Diabetes prevention interventions for women after gestational diabetes mellitus: an overview of reviews

Abstract Aims To present an overview of reviews of interventions for the prevention of diabetes in women after gestational diabetes mellitus (GDM) with the overall aim of gaining information in order to establish local interventions. Methods Six databases were searched for quantitative, qualitative or mixed‐methods systematic reviews. All types of interventions or screening programmes were eligible. The outcomes were effectiveness of reducing diabetes incidence, encouraging healthy behavioural changes and enhancing women's perceptions of their increased risks of developing type 2 diabetes following GDM. Results Eighteen reviews were included: three on screening programmes and seven on participation and risk perceptions. Interventions promoting physical activity, healthy diet, breastfeeding and antidiabetic medicine reported significantly decreased incidence of postpartum diabetes, up to 34% reduction after any breastfeeding compared to none. Effects were larger if the intervention began early after birth and lasted longer. Participation in screening rose up to 40% with face‐to‐face recruitment in a GDM healthcare setting. Interventions were mainly based in healthcare settings and involved up to nine health professions, councillors and peer educators, mostly dieticians. Women reported a lack of postpartum care and demonstrated a low knowledge of risk factors for developing type 2 diabetes. Typical barriers to participation were lack of awareness of increased risk and low levels of support from family. Conclusions Lifestyle interventions or pharmacological treatment postpartum was effective in decreasing diabetes incidence following GDM. Women's knowledge of the risk of diabetes and importance of physical activity was insufficient. Early face‐to‐face recruitment increased participation in screening. Programmes aimed at women following a diagnosis of GDM ought to provide professional and social support, promote screening, breastfeeding, knowledge of risk factors, be long‐lasting and offered early after birth, preferably by face‐to‐face recruitment.


| INTRODUC TI ON
Prevalence of gestational diabetes mellitus (GDM) is increasing globally, reportedly 2%-26% depending on ethnicity and the diagnostic criteria used. [1][2][3] GDM is related to several adverse outcomes during pregnancy and birth. 4 Complications include pre-eclampsia, shoulder dystocia, children born large for gestational age, neonatal hypoglycaemia and hyperbilirubinaemia. After GDM, lifetime risk of type 2 diabetes is increased, 5 and up to 50% of women with GDM will develop diabetes within 10 years. 6 The highest incidence is reported within 5 years after a GDM pregnancy 7 and varies according to the time of the postpartum examination 7 and diagnostic criteria. 8 A recent meta-analysis including more than 1.3 million individuals found the risk appears to be almost 10-fold higher for Type 2 diabetes and thereby for all-cause mortality. 9 GDM is also a predictor of obesity and diabetes later in life in the offspring. 10 New data confirm that women who develop GDM suffer from a latent metabolic disorder that comes to clinical attention during pregnancy. 11 Thus, GDM helps identify women who have a long-standing, high-risk cardiometabolic profile. 11 Known postpartum risk factors are 2-fold greater risk for elevated body mass index and >3-fold greater risk for an abnormal oral glucose tolerance test. 9 The worldwide increase in Type 2 diabetes has directed attention towards systematic follow-up programmes and clinical routines established to prevent progression of GDM to manifest Type 2 diabetes. 6 In Denmark, general practitioners are responsible for the postpartum follow-up. However, systematic follow-up programmes are lacking in routine clinical settings. 6 Some current approaches are considered not to be costeffective, 1,12-14 although they do help in delaying or preventing diabetes in women with GDM if a structured approach is used. 15 Adherence to preventive programmes seems challenged by women's low perception of the high risk of developing diabetes after GDM. 16 Women with previous GDM called for better continuation of postpartum care, 17 a finding which stresses the importance of programmes containing strategies for healthy lifestyle promotion. 18 A systematic overview of reviews from 2017 concluded that there was'no robust evidence to support the hypothesis that nonpharmacological interventions are effective at lowering the risk', 17 whereas another review concludes that any intervention is superior to no intervention. 19 Seemingly, there is no robust consensus on the content and effectiveness of interventions or the value of screening. 20 The present study is an overview of reviews of interventions for preventing Type 2 diabetes in women following GDM to explore the effectiveness, organization and stakeholders involved, and the perceived risks and barriers for participation in order to establish preventive local interventions.

| ME THODS
To perform the overview, the principles from the Joanna Briggs Institute (JBI) methodology were followed. 21,22 The protocol was registered a priori in the international prospective register of systematic reviews (PROSPERO), registration number: CRD42019131001.

| Searching
An initial search was conducted in the Cochrane Library, the JBI Database of Systematic Reviews and Implementation Reports, PubMed, Epistemonikos and PROSPERO. This displayed numerous systematic reviews about the topic; however, only one overview K E Y W O R D S exercise/physical activity, gestational diabetes, healthcare delivery, nutrition and diet, prevention of diabetes Already known regarding gestational diabetes • The incidence is increasing and follow-up is inadequate.
• One in two women with gestational diabetes develop diabetes within 10 years after birth the highest risk being within the first 5 years.

Findings regarding women with gestational diabetes
• Programmes including physical activity healthy diet and promotion of breastfeeding were effective in preventing diabetes.
• Recruitment should start early as this appears to be the time when women may be most motivated to make lifestyle changes.
• Emphasis should be placed on supporting women to adopt healthy lifestyles and breastfeed.
• Women lack knowledge about the risk of diabetes for themselves and their children and need professional follow-up and social support after giving birth.

Implications for clinical practice in women with gestational diabetes
• Preventive programmes should be offered early in the postpartum period preferably by face-to-face recruitment in local healthcare settings of systematic reviews included randomized controlled trials (RCTs) only. 17 Thus, an overview was decided upon that included qualitative as well as quantitative systematic reviews to draw on a broader range of evidence.
For this study, six databases (Cochrane Library, PubMed, JBI, Embase, CINAHL, Web of Science) were searched for eligible reviews following a 3-step search strategy. An initial search of PubMed was undertaken followed by an analysis of keywords and index terms. Secondly, the search strategy developed for PubMed was refined with assistance from a research librarian for use in the other databases. Thirdly, the reference lists of all included reviews were searched to find additional reviews.
The search was limited to reviews, systematic reviews, metaanalysis and meta-synthesis published in English, Danish, Norwegian and Swedish, published from 2009 to 2019. Predefined search filters regarding'systematic reviews' were applied or specific keywords were included in the search story in the databases, which have no predefined filters (Appendix 1: Search history).

| Inclusion
Eligible for inclusion was peer-reviewed quantitative, qualitative or mixed-methods systematic reviews including meta-analysis or metasynthesis reporting on the effect on incidence of diabetes among women following GDM, organizational aspects and stakeholders involved, women's risk perceptions and barriers for participation in interventions.
Inclusion criteria for the participants were as follows: women with previous GDM participating in postpartum interventions with no restrictions on country, socio-demographic factors (age, ethnicity, parity), socio-economic factors or health-related factors (comorbidity).
Intervention was defined as any pharmacological or nonpharmacological initiative to prevent diabetes in women with previous GDM, provided and organized in any settings, and involving any stakeholders.
Qualitative or mixed reviews exploring women's risk perceptions and determinants for participating in preventive interventions or living a healthy lifestyle were eligible for inclusion. Exclusion criteria were as follows: overviews and reviews that incorporated theoretical studies or text and opinion as their primary source of evidence, and programmes that included women with established diabetes diagnosed before pregnancy.

| Outcomes
The primary outcome was effectiveness in preventing diabetes presented with any estimates. Other outcomes were effect on lifestyle behaviour, data on the organization and stakeholders involved in the interventions. Furthermore, data on risk perceptions and participation barriers were extracted.

| Data extraction
Two reviewers independently screened the titles and abstracts of the identified eligible reviews. Secondly, the full text of all articles was screened by two reviewers when at least one reviewer deemed it potentially eligible. Any disagreement in assessment was solved by consensus. The selection process was recorded in a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram ( Figure 1). 23 Data were extracted by one author and checked for accuracy by a second author using a structured initial data extraction form based on the research question. The form was piloted in four reviews to become familiar with the source results and to ascertain the ease of extraction of data within the reviewers. Any disagreement was solved by consensus.
Characteristics of reviews and details of the interventions are presented in tables and analysed in a narrative summary.

| Quality assessment
For reviews selected for retrieval, the reported quality assessment tool, rating and eventual use of reporting checklist (eg PRISMA checklist) 23 were extracted. Quality assessment was conducted by couples of two reviewers using the standard JBI critical appraisal instrument for Systematic Reviews and Research Syntheses 21 (Appendix 2). Reviews not meeting a priori number (a minimum of 5 'yes') of 11 criteria were estimated to be of low methodological quality and were excluded. Any disagreements that arose between the reviewers were resolved by a third reviewer (AM).

| RE SULTS
The search identified 1996 articles, 999 of which were duplicates, and five additional records were identified from reference lists. Initial screening of abstracts and titles (DH, AM) left 84 articles for fulltext assessment for eligibility (DH, AM). After exclusion with reasons ( Figure 1), 18 systematic reviews met the inclusion criteria. The total number of participants (women with previous GDM) ranged from N = 256 24 to N = 122,877, 25 and sample sizes of the primary studies included in the reviews ranged from 91 to 116,671. 25 In total, 1,427,740 women were included in the overview. Eleven of the 18 systematic reviews consisted of quantitative primary studies, seven of which included RCTs, 18,19,[26][27][28][29][30] three included observational studies, 25,31,32 and one included a mix of quantitative studies 24 (Table 1).
Six reviews consisted of mixed-methods studies, [33][34][35][36][37][38] and one review included only qualitative studies. 39 The systematic reviews were conducted by researchers primarily in the Western world: Australia (n = 2), Canada (n = 1), Denmark (n = 2), the Netherlands (n = 1), New Zealand (n = 1), Switzerland (n = 1), the United Kingdom (n = 4) and the United States of America (n = 2), but one review was from researchers in Brazil, one was from Japan, and two were from China.

| Interventions
Most reviews included lifestyle interventions (diet and physical activity), and six reviews also included interventions promoting breastfeeding. [24][25][26][27][28]32 Details are described in Table 1. Seven reviews included primarily diet and physical activity interventions , 18,19,24,[26][27][28][29] and one review focused on both effectiveness and determinants for adherence to physical activity. 33 Six reviews included RCTs and cohort studies with both lifestyle and pharmacological interventions, 24,[33][34][35][36][37][38] The duration of the interventions varied substantially from 4 weeks 25 up to 3 years. 18,29 The time to follow-up also varied from 6 weeks to 16 and 19 years. 24,25 Duration of the screening programmes was not reported universally. 30,35,37 Three reviews included postpartum screening interventions in women with previous GDM, for example reminders and determinants for participation. 30,35,46 Ten of the quantitative reviews reported on measures of effectiveness, organization of interventions and the stakeholders involve d. 18,27,28,30,31,[34][35][36]37,38 Seven reviews including qualitative or mixed-method studies described determinants or barriers for participation, adherence to changes in lifestyle and women's risk perceptions (

| Intervention effect
Twelve reviews presented data on incidence or risk reduction of postpartum diabetes. 18,19,[24][25][26][27][28][29][30]32,33,37 All but two also presented estimates on risk associations between behaviours and postpartum diabetes. 29,37 However, meta-analyses were not performed due to heterogeneity of the study populations. Thus, the effectiveness of the interventions was presented in both descriptive and analytic terms (Table 2. The five reviews that included interventions promoting breastfeeding found a positive impact; one review concluded that exclusively breastfeeding for 6-9 weeks significantly reduced the risk of diabetes compared with formula at more than Women with a diagnosis of GDM in the index pregnancy.

N = 256
To assess the effects of reminder systems to increase uptake of testing for T2DM or impaired glucose tolerance in women with a history of GDM. Reminders of any modality (post, email, phone (direct call or short SMS text) to either women with a history of GDM or their health professional, or both.
Mode 3 m pp postal reminders 1) to the woman only, 2) to the physician only, 3) to both. Women and physicians were contacted 3 times during 1 y FU Duration

Effectiveness of (breastfeeding, diet, physical activity, pharmacological) interventions, and screening on reducing diabetes Stakeholders involved Organisation
Pedersen 2017 Denmark (20) Diet, PA No specific intervention or components were found superior NS reduction of T2DM incidence (tendency only) SS pooled estimate of absolute risk reduction (−5.02 per 100 (95% CI: −9.24;−0.80) SS effect in the subgroup of participants >40 y (T2DM incidence 8% in intervention group vs. 20% in control group, n = 175, p = 0.018 Tendency of poorer effect starting during pregnancy or very early pp (≤6 w) vs. interventions started >6 w pp SS changes were found for PA but not for diet Biomarkers of insulin resistance Generally, results were consistent within trials 2 showed NS effect on fasting glucose in spite of a SS intervention effect on other measures of insulin resistance

TA B L E 2 (Continued)
A mixed-method review found in four of 28 RCTs an increased level of physical activity 3-12 months after intervention; social support, childcare issues and cultural background impacted significantly on the effectiveness of interventions. 33 However, only a third of the

RCTs on exercise showed an effect on measures of physical activity
and on biomarkers of insulin resistance. 26  and changed dietary intake, although they were more effective in women without excessive gestational weight gain. 36 Pedersen et al.
concluded that lifestyle interventions in RCTs increased physical activity but not changes in diet (20). 19 Another review concluded that behavioural interventions had a significant effect on eating patterns during pregnancy and leisure time physical activity in the first year postpartum. 36 Timing, duration and recruitment to interventions Timing was of importance regarding effectiveness; early postpartum (2-6 months) interventions were most effective. 28 Jones et al. showed the start-up time for interventions was divided into three distinct periods: prenatal and early and late postpartum. 18 Pedersen et al. found that lifestyle interventions started during pregnancy were less effective than interventions implemented 6 weeks postpartum. 19 Furthermore, effect was superior if interventions lasted more than 1 year, but effect was less at 3 years of follow-up than after 1 year. 19 Recruitment method impacted upon participation rates. A review including eHealth interventions with a RCT design concluded that postpartum screening with follow-up by telephone increased screening rate and reduced fasting blood glucose levels in women with previous GDM. 37 Recruitment during pregnancy or the early postpartum period increased the participation rate more than 40%, especially if face-to-face contact was used in the GDM care setting. 31 In contrast, a mailed invitation and/or telephone contact later in the postpartum period decreased the participation to less than 15%. 31

| Perceived risk and barriers
Mixed-method studies found that the effectiveness of the interventions depended on incorporation of factors of importance for participation. 33 These were typically interacting behavioural factors, for example lack of support from family and professionals, cultural sensitivities and lack of resources and information. Typical barriers for women's participation were lack of information during pregnancy, lack of knowledge of risk factors, preventive behaviours and, explicitly, the role of physical activity. [34][35][36]38,39 For some women, the importance of physical activity was perceived to be relevant only to control blood glucose and lose weight during pregnancy. Thus, only 7% of women believed that physical activity would decrease the risk of diabetes later in life. 36 One review that included physical activity interventions found that use of pedometers was not effective. 33 Another review concluded that although women may continue eating healthy postpartum, some stopped being concerned with what they ate because they perceived that their diet no longer had an impact on the health of the child. 35 Furthermore, during breastfeeding some women increased their food intake. 35 The review found that only a minority of women were conscious of their high risk of developing diabetes later in life. 35 Despite an intention to maintain a healthy lifestyle, most women did not, and only one in three reported a sufficient level of daily physical activity. 35

| Organization of interventions
Twelve of the systematic reviews reported on organizational aspects of the interventions (Tables 2 and 3). 18,19,27,28,30,31,[34][35][36]37,38,39 ranging from unspecified to involving several settings: participants' home, community-based practice or health centre, GDM care setting, public/urban hospital, university health system, private practice, women's hospital, clinic or ward, pregnancy service, urban antenatal clinic, private obstetrician clinic, GDM clinic or unit, medical centre at tertiary hospital and university prenatal clinic. 31 One review reported behavioural interventions in home-based settings only, 18 and another review reported interventions in fitness centres. 19 Women expressed preferences for programmes that allowed access from home (eg Internet-based or telephone intervention), thereby overcoming accessibility issues. 36 Women also expressed a need for support from a lifestyle coach and provision of family friendly programmes. 36

| Stakeholders involved
Fifteen of the systematic reviews reported on specific stakeholders involved in the interventions (Tables 2 and 3). 18,19,24,[27][28][29][30][31][33][34][35][36]37,38,39 These ranged from a few unspecified healthcare professionals and researchers up to nine different professions (trained counsellor, exercise physiologist, dietician, lifestyle behaviour case manager, research nutritionist, lactation consultant, peer educator, diabetes educator and research nurse). 27 The most prevalent stakeholders involved were dieticians,however, their role was not described in detail. One review concluded that midwives played an important role as primary carers, as they were ideally positioned to educate and engage women in lifestyle programmes during pregnancy and following the postpartum period. 36 In postpartum screening programmes that only involved obstetricians. 37 both women and clinicians were more satisfied with eHealth programmes with remote monitoring, and planned and unplanned clinic visits were reduced by 50% and 66%, respectively.37 In contrast to the intensive GDM monitoring during pregnancy, the women reported they felt abandoned by healthcare providers postpartum and found difficulties balancing household demands and following a healthy lifestyle. 36 A need was identified for more proactive support and postpartum care, together with the need for information regarding the risk and complications of diabetes for themselves and their offspring. 38

TA B L E 3 Findings from systematic reviews including qualitative studies
Systematic review

Determinants
Putting others before yourself, putting off lifestyle change, lack of support from healthcare professionals, being a healthy role model for families, accounting for childcare issues, social support and cultural sensitivities Interventions (Random control trials (RCTs) that incorporated these factors were associated with effectiveness Education about how to reduce future risk of type 2 diabetes mellitus (T2DM) and pedometers in interventions were not associated with effectiveness Healthcare professionals Doctors Healthcare providers Practitioners Researchers -

Dasqupta 2018
Canada (32) Participation (calculated as the proportion of) those invited who actually enrol in different intervention programs varied substantially Penetration (coverage of the target population) calculated as the proportion invited to participate in interventions for preventing diabetes was 85-100% When recruitment occurred during pregnancy or early postpartum (pp), participation was >40% or more, especially if face-to-face contact was used within the gestational diabetes mellitus (GDM) care setting, but participation <15% in mid/late pp with mailed invitation and/or telephone contact Lifestyle change influences Determinants (interacting influences on pp behaviour): Role as mother and priorities; social support from family and friends; demands of life; personal preferences and experiences; diabetes risk perception and information; finances and resources; format of interventions Barriers Women identified themselves primarily as mothers who prioritized their family above themselves, and needed resources, time, energy, information and support to encourage healthy diets and levels of activity Important to adapt interventions to the target population and facilitate familyfriendly changes because the mother's own diabetes risk was unlikely to motivate change without her perceiving benefits for her children Some of the most beneficial aspects of groups (e.g. forming supportive relationships) are impractical for most to commit to in the long term Adherence to health behaviours: Health behaviours, impact on adoption of: women's own perception of health, risk perception, risk and knowledge regarding diabetes, impact of health beliefs and psychosocial factors, social support, self-efficacy Determinants Information during pregnancy, recall of advice/remembered receiving diabetes prevention information, perception and awareness of risk of diabetes, knowledge of risk factors and preventive behaviours, knowledge on diabetes and role of PA, social support from partner, family and friends, appropriate childcare Partner/family support, high social support, high self-efficacy, companions, community safety, transportation, centre-based programme Barriers Lack of assistance for child care/ constraints related to children, lack of time, time constraints, enjoyment of activity, necessity to prevent later health problems, self-perceived health status, continuing support and education post partum, beliefs about health and illness, perceived risk, self-efficacy, perceived personal control, beliefs in the benefits and barriers of lifestyle modification, financial constraints, lack of motivation/fatigue, Determinants for healthy lifestyle pp (diet, PA) Despite women expressed they intended to live a healthy lifestyle pp, it was generally not achieved. Among women with GDM in the past 6 months (m) −2 year (y) unhealthy diet was prevalent, only 34% reported sufficient PA. Women with previous GDM do not perceive themselves to be at increased risk of future diabetes. 90% of women (US population) recognized GDM as a risk factor for future diabetes, only 16% believed they themselves were at high risk, though the proportion increased to 39% when asked to estimate their risk assuming they maintained their current lifestyle. 40% of women with a history of GDM were very worried about developing diabetes in the future, 46% a little worried and 14% not worried at all. Some women increase their food intake during breastfeeding Determinants for diet Self-efficacy was associated with high vegetable consumption, ability to cook healthy foods, and reporting that healthy diet is not a difficult change and that dislike of healthy foods by other household members is not a barrier for them. Moreover, self-efficacy when busy and not reporting a dislike of healthy foods by others at home were associated with high fruit consumption Determinants for PA Independently associated with high self-efficacy and social support. Barriers for PA Lack of time and/or energy, child care support, motivation, knowledge about GDM, social support, support from health care provider, enjoyment of PA, not feeling well, emotional distress, financial barriers, domestic responsibilities such as cooking, feeling of solitude, dullness and isolation from family and friends, poor body image, bad weather, considering oneself to be too young to be on a restricted diet, obstacles at work, unsuitable local neighbourhood, no access to exercise equipment, cultural expectations, bad weather; considering oneself to be too young to be on a restricted diet; unsuitable local+neighbourhood or no access to exercise equipment; cultural expectations e.g. needs of women come last in the family. Women who perceived themselves to be at no or slight risk of diabetes were less likely to modify their lifestyle. Many women tried to continue eating healthy pp to protect their health However, some pp women felt they no longer had to worry about what they were eating as it would no longer impact the health of the baby. Intentions of healthy lifestyle may be there, but many do not succeed in continuing modifications. May be influenced by their perception of risk of future diabetes and particularly by self-efficacy and social support Barriers to screening pp Not considering the test necessary, declining testing, unable to complete test, testing not affordable, uninformed, lack of understanding of need for test, practice being too busy, time pressure, lost requisition, recent delivery experience, baby's health issues, adjustment to the new baby (emotional stress, feeling overwhelmed and lack of time and burden of child care), concerns about pp and future health (feeling healthy and not in need for care, and fear of receiving bad news), experiences with medical care and services (dissatisfaction with care and logistics of accessing care).

| Principal findings
This overview included 18 systematic reviews, seven of which consisted of qualitative or mixed studies. Eleven of the reviews reported on effectiveness, organization as well as stakeholders involved in the interventions.

| Effectiveness, timing, duration and recruitment to interventions
The scope, components and duration of the interventions varied.
A majority of the lifestyle programmes including breastfeeding, ; both interventions were highly efficient in reducing progression to Type 2 diabetes, 15 after 10 year, the incidence was reduced by 35%-40%. 47 Despite the long term increased risk among the women, the interventions were in general followed up for only a few months, but showed the importance to maintain support for lifestyle changes for a longer period. 40 The most comprehensive meta-analysis found that among women (average age 30 years) minimal changes in anthropometric measures over a short period translate into a 25% risk reduction of diabetes. 40 A need for long-term follow-up was similarly underlined by qualitative systematic reviews, where a need was expressed for GDM follow-up and proactive support from lifestyle coaches and healthcare professionals.

Systematic review
Determinants and barriers for diabetes prevention (lifestyle behaviours, diet, physical activity, and screening)

Stakeholders involved Organisation
Participation determinants Preference for a programme of support that allowed access from home (e.g. internet based) and/or support from 'lifestyle coach'. Early pp interventions using telephone experienced a greater percentage of weight loss and lifestyle behaviour changes. Increased social support and facilitating increased PA self-efficacy, as well as a ''family friendly'' approach, may help increase lifestyle recommendations A healthy diet (more vegetables and less fried foods) was too great a change from their current behaviours PA Concern about progression to diabetes were not observed to increase their levels of PA or lose weight as advised during pregnancies. Women exercised in pregnancy to control their blood glucose levels, whereas pp exercise was perceived as important only to assist weight loss. Participation in screening and lifestyle programmes is promoted by early postpartum recruitment, especially with face-to-face contact in GDM healthcare setting. Thus, implementation up to 6 weeks postpartum proved to be superior; preventive interventions should be long-lasting to be most effective although some interventions showed a decreased effect after 3 years.

| Organization of interventions and stakeholders involved
Most interventions involved professionals from different fields; the roles or skills of the professionals have to be described in detail to legitimate their involvement. Midwives play an important role as primary carers for women during pregnancy and childbirth and may play a part after birth in engaging women in lifestyle programmes, 36 trusted dieticians and coaches for training in family friendly programmes are needed. Furthermore, the findings stressed the importance of acknowledgment of the women's need of support from professionals and families for participation. This is in accordance with a previous review, which also found it was necessary to involve the healthcare system as well as the family context in preventive programmes. 48

| Perceived risks and barriers for participation
The findings revealed lack of knowledge of risk factors among the women. Furthermore, the qualitative reviews found that women reported a lack of postpartum care and demonstrated a lack of knowledge that GDM begets type 2 diabetes. 6 Thus, the women expressed a need for reliable information and support from professionals (eg information about risk of diabetes for the woman as well as their child).
A theory-based study found effectiveness of a health promotion intervention among adults at high risk of diabetes. 49

| Strengths and weaknesses of the study
An overview of reviews considers the highest level of evidence; the methodology followed was rigorous and comprehensive, with duplicate screeners throughout title and abstract screening, full-text review, quality assessment and data extraction.
This overview explored effectiveness and determinants for preventive interventions among women following GDM. According to the aim, drawing a broader range of evidence, this overview added value as it included qualitative data in order to explore typical barriers for participation in certain interventions. 38  Only three reviews were from emerging countries, thus limiting the validity of our findings and conclusions to settings in the Western world.
A review on clinicians' views was not included due to criteria for participants; the findings showed gaps in postpartum screening practice, and a need to improve collaboration among stakeholders and education about GDM. 46

| Future preventive interventions
The need for interventions after GDM is evident, and most findings and recommendations are reproducible for programmes in a local healthcare setting. Supporting an intended healthy lifestyle postpartum seems to be the challenge; unhealthy diet and insufficient level of daily physical activity were very common among the women.
Participation barriers regarding screening should be taken into account, for example healthcare provider (specialist or family physician) not seeing the patient, lack of communication and collaboration between healthcare providers and the women, inconsistent guidelines or lack of familiarity with guidelines, and no awareness about the woman's history of GDM. 35 To motivate women to take advantage of healthcare opportunities, automatic reminders in patient charts or electronic medical records would be beneficial. Postpartum screening programmes may increase participation by using telephone follow-up and eHealth.
Behavioural interventions are often provided face-to-face; however, eHealth may prove less costly; a meta-analysis found strong evidence for use of mobile phone apps for lifestyle modification in diabetes. 50 The lack of awareness among the women for the need of screening should be addressed, and healthcare providers should adhere to newest available guidelines.
A combination of approaches may be most appropriate, for example online information, target-setting and options to arrange video calls with dieticians and contact with local groups of women who also had experienced GDM. 39

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interests with respect to the research, authorship or funding of this research project.

AUTH O R CO NTR I B UTI O N S
All authors have given final approval of the version to be published.

DATA AVA I L A B I L I T Y
Data are available in the included systematic reviews and Tables 1-3.