performed according to the International Association of Diabetes and Pregnancy Study Groups criteria.13 Gestational diabetes mellitus was confirmed if at least one glucose value was at or above the following: fasting 92 mg/dL or greater, 1 hour 180 mg/dL or greater, and 2 hour 153 mg/dL or greater. Additional data were recorded from the medical charts, including relevant medical and family history of diabetes.
At 34 weeks of gestation, maternal weight was recorded. We assessed fetal biometry ultrasonographically at 34 weeks of gestation including a measurement of fetal anterior abdominal wall width, fetal midthigh subcutaneous tissue, and fetal thigh circumference as reported previously.14 At delivery, cord blood was taken and neonates' birth weight, length, and head circumference were recorded to calculate the Ponderal index (100×mass in g/height in cm3). The Gestation Network's Bulk Calculator 6.2.3 UK was used to calculate birth weight centiles. The following data were also obtained from medical records: time of labor onset, mode of delivery, adverse maternal events, and admissions to the neonatal unit.
All participants in both groups were provided with two 3-day food diaries to quantify glycemic index and glycemic load intakes and two validated pregnancy exercise and lifestyle surveys to ascertain self-reported physical activity levels20—at baseline (preintervention) and in the third trimester (postintervention). The glycemic index and glycemic load of foods were calculated from food diaries analyzed with Nutritics Professional 3.09. The glycemic load of food pertains to a number that approximates how much a particular food will increase an individual's blood glucose level after consuming it. The glycemic index is a measure of how foods that contain carbohydrates increase blood glucose levels.
The prespecified primary outcome was incidence of GDM at 28–30 weeks of gestation. Prespecified maternal secondary outcomes were: gestational weight gain, glycemic index, glycemic load, and self-reported exercise levels in the third trimester. Other secondary outcomes collected for exploratory and subgroup analysis included neonatal birth weight, Ponderal index, and the incidence of large for gestational age (LGA), that is, birth weight greater than the 90th centile.
The rate of GDM in those with BMIs greater than 25 is 12–15%.21 To reduce GDM to a rate of 7.2%, we estimated that 506 women would be required to have 80% power to detect this effect size at a significance of .05, that is, 253 in each arm. The Finnish Gestational Diabetes Prevention (RADIEL) study was a randomized controlled trial of a lifestyle intervention in the pregnancies of obese women and reduced the incidence of GDM from 21.6% to 13.9%, that is, a 35.6% reduction.2 Therefore, we considered that a reduction from 15% to 7.2% could be achievable in this population. All statistical analyses were performed using IBM SPSS for Windows 22.0. Participants were analyzed per their originally assigned groups. Variables were graphically assessed for normality using histograms. Primary analysis was performed with a χ2 test with GDM as the primary outcome. Independent-sample t tests were used for comparison of means and χ2 and Fisher exact tests used to compare categorical variables between the two groups. Treatment effects for binary endpoints are expressed as risk ratios (relative risk) with 95% CIs and mean differences with 95% CIs for continuous variables. Analysis of the mean change in maternal anthropometric and metabolic measurements was performed using independent-sample t tests. We apply a correction to control the false discovery rate over 47 secondary and exploratory efficacy tests using the Benjamini-Hochberg procedure. This is a less conservative procedure than the common Bonferroni correction; it controls the number of false-positive results rather than the probability of at least one false-positive (Appendix 1, http://links.lww.com/AOG/B90).22
Between March 2013 and February 2016, 1,858 women were assessed for eligibility and 565 women were recruited and randomized (278 in the intervention and 287 in the control arm) (Fig. 1). A total of 241 (87%) and 257 (90%) women completed the oral glucose tolerance test and could be assessed for the primary outcome. Mean BMI of participants was 29.3 and mean gestational age 15.5 weeks. The groups had similar demographic characteristics (Table 1). Mean±SD of participants in the intervention group who used the application was, on average, 2.4±2.8 times per week and for an average of 16.2±8 weeks.
The primary study outcome (incidence of GDM) did not differ between the two groups, 37 of 241 (15.4%) in the intervention group compared with 36 of 257 (14.1%) in the control group (relative risk 1.1, 95% CI 0.71–1.66, P=.71).
Women in the intervention group had significantly less gestational weight gain (Table 2). The intervention resulted in lower dietary glycemic index and glycemic load and increased exercise compared with women in a control group (Table 2). However, only differences in dietary glycemic load and exercise participation persisted after multiple correction testing.
After the intervention, there was no difference between the groups in midupper arm circumference (31.5±3 cm vs 31.7±2.7 cm, P=.5), but less whole body fat mass in the intervention group compared with women in the control group, in the third trimester (30.9±7.3 kg vs 33.3±8.3 kg, P=.04, adjusted P=.12). There was no difference between the groups in mean±SD blood loss at delivery (449.5±275.4 mL vs 433.1±245.7 mL, P=.82), postpartum hemorrhage (blood loss greater than 500 mL) (19.5% vs 19.6%, P=.99), antenatal admissions (34.2% vs 32.1% P=.59), or other labor and delivery outcomes (Table 3). There was no difference in maternal lipid profile and markers of glucose homeostasis between groups (Table 2).
There was no difference between the groups in the following prenatal and postnatal characteristics: estimated fetal weight at 34 weeks of gestation (2,520.4±292 g vs 2,535.3±283.8 g, P=.57), Ponderal index (2.71±0.3 vs 2.75±0.5, P=.38), birth weight, and cord blood parameters (Table 3). There was one unexplained stillbirth at greater than 37 weeks of gestation and one preterm delivery with subsequent neonatal death in the intervention group. There was one miscarriage in the intervention group and two in the control group.
This multifaceted behavioral lifestyle intervention incorporating dietary and physical activity advice with reinforcement using mobile health technology did not lower the rate of GDM. However, among prespecified secondary outcomes, the intervention was found to exert a beneficial influence on lifestyle behaviors by way of reducing dietary glycemic load and increasing exercise participation.
Because glucose is important both in the development of GDM and for fetal growth, attenuation of maternal glucose in pregnancy by manipulation diet and exercise is a reasonable clinical target for intervention trials. However, it is possible that the magnitude of the observed dietary and physical activity changes in this lifestyle intervention was insufficient to alter GDM rates. Our results are consistent with others; two recently published large randomized controlled trials were unsuccessful in showing a positive effect of a behavioral lifestyle intervention on GDM.9,23 Recent interest has focused on pharmacologic interventions, but two recent studies did not find that metformin-supported lifestyle interventions reduced GDM rates (Balani J, Hyer S, Shehata H. Metformin versus placebo in obese pregnant women without diabetes [letter]. N Engl J Med 2016;374:2502.).24 Interventions during the preconception period may hold potential to reduce GDM25; however, considerable work is required to show that prepregnancy interventions are effective.
This intervention did result in greater self-reported exercise participation by the third trimester. This is important because physical activity tends to decline with advancing gestation. The intervention involved a face-to-face education session and incorporated SMART goals, techniques that have been shown to improve exercise behaviors in pregnancy.10 This intervention also resulted in lower dietary glycemic load intakes by the third trimester. A low glycemic index dietary intervention has been shown previously to improve maternal dietary intakes,26 reduce GWG, and improve glucose homeostasis.20 Future dietary analysis in this cohort will confirm whether women improved other macro- and micronutrient intakes. Initial unadjusted analysis showed that the mean gestational weight gain in the intervention group was significantly lower than the control group. In addition, a post hoc exploratory analysis revealed fewer neonates in the intervention group were born LGA. Because this was not a prespecified outcome, this finding awaits replication. A potential reduction in LGA is important because neonates at the higher end of the birth weight centile have increased rates of childhood obesity and future cardiovascular and metabolic morbidity.27 Our preliminary findings are supported by the results of the LIMIT study, which noted lower rates of neonates born at greater than 4 kg,23 and the Nelli study, which reported lower birth weight and fewer neonates born LGA.8 All of these results, however, have to be interpreted with caution because they did not persist after multiple testing corrections.
Few studies have looked at the effect of mobile health technology to improve pregnancy outcomes. This trial was well designed with stratified randomization ensuring equal numbers of overweight and obese women in each group. Enrollment rates and prespecified sample sizes were also achieved. This study is also novel in its incorporation of mobile health technology and behavior change techniques in the intervention. We argue the adjustment for multiple testing for secondary and exploratory efficacy outcomes gives us more reassurance that the intervention effect on exercise and glycemic load is robust. We note that this interpretation of the adjusted P values is in contrast to a more traditional Bonferroni correction. The false discovery rate procedure enables one to take the unadjusted P<.05 level and inspect the adjusted P value as a direct estimate of the proportion expected to be false below that threshold given the number of tests conducted. Although not a solution for the multiplicity problem, this approach frames the problem for the reader and gives them more useful information with which to make a determination of “significance.” The primary limitation of the study is that dietary intake and exercise behaviors were self-reported and potentially subject to underreporting or overreporting. Also, the absence of access to a smartphone potentially excluded 30% of lesser affluent and able women. Additionally, we may have had a slightly different cohort and study results if early pregnancy screening for diabetes was performed. However, routine practice in many countries does not involve screening for diabetes early in pregnancy based on risk factors.
Although this mobile health-supported intervention did not affect GDM rates, it did result in some maternal benefits. This study used mobile health technology to support a traditional face-to-face consultation and facilitated positive behavior change. The use of mobile health technology supporting a low glycemic index diet and physical activity intervention with routine antenatal care in at-risk women is a simple, safe, and effective measure to improve maternal exercise levels and improve dietary glycemic load.
1. Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop-Conference on Gestational Diabetes Mellitus. The Organizing Committee. Diabetes Care 1998;21(suppl 2):B161–7.
2. Koivusalo SB, Rönö K, Klemetti MM, Roine RP, Lindström J, Erkkola M, et al. Gestational diabetes mellitus can Be prevented by lifestyle intervention: the Finnish gestational diabetes prevention study (RADIEL): a randomized controlled trial. Diabetes Care 2016;39:24–30.
3. Baci Y, Üstüner I, Keskin HL, Ersoy R, Avşar AF. Effect of maternal obesity and weight gain on gestational diabetes mellitus. Gynecol Endocrinol 2013;29:133–6.
4. Anand SS, Gupta M, Teo KK, Schulze KM, Desai D, Abdalla N, et al. Causes and consequences of gestational diabetes in South Asians living in Canada: results from a prospective cohort study. CMAJ Open 2017;5:E604–11.
5. Sui Z, Moran LJ, Dodd JM. Physical activity levels during pregnancy and gestational weight gain among women who are overweight or obese. Health Promot J Austr 2013;24:206–13.
6. Vinter CA, Jensen DM, Ovesen P, Beck-Nielsen H, Jørgensen JS. The LiP (Lifestyle in Pregnancy) study: a randomized controlled trial of lifestyle intervention in 360 obese pregnant women. Diabetes Care 2011;34:2502–7.
7. Korpi-Hyövälti EA, Laaksonen DE, Schwab US, Vanhapiha TH, Vihla KR, Heinonen ST, et al. Feasibility of a lifestyle intervention in early pregnancy to prevent deterioration of glucose tolerance. BMC Public Health 2011;11:179.
8. Luoto R, Kinnunen TI, Aittasalo M, Kolu P, Raitanen J, Ojala K, et al. Primary prevention of gestational diabetes mellitus and large-for-gestational-age newborns by lifestyle counseling: a cluster-randomized controlled trial. PLoS Med 2011:8:e1001036.
9. Poston L, Bell R, Croker H, Flynn AC, Godfrey KM, Goff L, et al. Effect of a behavioural intervention in obese pregnant women (the UPBEAT study): a multicentre, randomised controlled trial. Lancet Diabetes Endocrinol 2015;3:767–77.
10. Currie S, Sinclair M, Murphy MH, Madden E, Dunwoody L, Liddle D. Reducing the decline in physical activity during pregnancy: a systematic review of behaviour change interventions. PLoS One 2013;8:e66385.
11. Free C, Phillips G, Galli L, Watson L, Felix L, Edwards P, et al. The effectiveness of mobile-health technology-based health behaviour change or disease management interventions for health care consumers: a systematic review. PLoS Med 2013;10:e1001362.
12. O'Higgins A, Murphy OC, Egan A, Mullaney L, Sheehan S, Turner MJ. The use of digital media by women using the maternity services in a developed country. Ir Med J 2014;107:313–5.
13. International Association of Diabetes and Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care 2010;33:676–82.
14. Kennelly MA, Ainscough K, Lindsay K, Gibney E, Mc Carthy M, McAuliffe FM. Pregnancy, exercise and nutrition research study with smart phone app support (Pears): study protocol of a randomized controlled trial. Contemp Clin Trials 2015;46:92–9.
15. Dietary reference values for food energy and nutrients for the United Kingdom. Report of the Panel on Dietary Reference Values of the Committee on Medical Aspects of Food Policy. Rep Health Soc Subj (Lond) 1991;41:1–210.
16. Artal R, O'Toole M. Guidelines of the American College of Obstetricians and Gynecologists for exercise during pregnancy and the postpartum period. Br J Sports Med 2003;37:6–12.
17. Kessler HS, Sisson SB, Short KR. The potential for high-intensity interval training to reduce cardiometabolic disease risk. Sports Med 2012;42:489–509.
18. Carver CS, Scheier MF. Control theory: a useful conceptual framework for personality-social, clinical, and health psychology. Psychol Bull 1982;92:111–35.
19. Bandura A. Health promotion by social cognitive means. Health Educ Behav 2004;31:143–64.
20. Walsh JM, McGowan CA, Mahony R, Foley ME, McAuliffe FM. Low glycaemic index diet in pregnancy to prevent macrosomia (ROLO study): randomised control trial. BMJ 2012;345:e5605.
21. Torloni MR, Betrán AP, Horta BL, Nakamura MU, Atallah AN, Moron AF, et al. Prepregnancy BMI and the risk of gestational diabetes: a systematic review of the literature with meta-analysis. Obes Rev 2009;10:194–203.
22. Glickman ME, Rao SR, Schultz MR. False discovery rate control is a recommended alternative to Bonferroni-type adjustments in health studies. J Clin Epidemiol 2014;67:850–7.
23. Dodd JM, Turnbull D, McPhee AJ, Deussen AR, Grivell RM, Yelland LN, et al. Antenatal lifestyle advice for women who are overweight or obese: LIMIT randomised trial. BMJ 2014:348:g1285.
24. Chiswick C, Reynolds RM, Denison F, Drake AJ, Forbes S, Newby DE, et al. Effect of metformin on maternal and fetal outcomes in obese pregnant women (EMPOWaR): a randomised, double-blind, placebo-controlled trial. Lancet Diabetes Endocrinol 2015;3:778–86.
25. Zhang C, Tobias DK, Chavarro JE, Bao W, Wang D, Ley SH, et al. Adherence to healthy lifestyle and risk of gestational diabetes mellitus: prospective cohort study. BMJ 2014;349:g5450.
26. McGowan CA, Walsh JM, Byrne J, Curran S, McAuliffe FM. The influence of a low glycemic index dietary intervention on maternal dietary intake, glycemic index and gestational weight gain during pregnancy: a randomized controlled trial. Nutr J 2013;12:140.
27. Hermann GM, Dallas LM, Haskell SE, Roghair RD. Neonatal macrosomia is an independent risk factor for adult metabolic syndrome. Neonatology 2010;98:238–44.
Supplemental Digital Content
© 2018 by American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.