The development of the suggested intervention was intricately woven with principles drawn from a conceptual model based on health behavior change theory during pregnancy and illustrates some key elements with motivation and self-efficacy as central health behavior change constructs (46) in facilitating the transformation of pregnant individuals’ dietary habits. Furthermore, the intervention was thoughtfully aligned with the latest guidance provided by the CFG. It encompasses a three-tiered approach designed to empower expectant mothers with the knowledge, skills, and motivation necessary for optimal diet during pregnancy.

The main component of the suggested intervention is four individual 1-h in-person (or virtual, according to participants’ preference) counseling sessions conducted by an RD at key junctures in the pregnancy journey (12, 18, 24, and 30 weeks). The RD uses motivational interviewing which has been has been shown to have had promising results when used in nutritional interventions (47–50). Nutritional counseling during pregnancy is not only the most effective intervention for improving pregnant women’s knowledge and understanding but could also prevent multiple perinatal complications (51–53). Within the individualized counseling sessions, participants will be encouraged to elaborate Specific, Measurable, Attainable, Relevant, Timely (S.M.A.R.T.) goals regarding their eating habits per the CFG guidelines. Hence, every participant’s goals will be tailored to their own verbalized priorities and motivations to improve compliance. During these counseling sessions, the RD will also help participants to identify potential barriers and solutions to achieve their goals. Between the individual counseling sessions, four phone call follow-ups with the RD will be provided to review the goals and solutions chosen by participants (15, 21, 27, and 33 gestational weeks).

The intervention also includes access to 10 informative brief video clips about healthy eating during pregnancy, which are hosted on a private Web platform. The video clips were developed by a team of RDs from our research team, in line with the data demonstrating that (1) education is a necessary component for behavior change (54), and (2) the lack of available resources appears to be a key barrier to healthy eating (55, 56). Their conception also involved experts in prenatal nutrition, communication, and three patient partners. The content was based on current evidence, as well as, on national and provincial recommendations including the CFG and both Health Canada and Quebec National Institute of Public Health’s guidelines for pregnancy. These brief (4–14 min) and informative nutritional video clips cover concerns about healthy eating throughout pregnancy and aim to increase nutrition knowledge to improve motivation (57) and facilitate the achievement of their set goals. The video clips have the following title and the development process as well as their content is available in Supplementary Tables S1, S2.

Lastly, participants randomized to the intervention group are invited to join an online private virtual community on Facebook to discuss and share experiences, thus maintaining motivation and providing social support, as requested by pregnant individuals (58). The Facebook group is animated and moderated by an RD. Each week, the RD shares a new publication that addresses one of the four dimensions of social support (emotional, instrumental, informative, and evaluative) (59) to guide the discussion between participants. This virtual support community was established to provide continuous motivation, peer support, and the exchange of experiences to bolster participants’ commitment to their nutritional goals.

This multi-dimensional approach aims to empower pregnant individuals to make positive and sustainable dietary choices throughout their pregnancy. Moreover, this intervention has been strategically designed to encompass a multitude of behavior change techniques, including goal setting, education, problem-solving, coping planning, and social support (60). By incorporating these techniques, the intervention aims to create a comprehensive framework that not only educates and informs but also enables and supports pregnant individuals in their journey toward healthy eating during pregnancy. These techniques are thoughtfully intertwined within the counseling sessions, the video clips, and the virtual support community, fostering motivation, self-efficacy, and the development of practical skills necessary for achieving and maintaining optimal dietary practices. The synergistic effect of these techniques provides a robust foundation for behavior change, ultimately enabling participants to make informed choices and establish healthy eating habits during pregnancy.

The chosen time frame for intervention commencement (12 weeks), was determined in accordance with the Canadian perinatal care practices. In most cases, contact with medical professionals for pregnant individuals does not occur until around 10–12 weeks, coinciding with the confirmation of pregnancy through ultrasound assessments, such as nuchal translucency and dating ultrasound. Additionally, it is well-acknowledged in medical literature that a significant proportion of miscarriages occurs within the first 12 weeks of pregnancy. Considering these factors, initiating the intervention at 12 weeks allows for a more stable participation, reducing the risk of early dropouts due to miscarriage.

The chosen hybrid mode of delivery, combining in-person, virtual, and phone counseling, is based on previous studies suggesting that in-person counseling with multiple visits might not be feasible and can decrease treatment efficacy (61). The use of Web-based resources to provide nutrition information is based on the reported preferences of French-Canadian pregnant individuals (62). Data also suggest that evidence-based online educational interventions are credible, low cost, and easily accessible, while also complementing the advice provided by healthcare providers (56).

Our choice to prioritize nutrition intervention, rather than covering areas such as physical activity or other lifestyle habits, is primarily based on the simplification of the intervention that takes into account the reality that pregnant individuals are often faced with a multitude of responsibilities and concerns, which can lead them to feel overwhelmed, especially as their time to meet with several health professionals is limited (63). This simpler approach aims to enhance adherence to the program and facilitate the achievement of individual goals and have the potential to be implemented with perinatal care. This choice is also based on the fact that diet-based interventions have been associated with a reduced risk of GDM while mixed-approach interventions seem to be less effective (64). Moreover, current literature clearly demonstrates that pregnant individuals are motivated to eat well during pregnancy and require help to do so (65, 66). Finally, by choosing to target nutrition, we also avoid having to interrupt the intervention due to physical activity contraindications (67, 68). However, we do not deny the importance of staying active during pregnancy, which is suggested by recent Canadian recommendations (67). Physical activity will be measured every trimester to control for this confounding factor. Participants having questions about physical activity will be referred to their general practitioner or a kinesiologist.

Participants randomized to the control group receive usual prenatal care from their general practitioner without access to the nutritional intervention. The usual prenatal care provided to all pregnant individuals in the Province of Quebec is provided by general practitioners, midwives, and obstetricians in private practice, family medicine groups, or in hospitals or birthing centers. Pregnant individuals in Quebec are encouraged to seek prenatal care early in their pregnancy and to meet with a healthcare provider or midwife as soon as the pregnancy is confirmed. Individuals have the option to choose between receiving care from a family doctor, obstetrician, or midwife, depending on their preferences and specific health needs. Midwives provide care as part of a team approach and are available for low-risk pregnancies. Throughout pregnancy, individuals attend a series of prenatal visits with their chosen healthcare provider. The frequency of these visits may vary but typically consists of monthly visits in early- and mid-pregnancy and bimonthly visits in late pregnancy. These visits generally include health assessments, monitoring of fetal development, discussions about pregnancy-related topics, and various routine tests and screening, such as ultrasound and blood tests. Usual prenatal care also involves education and counseling on topics such as childbirth preparation, breastfeeding, and postpartum care. Expectant mothers can also receive guidance through web-based or printed documents from national, provincial, or local resources, which can provide information on the reasons and ways to achieve and maintain a balanced diet and proper nutrition during pregnancy. They may be referred to RDs if considered necessary by their healthcare provider. In general, individuals carrying more than one fetus and those with complex medical histories (including a history of bariatric surgery or eating disorders), pre-existing health problems and GDM can be referred to an RD during pregnancy for appropriate dietary management. As a result, only a small proportion of pregnant individuals in Quebec have access to nutritional management as part of their usual prenatal care.

Participants with a positive diabetes diagnosis in the first trimester are excluded. However, if a GDM is diagnosed during the second or third research visits, participants remain in the project. In the second and third trimesters (V2 and V3), GDM will be diagnosed using the alternate one-step approach suggested by Diabetes Canada’s latest guidelines and the International Association of the Diabetes and Pregnancy Study Groups based on the Hyperglycemia and Adverse Pregnancy Outcome study (5). Therefore, GDM will be diagnosed if one plasma glucose value is abnormal (FPG ≥ 5.1 mmol/L, 1 h following the ingestion of 75 g of glucose ≥10.0 mmol/L, 2 h following the ingestion of 75 g of glucose ≥8.5 mmol/L) (2). The usual conduct for GDM management is to refer to the multidisciplinary team for pregnancies at risk in hospitals or clinics and nutritional therapy is usually initiated within a few weeks (2). Hence, participants diagnosed with GDM will be referred and followed by a multidisciplinary team including RDs, regardless of their group allocation. Since the diagnosis of GDM in the second trimester and the resulting treatment received by the participants may interfere with our results, the effect of this added care will be monitored through retrospective patient medical record data collection and considered in our statistical analyses. However, a late diagnosis at V3 and its subsequent management are not expected to interfere with the results collected at V3 and will therefore not be considered in analyzing our primary and secondary outcomes. Participants are also instructed to contact the research staff between visits or phone calls if they are diagnosed with GDM outside of research visits or if any relevant situations occur during their participation, making them no longer eligible for the study or unable to complete their participation.

At each visit, glucose, insulin, and C-peptide will be measured on fresh blood samples drawn after a 12-h fast, as well as 15, 30, 60, and 90, 120 min after the ingestion of 75 g of glucose. Glucose will be measured enzymatically by the hexokinase method whereas insulin and C-peptide will be measured with an electrochemiluminescence immunoassay. The obtained data will be used to calculate the incremental area under the curve for glucose (iAUC glucose) using the trapezoid method from 0 to 120 min. The following indexes will also be calculated: HOMA-IS [Fasting insulin (μU/ml) × fasting glucose (mmol/L)/22.5] (39); Matsuda index {10,000/[fasting glucose × fasting insulin × (mean glucose × mean insulin)]^1/2} (40); Insulinogenic index [(insulin at 30 min − fasting insulin)/(glucose at 30 min fasting glucose)] (41). The disposition index, which serves as an integrated measure that reflects beta-cell function, will be calculated using Matsuda index × insulinogenic index (41). GDM incidence will be recorded throughout the study based on OGTTs performed during the research visits or based on medical records if not diagnosed during research visits detailed (2).

Three automated, self-administered, and validated 24-h dietary recalls (R24W) using the Rappel de 24 heures Web virtual platform, developed and hosted by the Institut sur la nutrition et les aliments fonctionnels, at Université Laval and validated among Canadian pregnant individuals (69–71). It will be used to measure food intake at each trimester and 3 months postpartum. The R24W platform allows the research staff to randomly select 3 days (2 weekdays and 1 weekend day) in each trimester. On the days the R24W needs to be filled, an email will be sent to the participants, asking them to record all foods and beverages they had the day before (24-h period). Individuals who do not have access to a computer, tablet, or cell phone to complete these online questionnaires (and other questionnaires detailed below) will be invited to complete them on-site, during their on-site visit or the following days at the research center, where a tablet will be provided for the time needed to complete the questionnaires. The 2018 version of the 2015 edition of the Canadian Nutrient File (72) is used to code every food item that is offered on the web platform, enabling the automatic extraction of nutritional data. If a participant fails to complete the R24W, subsequent emails will be sent, within the trimester, until three R24Ws are completed.

Diet quality will be measured using the Healthy Eating Food Index 2019 (HEFI-2019) score, using the mean intakes from repeated R24W (73). This score was recently developed and validated among Canadian adults (74). The HEFI-2019 includes 10 components, of which five are based on foods, one on beverages, and four on nutrients, with a total score of 80 points, with a higher score representing a greater level of adherence to the CFG recommendations on healthy food choices. Since the intervention is based on the 2019 Food Guide, we believe this is the best index to use.

Quality of life will be evaluated in the third trimester using the 36-Item Short Survey (SF-36) (75), which has been previously used among pregnant populations across the globe (76, 77). The SF-36 includes a multi-item scale that assesses eight health concepts: (1) limitations in physical activities because of health problems; (2) limitations in social activities because of physical or emotional problems; (3) limitations in usual role activities because of physical health problems; (4) body pain; (5) general mental health (psychological distress and well-being); (6) limitations in usual role activities because of emotional problems; (7) vitality (energy and fatigue); and (8) general health perceptions.

In this prospective study, we will thoroughly assess both the acceptability and feasibility of the proposed nutritional intervention. To evaluate acceptability, we will employ a mixed-methods approach. Quantitative data will be collected through a 38-item in-house questionnaire among participants receiving the intervention only. This survey aims to capture their perceptions of the intervention’s relevance and appropriateness, and to gather information on their overall satisfaction with the components of the intervention, including the individual counseling sessions, the video clips, and the virtual community, using Likert scales. A content analysis of the Facebook group could also provide qualitative insights into participant engagement and interactions within the community, assessing the depth of discussions, sharing of experiences, and support provided. Attendance to the sessions with the nutritionist will be recorded, as well as their duration. Participants will be questioned regarding their viewing of the video clips at each research visit and intervention follow-up. The engagement metrics on the virtual community include observations such as participation (i.e., if the participant have joined the Facebook group or not), as well as, the number of views, likes, and comments by participants. This multi-faceted approach provides a comprehensive view of participant adherence and engagement with each intervention component throughout the study. Feasibility will be assessed by comparing recruitment rates to predetermined targets. Reasons for non-participation will also be collected from eligible participants who declined participation. Additionally, participant retention rates will be monitored throughout the study, considering a dropout rate of less than 20% as indicative of feasibility. Barriers to participation and reasons for drop-out will also be recorded. Adherence to the intervention will be assessed using activity logs and tracking of engagement metrics within the platform on which the video clips are hosted and the virtual community. An assessment of the intervention’s compatibility with the protocol, available resources, and timeline will be conducted through continuous process evaluations. Therefore, we plan to document resource utilization (time, personnel, and financial resources required for intervention delivery) to compare these to the initially projected allocation and to assess whether the intervention was delivered within the planned time frame, identifying any delays or deviations from the study schedule. This will allow us to comprehensively assess intervention delivery, identifying challenges, successes, and opportunities for refinement with the research team and setting. The outcomes of these evaluations will be pivotal in refining the RCT design, optimizing intervention delivery, and enhancing overall study viability throughout the project.

Even though the proposed project is not expected to pose any risks for participants, the proposed research protocol incorporates measures to ensure participant safety and minimize risks associated with the nutritional intervention and other study procedures. Trimestrial administration of standard 2-h OGTT is considered safe for both the expecting mothers and their fetuses, as this test is performed as part of usual prenatal care. Data monitoring will be constantly done by the treating physicians and the results of all OGTTs will be validated by an endocrinologist. Hence, participation in this project can allow early identification and prompt management of GDM in some participants. The nutritional intervention provided aims to improve diet quality and glycemic profile. Thus, it is very unlikely to cause any safety issues as it is already recommended during pregnancy in individuals with GDM. Moreover, since this intervention does not involve any caloric restriction or strict recommendations, it is not expected to pose any risks such as insufficient weight gain or ketonuria for the participants.

Since the study is open-label, even if very unlikely, it will be easy to determine whether clinical problems that occur are related to the intervention. However, to ensure safety monitoring, adverse events will be closely monitored. Adverse events will be classified into two categories:

During their participation, all study participants will be closely monitored for the occurrence of adverse events. Upon identification of an adverse event, the investigator will assess its severity (mild, moderate, or severe), its expectancy, and its causality. If needed, the investigator will take appropriate actions. The trial steering and monitoring committees will also be notified with relevant information on adverse events for their review.

The sample size calculation for our study is predicated on relevant prior research (5, 91). First, data from one of our previous studies conducted among a sample including pregnant people at risk for GDM indicates that the mean change in FPG from the first to the third trimester was an increase of 0.4 ± 0.8 mmol/L, which is what we expect to find in the control group (91). Second, it was previously demonstrated that an increase of 0.4 mmol/L (1 SD) in FPG in the third trimester is significantly associated with higher odds of unfavorable perinatal outcomes. Our sample size calculation was therefore primarily centered on these findings, as it aims to enable the detection of a minimal clinically significant difference between groups, specifically a 0.4 mmol/L change in delta FPG levels from the first to the third trimester. We postulate that the control group will experience an FPG increase of 0.4 ± 0.8 mmol/L, per previous observations, while the intervention group is anticipated to maintain stable FPG levels (0.0 ± 0.8 mmol/L) (effect size of 0.5). To ensure a statistical power of 80% in detecting this clinically significant 0.4 mmol/L difference between groups, with a significance level (alpha) set at 0.05, we have factored in two considerations. First, we assume a 4% pregnancy loss (92), and an 8% premature delivery (93). Second, given the nature of the study and drawing from experiences and other relevant research, we conservatively anticipate a dropout rate of 20% (24, 26, 89). Consequently, we are in the process of recruiting a total of 150 pregnant women for our study.

The primary analyses will follow the complete case intention-to-treat principle. Changes from first- to third-trimester measurements of the primary and secondary outcomes between the intervention and control groups will be assessed using a two-sample one-sided t-test for means. To determine whether the results obtained depend on an improvement in diet quality, a generalized linear model will be used. Sensitivity analyses will be performed on all outcomes to (1) determine if missingness of the outcome data is associated with prognostic characteristics at baseline and treatment group; (2) compare the results obtained with those following an intention-to-treat analysis where missing outcomes are handled using multiple imputation by chained equations, an analysis on a complete case per protocol, and a per-protocol analysis where missing outcomes are handled using multiple imputation by chained equations; and (3) to explore potential sources of heterogeneity in treatment effects. We thus plan to test the effect of various covariables including age, pre-pregnancy BMI, parity, gestational weight gain, GDM risk factors, fasting blood glucose in the first trimester, GDM diagnosis in the second trimester, adherence to the intervention (with a weighted model based on the level of participation in the intervention), and intervention delivery mode (in person of virtually). Those analyses will ensure a more comprehensive understanding of how different factors may influence the intervention’s outcomes in diverse subgroups of the study population. Programming codes will be tested with a dummy data set to ensure their validity. No preliminary or interim analyses are planned. An exploratory health-economic analysis could be conducted post hoc.