Design

A nonrandomised single-arm design with a 12-week follow -up, all participants received the WWDP+ intervention. Feasibility outcomes were assessed at 3-months and 6-months post baseline while effective signal outcomes were assessed at baseline, 3-months and 6-months post baseline. The study protocol can be found in S1 File.

Study setting and recruitment

Participants were recruited through a four week, funded, social media campaign targeting midlife women with T2DM through Facebook and Google adverts. Women were invited to respond if they were interested in engaging with diabetes research in any of the following ways 1) Advising our team on research directions (public and patient involvement), 2) participating in research or 3) becoming a peer supporter for other women living with T2DM. A Microsoft form was accessible for women with T2DM to complete, to provide their demographic and contact details and to indicate which activities they were interested in. Those interested in research participation were emailed and given a project weblink for further information. A study-specific webpage included the aim of the study, eligibility criteria, a participant information sheet, and a link for a screening questionnaire via Qualtrics [20].

Women confirming interest through the webpages were contacted, given the opportunity to ask further questions and then asked to confirm their eligibility within 24 hours by completing a brief screening question. The screening questions comprised basic data on their physical health, demographic information and the Diabetes Distress scale (DDS) [21] which had indicated a strong effect signal in the previous trial [17,18]. The DDS measures the extent of diabetes distress related to their medical regimen, emotional, physician care and interpersonal domains. Those scoring over 2 (from a scale of 1–5) on the DDS were eligible because the prior feasibility trial reduced elevated diabetes distress. Pre-specified threshold reflects the level at which distress becomes clinically meaningful, distinguishing participants experiencing minimal or no distress (<2) from those with moderate to high distress (≥2). Using this cut-off allows consistent classification of participants and aligns with established conventions for interpreting DDS scores [12]. Participants received the informed consent form online and provided written consent by digitally signing it on the Qualtrics [20] within 7 days before participating in the study. All participant recruitment including follow up were conducted between May 2024 and December 2024.

Inclusion criteria were: 1) women living in the UK with access to diabetes care in the NHS, 2) aged ≥ 45 and ≤65 years with a current diagnosis of T2DM 3) taking insulin or oral medication, 4) receiving diabetes care only through their GP 5) able to read/speak English sufficiently to take part, 6) access to a computer or laptop, internet access and IT literacy (defined as being able to buy a product over the internet, search for a topic and send an email), 7) agreeing to having a Facebook account, 8) scored 2 or above on the Diabetes Distress scale [21]. Exclusion criteria were concurrent participation in another research study that could interfere with the assessment of the programme’s outcomes.

Sample size

The target sample size was based on guidelines for feasibility studies to test procedures and processes and to estimate some of the parameters for the definitive trial [22]. In line with methodological recommendations for pilot and feasibility trials [23,24] which indicate that relatively small samples can be adequate for estimating feasibility parameters, but larger samples are required when the primary aim is precise estimation of a standard deviation, this study aimed to recruit a total of 40 participants.

Materials

Assessments were conducted at three time points including baseline (T1) and immediately post-intervention (T2) and 6-months post baseline (T3) of the primary outcomes, with goal setting assessments at 3- and 12-weeks through online survey via Qualtrics [20]. An online survey link was sent to each participant’s email address. Completion of the baseline and post and follow up questionnaires required around 30–60 minutes. The programme materials were developed and delivered by the research team and peer supporters for peer group. Participants had access to the online peer support group between 9:00 am and 5:00 pm daily, with the peer group being moderated during these hours and programme resources available via the website. The 12-week intervention followed a structured weekly flow with different educational modules in the E-book and peer support sessions in the group (S2 File).

Outcomes which improved during the previous WWDP trial [17] were targeted. All baseline and post study DDS were assessed within 48 hours of receipt of the completed survey. If the score was 3.5 or over, indicating significant distress [12], participants were contacted by the research team and signposted to resources to support their psychological wellbeing, e.g., the Samaritans or GP depending on how immediate the research team perceived their need for support either over the phone or email.

Feasibility outcome measures

Feasibility outcomes related to the new programme components, and therefore not evaluated in the previous trial [17], were considered primary outcomes:

• Recruitment rate via the social media campaign- calculated as the proportion of eligible women who consented to participate and completed baseline assessments.
• Intervention/Peer Group attendance – calculated as the proportion of participants who logged into the online peer group site on at least three occasions over the 12 weeks.
• Data quality – calculated as the proportion of participants who provided data at all three time points.
• Pre-defined progression criteria for the trial to assess feasibility were set at achieving at least a 50% recruitment rate, 70% program completion rate and data completeness of 70%.

Secondary outcome measures (self-reported)

• Diabetes Distress was measured using the Diabetes Distress Scale [21] that contains 17-items categorised into four subscales labelled as Emotional Burden, Regimen Distress, Interpersonal Distress and Physician Distress.
• Confidence in self-management was measured using the Diabetes Management Self-Efficacy Scale (DMSES UK) [25].
• The Short-Form health survey (SF-36) is 36 item self-report scale assessed health-related QOL across 8 domains: physical functioning, physical health role limitations, emotional problems, role limitations, pain, mental health, social functioning, general health [26].
• Assessing positive or negative lifestyle factors was measured using International Physical Activity Questionnaire (IPAQ) [27] assessed physical activity whereas measuring dietary behaviours on diabetes management were assessed through UK version Diabetes and Diet Questionnaire (UKDDQ) [28].
• To assess smoking prevalence, a subset of key questions from the Global Adult Tobacco Survey [29] were used.
• Participants’ sleep activity and quality was measured using the 21-item General Sleep Disturbance Scale [30].
• Furthermore, the standard Greene Climacteric Scale assessed menopausal symptoms, with validated 21 items assessing subscales for vasomotor, somatic, psychological (anxiety and depression) with strong reliability [31].
• The Goal Assessment Scale (GAS) [32] was used to assess whether participants had achieved their goals within seven days of receiving the form in week 12 using a + 2 to −2 scale.
• Physiological measures: Anthropometric measures covered body weight, height, waist and hip circumferences and body mass index (BMI). Women were provided with a video from reliable source by Diabetes UK [33] which demonstrated how to take the various measurements using a tape measure and scales, and how to calculate their BMI.

The Women’s Wellness in Diabetes Programme + (WWDP+) Intervention

The WWDP+ [19] comprised three components a) website and e-book of 12 chapters/weeks including reflections, information and interactive exercises b) Personalised goal prioritisation activity and c) a moderated peer support Facebook group.

Website and eBook component of WWDP+

The WWDP+ is a 12-week multi-modal intervention delivered through an eHealth website [17–19]. The online intervention encourages women to reduce diabetes complications and promote healthy behaviours and is theoretically underpinned by Social Cognitive Theory [34] with an emphasis on behaviour change through mastery experiences, vicarious experiences, and verbal encouragement through self-efficacy theory. It includes a web interface, including podcasts, and an interactive electronic book (eBook) accessible from any electronic device offering intervention instructions and guides for participants, a logbook to record relevant health and lifestyle information weekly and factsheets on diabetes-specific selected topics. Table 1 outlines the targeted health knowledge and behaviours addressed in the intervention. Further information on the intervention can be found in S1 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Targeted health knowledge and behaviours [17].

https://doi.org/10.1371/journal.pone.0345517.t001

Peer support component

Online peer support is facilitated by membership of a closed Facebook group and provides the opportunity to meet other women who are living with T2DM, to share experiences and information, learn more about diabetes and living with a lifelong condition, and contribute to communal online activities. Participants were expected to log into the online peer group on at least three times and/or visit to the website once per week for four of the twelve weeks. Specialised trained peer supporters moderated the peer activity daily from Monday to Friday. These moderators were midlife women with T2DM who had undergone over ten hours of training, delivered by DB and an experienced peer support trainer. The training covered both peer support content and technical aspects of online moderation supported by a peer supporter guidance handbook.

Personal goal setting component

As part of the intervention, participants received via email a “Personal Goal Evaluation” (PGF) form with a unique identification number and including a Goal Assessment Scale (GAS) [32]. Participants were encouraged to read week 1 and 2 of the eBook and collaborate with their peer group to set up their goals. They had until the end of week 3 to finalise their goals of participation. Participants were encouraged to set “SMART” (i.e., specific, measurable, attainable, realistic, and timely) goals and to focus preferably on one or two goals throughout the intervention. To enhance commitment, participants were also encouraged to think about potential barriers and solutions to their action-plans. They rated their confidence (from 1 to 10) in achieving their chosen goal. Those with a confidence level of less than 7 were encouraged to re-visit their goal or explore ways to boost their confidence. By week 3, participants were expected to be ready for goal implementation. Once participants selected personalised goals at 3 weeks, which were subsequently rated post-intervention using GAS. Each goal was scored on a five-point scale from –2 (much less than expected outcome) to +2 (much more than expected outcome), with 0 representing the expected level of achievement. For participants with multiple goals, scores were grouped according to common goals and analysed separately for each goal. Goals common to multiple participants were aggregated to show the overall goal attainment. This approach allowed goal attainment to be summarised both at the individual and goal-specific level.

Data analysis

Descriptive statistics were used to characterize rates of demographic characteristics, recruitment, and data completeness. For continuous data, means and standard deviations were calculated, as well as frequency and percentages for categorical variables. To explore the potential impact of the intervention, statistical analyses were performed to assess changes in secondary outcomes over time. For normally distributed continuous variables through Shapiro-Wilk test, paired t-tests for variable on IPAQ and UKDDQ were used to compare pre- and post-intervention means, while the Wilcoxon signed-rank test for variables on DDS, DMSES UK, SF-36, HbA1c, BMI, weight, GSDS was applied for non-normally distributed data. Mean changes and standard deviations were reported, and where relevant, effect sizes were calculated to estimate the magnitude of change. Given that this study was not powered for hypothesis testing, statistical significance was not the primary focus; instead, results were interpreted with reference to Minimal Clinically Important Differences (MCID) from existing literature where available. Data were analysed using the Statistical Package for Social Sciences (SPSS) version 29.0 for Windows. Due to the nature of the intervention and the non-randomised design of the study, neither the participants nor the researchers delivering the intervention were blinded. However, the outcome assessor responsible for collecting follow-up data was blinded to participants’ intervention exposure. Each participant was assigned a unique identification (ID) number, and all outcome data were stored and analysed using these blinded IDs. The outcome assessor accessed and evaluated results based solely on the de-identified dataset to minimise potential assessment bias. For each analysis, only participants with complete data at the relevant time points were included. For paired comparisons (pre-post-test and follow-up), participants were required to have data at both time points; those with missing data were excluded from that specific analysis (listwise deletion). There were no missing data within each time point, as all participants who were included completed all items at each assessment.

Ethical approval

Ethical approval was granted by King’s College London Ethics Committee (Ref: HR/DP-23/24–34435) on (05/03/2024) and (Ref: RESCM-23/24–34435) (10/05/2024). This trial was retrospectively registered at ISRCTN (Registration number: ISRCTN93338547). At the time of study initiation, the study was initially not considered to fall under the category requiring prospective registration. The authors confirm that all ongoing/completed and related trials for this intervention are registered.

Baseline characteristics

Thirty-four midlife women with T2DM participated to the study. Participants had a mean age of 55.38 years±5.15. The mean duration of T2DM was 6.83 ± 5.1 years and ranged from 1 to 20 years, with 65% of participants having a self-reported HbA1c level exceeding 7% (53 mmol/mol). Mean Body Mass Index (BMI) of the sample was 33.13 ± 7.31.

Most participants had a family history of diabetes (n = 25). Participants reported that 24/34 were treated with oral hypoglycaemic agents (OHA) only, 8/34 with OHA and insulin and 1/34 with insulin only. The majority of participants were married (n = 20). Regarding educational attainment, 14/34 had completed undergraduate studies, 8/34 held postgraduate qualifications, and 8/34 had completed college education. In terms of working status, 13/34 women reporting working full-time and 12/34 women with working part-time. The detailed baseline demographic data is provided in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Socio-demographic characteristic of participants (n = 34).

https://doi.org/10.1371/journal.pone.0345517.t002

Feasibility outcomes

Eighty-three women expressed interest in being a research participant and they were sent information about this study (Fig 1). Forty-three (52%) of these women were deemed ineligible through not meeting inclusion criteria or did not give their consent. The remaining forty were then invited to complete a baseline questionnaire. Six women did not complete the baseline assessment, and thirty-four women entered the study. The feasibility targets were assessed using pre-defined progression criteria; thus, the recruitment rate of 77% of eligible women (n = 34/44) met the pre-determined recruitment target of 50% at baseline.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT flow diagram.

https://doi.org/10.1371/journal.pone.0345517.g001

In terms of data completeness, 88% of participants completed the 3-week personal goal-setting forms, 82% completed the 12-week personal goal assessment forms, 82% completed the 3-month assessments, and 79% completed the 6-month post-baseline assessments. These completion rates exceeded the pre-determined threshold of 70%.

A total of 30 women (88%) completed the intervention. For peer group attendance rates, the data suggested that at least 28 women (82%) logged onto the online peer group on at least three occasions over the 12 weeks. This met the pre-determined peer group attendance rate of 70%. As Fig 2 suggests, despite stable group engagement during the first eight weeks of online peer group use, fluctuations were observed in the final four weeks during which fewer women logged on. Of the 215 posts analysed, 43 were from peer supporters and were excluded from the analyses. The remaining posts, generated by participants, showed the highest engagement on Mondays, Tuesdays, and Wednesdays, mainly between 12:00 pm and 8:00 pm. This provides insight into peak activity periods. Further details of patterns of daily use can be found in Fig 2 and S1 Fig.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Patterns of peer support over the 12-weeks intervention.

https://doi.org/10.1371/journal.pone.0345517.g002

Goal setting findings

Among the 30 women that completed the goal setting by 3-week mark, 30% chose one goal while the majority (70%) selected two goals to focus on. Table 3 shows from the eight available options, participants chose the goals.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. The proportion of the preferred goals at week 3.

https://doi.org/10.1371/journal.pone.0345517.t003

At T2, 28 women (82%) completed their personal goal assessment forms, with data showing that 61% had better than expected progress with their goal, 7% had made expected progress and 32% had fared worse than they had expected.

Health related outcomes findings

Paired comparisons were conducted between T1 and T2, and between T1 and T3. Only participants with data available at both time points were included (Table 4). Participants experienced a mean reduction in diabetes distress by 1.52 from T1 to T3. Particularly, emotional burden (1.19), regimen distress (1.66), interpersonal distress (1.59), and physician distress (1.71) were found. These reductions exceeded the MCID of 0.25 established for the DDS [35]. Compared to the baseline, there was an increase in the mean of diabetes management self-efficacy as measured by the DMSES UK at T3 (42.02). Self-reported HbA1c levels also showed a mean decrease of 13 points exceeding the MCID for HbA1c of 4–5 mmol [36] indicating a clinically meaningful difference at T3. Furthermore, the MCID of health-related quality of life suggests three to five points [37,38], therefore, the findings on general health also exceed 27.49 points which showed significant improvement.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. The health-related data of study participants across three time points of the study. Descriptive statistics are presented as mean ± SD or median (IQR), as appropriate.

https://doi.org/10.1371/journal.pone.0345517.t004

The Wilcoxon signed-rank and paired t-tests were run to compare the intervention effects between T1 and T3. Findings on DDS (z = −3.81; p < 0.001) suggested a clear trend towards improvements at T3. Subsequently, a change in sub-outcomes of DDS which are emotional burden (z = −3.02; p < 0.003), regimen distress (z = −3.90; p < 0.001), interpersonal distress (z = −3.70; p < 0.001), and physician distress (z = −4.11; p < 0.001) were observed at T3. For self-efficacy findings (z = −3.86; p < 0.001) indicated a difference between T1 and T3. A trend of improvement was observed in self-reported HbA1c levels (z = −2.27; p < 0.02). Moreover, self-reported quality of life (z = 2.48; p < 0.01) showed a potential change from T1 to T3. Findings also revealed reductions in menopausal symptoms (z = −3.94; p < 0.001).

Other assessments were observed in general sleep disturbance (z = −3.10; p < 0.002). Physical activity in minutes per week was a mean of 2670 (1575) and there were observations respectively on weight at T3 (z = −1.77; p < 0.076) and on BMI (z = −1.96; p < 0.049). In addition, the confidence levels of women about their ability to change diet were found mean of 8.22 (p < 0.02) indicating improvement.

Moreover, improvements were observed in their concerns regarding their weight. Indeed, most of the participants (86%) indicated either moderate, little or no concern, and only 14% of participants remained very concerned about weight at T3. The study also found a decreased mean in unhealthy dietary choices (2.43) and an increase in healthy dietary choices (4.43) post-intervention. 68% of participants indicated that they had achieved their individualised goal attainment score overall.

At T1, nearly 12% of participants stated they were current smokers and up the data indicated that 8% of participants were still smokers at T3. Only one participant indicated that they were seriously thinking about quitting smoking within a month after intervention. All scores are presented in detail in Table 4.