Exposure to and engagement with digital psychoeducational content and community related to maternal mental health by perinatal persons and mothers: design of an online survey with optional follow-up and participant characteristics

Overview

We recruited adults exposed to or who engage with Momwell via social media or other digital platforms to complete an online survey. We invited participants who provided permission for us to re-contact them to complete a second survey 2-3 months later. The University of Connecticut Institutional Review Board (IRB) approved this study.

Momwell

Founded and led by Erica Djossa, a registered psychotherapist based in Ontario, Canada, Momwell offers a mom-centered model of care that seeks to educate, empower, and support mothers (momwell.com). Momwell offers individual telehealth psychotherapy to clients in Canada and the United States and self-paced virtual workshops and courses led by Ms. Djossa and her team of licensed psychotherapists. In addition to these for-fee services, Momwell provides evidence-based psychoeducational content about motherhood, parenting, and maternal mental health to the public without cost through their social media feeds (Instagram, Facebook, and TikTok), weekly podcast, and blog with posts corresponding to the podcast episodes. A weekly email newsletter alerts readers to new podcast episodes and blog posts, and provides other updates including new workshop/course offerings. In the current study, we focused on exposure to and engagement with Momwell psychoeducational content and community via their social media feeds, podcast, blog, and email newsletter (i.e., resources freely available to the public).

Recruitment and Eligibility Screening

In July-September 2023, we recruited adults who follow Momwell on any of their social media platforms, listen to their podcast, read their blog, and/or receive their email newsletter. The Momwell team posted recruitment messages that included a link to an eligibility screener via Qualtrics. Recruitment messages were posted as public social media posts and Instagram Stories. On Instagram, interested individuals could comment “STUDY” to be sent a link to the eligibility screener via direct message. An ad was included in a Momwell podcast in mid-July 2023. Additionally, a message about the study was included in Momwell’s weekly email newsletter. Individuals were recruited to participate in two cohorts (Cohort 1 and Cohort 2); participants in Cohort 2 completed shorter surveys at both timepoints. Inclusion criteria for both cohorts included age 18 years or older, follows or engages with Momwell on at least one of their digital platforms (e.g., Instagram, Facebook, Facebook group, TikTok, podcast, blog, and/or email newsletters), comfortable participating in the study in English, and able and willing to provide informed consent. To participate in Cohort 1, individuals additionally had to be (1) a perinatal person (e.g., currently pregnant and/or within 12 months post-partum) and/or mother (e.g., identifies as mother of at least one child under the age of 18 who lives with them at least part-time), and (2) currently living in Canada or the United States. As part of the eligibility screener, respondents indicated whether they had a preference of completing the longer or shorter survey (or had no preference); participants who expressed preference for the shorter survey were all invited to complete the shorter survey.

As some recruitment messages were posted publicly, we recognized the potential for attracting the attention of bots or individuals trying to participate under false pretenses in order to receive the gift card incentive [21–24]. As recommended [23,24], we employed a multi-faceted strategy to ensure that enrolled participants were truly eligible, including bot/fraud detection features of Qualtrics, duplicate respondents (e.g., email, IP address), and checking consistency of information (e.g., specific recruitment link clicked on versus where respondent reported they heard about the study). Individuals whose responses to the eligibility screener suggested that they were a bot or fraudulent individual were not invited to complete the survey.

Data collection

Eligible individuals were emailed invitations to complete an online survey via Qualtrics. Individuals who had not responded to the invitation were sent a reminder one day later. Non-responders were sent two additional invitations a few weeks later. Participants who partially completed the survey were sent a reminder email with a new survey link. Prior to completing the survey, participants were shown an Information Sheet (specific for Cohort 1 and Cohort 2) and provided an opportunity to download a PDF of the Information Sheet. Participants consented to the research study electronically, and were then directed to the survey. The survey included both quantitative and open-ended questions. The survey for Cohort 1 participants was designed to take 30-45 minutes to complete (median [IQR]: 52 [36–95] minutes among those who completed at least 80% of the survey). Cohort 1 participants who completed at least 80% of the survey received a gift card ($15 USD or $20 CAD). The survey for Cohort 2 participants was a shortened version of the Cohort 1 survey, comprised of a subset of questions and scales. It was designed to take 15-20 minutes to complete (median [IQR]: 29 [21–45] minutes among completers). Cohort 2 participants who completed at least 80% of the survey were entered into a gift card lottery. One in every 30 participants were randomly selected to receive a gift card ($60 USD or $80 CAD).

We reviewed participants’ responses and survey meta-data to detect likely fraudulent respondents that were not detected at eligibility screening (e.g., discrepancies between meta-data and survey responses, discrepancies between data provided in eligibility screener and survey). As recommended,[23,24] we conducted additional data quality checks at this phase of data collection; we reviewed surveys for patterns of responses indicating inattentiveness or fraudulent responses (e.g., straight-lined psychological measures with reversed items; failed an attention check item; missing, extremely brief, or nonsensical responses to open-ended questions). We contacted participants with multiple flags via phone and/or email to clarify their responses or ask them to provide responses to the open-ended survey questions. Individuals we could not reach or who provided conflicting responses were excluded.

The final question in the survey asked participants for permission to contact them about future research studies, specifically a second survey in about 2 months. Participants who responded affirmatively were invited to complete an optional second survey 2-3 months later (October-December 2023). Study procedures (e.g., survey invitations, reminders, consent process) were the same as for the main survey. Survey response review was also similar to the main survey; we additionally compared consistency of responses in the two surveys (e.g., pregnancy/post-partum status of participants who were pregnant when they completed the main survey). Similar to the main survey, the second surveys were designed to take 30-45 minutes (Cohort 1; median [IQR] duration: 34 [24–62] minutes among those who completed at least 80% of the survey) and 15-20 minutes (Cohort 2; median [IQR] duration: 12 [9–22] minutes among those who completed at least 80% of the survey). As with the main survey, Cohort 1 participants received a gift card after completing the survey ($15 USD or $20 CAD), and Cohort 2 participants who completed the survey were entered into a gift card lottery in which 1 in 30 participants were randomly s elected to receive a gift card ($60 USD or $80 CAD).

To ensure that participants had opportunity to be exposed to and engage with content related to perinatal mental health, the Momwell content development team created and posted 3 psychoeducational posts specifically related to perinatal mental health per week for 16 weeks starting at the start of recruitment for the main survey.

Measures

The surveys included a rich set of validated psychological measures, study-specific quantitative questions, and open-ended questions that assessed exposure to and engagement with Momwell psychoeducation content and community, maternal mental health, parenting relationships, parenting self-efficacy, and other participant characteristics and behaviors. Measures are described below and in Table 1.

Exposure to and engagement with Momwell psychoeducational content and community

As part of the eligibility screener, main survey, and second survey, participants were asked a variety of questions about their exposure to and engagement with Momwell content and community to capture different dimensions of exposure and engagement including duration, frequency, and depth of active engagement (Table 1).

In the eligibility screener, participants were asked whether they listen to the Momwell podcast, follow Momwell on Instagram, Facebook, or TikTok, are a member of the “Momwell Community” Facebook group, follow the Momwell blog, or receive the Momwell email newsletters. From this we categorized participants as social media followers (i.e., follows Momwell on Instagram, Facebook, or TikTok, or is a member of the Momwell Facebook group) or not. We also categorized whether participants listen to the Momwell podcast and/or read the Momwell blog, as both these platforms provide a deeper dive into psychoeducational topics, and blog posts are published that correspond to each podcast episode.

Those reporting exposure to content on each platform were asked how recently they started consuming content from that platform (within the past 7 days, at least 7 days ago but less than 3 months ago, at least 3 months ago but less than 6 months ago, at least 6 months ago but less than 12 months ago, and at least 12 months ago). As Momwell rebranded themselves as “Momwell” from “Happy as a Mother” in January 2023, we prompted participants to include involvement with “Happy as a Mother”. To capture duration of exposure to Momwell content, we calculated the longest duration of exposure on any platform (i.e., how long ago they joined via their first digital platform). We also calculated how recently they joined a new platform.

For each platform, participants reported how often they engaged in specific ways applicable to that platform within the past 4 weeks (or since they started following Momwell on that platform if more recently than 4 weeks ago) on a 4-item Likert scale (“every day”, “3+ times per week but not every day”, “1-2 times”, and “not at all”). Activities included passive exposure to content (e.g., read posts, watch videos, listen to podcast, read blog) and more active and visible forms of engagement (e.g., liking/reacting to posts, replying to post).

Participants who completed the second survey who did not report exposure to Momwell content on a particular platform at eligibility screening were asked about exposure to Momwell content on that platform during the second survey. Then, existing and new users of each platform were asked the same questions about their exposure to and engagement with Momwell content and community since they completed the main survey, with the recall period of since they completed the first survey (e.g., in [MONTH] 2023, e.g., August 2023).

On the main survey, participants were asked several questions related to their perceived impact of Momwell content on their lives. First, they were asked “what Momwell social media post, blog post, or podcast has had the biggest impact on your life? How did it impact you?”.

On the second survey, they were asked similar open-ended questions about the perceived impact of Momwell content, but specific to each platform. In both surveys, participants were also asked whether they had taken several actions as a result of their exposure to Momwell content (e.g., talked with a doctor or health care professional about something you heard in the Momwell podcast or saw on a social media post). In the main survey, participants reported the extent to which they agreed with several statements about changes since joining the Momwell community (e.g., I feel more aware of the signs and symptoms of mental health conditions, I feel more confident in my approach to parenting). In the main survey, participants were also asked what would help them engage more with the Momwell community and what topics or content they would like to see more of from Momwell.

While our primary interest was in exposure to and engagement with the free psychoeducational content and peer community provided by Momwell through their social media feeds, podcast, and blog, we also asked participants about paid services available from Momwell. Specifically, we asked participants who had joined the Momwell community more than 7 days ago whether they had ever purchased a guided journal or other self-paced tool from Momwell, enrolled in a Momwell workshop or course, or enrolled in therapy (in-person or remote) with a Momwell therapist. From responses, we calculated the proportion who had purchased any paid services.

Mental health

Our main measures of maternal mental health are depression, anxiety, and perceived stress. Participants in both cohorts were asked to complete these measures in the main and second surveys (Table 1). Depressive symptoms were assessed using the 8-item Patient Health Questionnaire (PHQ-8) [25,26]. The scale asks participants to indicate on a 4-point scale how often they have been bothered by various depressive symptoms in the past 2 weeks. Response options include “Not at all”, “Several days”, “More than half the days”, and “Nearly every day”. Responses are summed to create a total score ranging from 0 to 24 with a greater score indicating greater reported depressive symptoms. Internal consistency was acceptable, with a Cronbach’s α = 0.85 in the main survey and α = 0.86 in the second compared to the original measure (α = 0.86-0.89 across samples) [25].

Symptoms of anxiety were assessed using the Generalized Anxiety Disorder 7-item scale (GAD-7) [27]. Participants were asked to report how often they had been bothered by symptoms of anxiety over the past 2 weeks. Response options include “Not at all”, “Several days”, “More than half the days”, and “Nearly every day”. Responses were summed to create a total score ranging from 0 to 21 with a greater score indicating higher symptoms of anxiety. Internal consistency was acceptable, with a Cronbach’s α = 0.91 in the main survey and α = 0.90 in the second compared to the original measure (α = 0.92) [27]. Participants also reported whether they had ever been diagnosed with depression or anxiety during pregnancy, the post-partum period, or a non-perinatal period of life. Participants reporting a history of depression (or anxiety) were asked if they currently had depression (or anxiety).

Participants who scored 10 or higher on the PHQ8 or GAD7 or who endorsed feeling as if they were currently experiencing depression or anxiety were flagged and shown a message noting that they might be experiencing some feelings of depression (or anxiety, or depression and/or anxiety, depending on which was flagged) and encouraging them to connect with a mental health care professional for assessment, support, and therapy if appropriate, and that their primary care provider or obstetrician/gynecologist may also be a good resource for support or to help you connect with a mental health care professional. Participants were then offered the option of downloading a PDF mental health resource guide developed by the research team (options for participants from the US, Canada, and other countries). Participants were asked if they wanted a copy of the guide sent to them via email, and research staff emailed a copy of the appropriate mental health guide to participants answering affirmatively.

The 10-item Perceived Stress Scale (PSS-10)[28] was used to assess perceived stress during the past 2 weeks. Participants selected how often they experienced the feelings and thoughts presented in the items via response options ranging from “Never” to “Very often”. Responses are summed to create a total score with a range from 0 to 40 with a higher score indicating higher perceived stress. Internal consistency was acceptable, with a Cronbach’s α = 0.88 in the main survey and α = 0.88 in the second compared to the original measure (α = 0.78) [28].

Parenting support and self-efficacy

Participants in Cohort 1 completed a set of measures related to parenting support and self-efficacy (Table 1). We first asked participants with children whether they had a parenting partner (i.e., a partner who is assisting the participant with parenting, with a note that for many people, their parenting partner is their spouse or the child’s other parent).

Participants who reported a parenting partner completed a modified version of the Postpartum Partner Support Scale (PPSS) [29], a validated 20-item measure that assesses support provided by their husbands or partner. To better represent a variety of partnerships and family structures, we modified the PPSS to replace “him” (i.e., husband) with “my partner”. We also modified questions to refer to “our children” rather than “our baby”. Responses are summed to create a total score with a range from 20 to 80 with higher scores indicating higher parenting support. Internal consistency was acceptable, with a Cronbach’s α = 0.95 in the main survey and α = 0.96 in the second compared to the original measure (α = 0.96) [29].

Participants who reported a parenting partner completed a modified version of the 14-item Brief Co-Parenting Relationship Scale (BCRS) [30] to assess interparental relationships. We modified the phrasing of items and measure instructions to refer to “child(ren)”versus “child” to better reflect families of various sizes. Responses are averaged, with scores ranging from 0-6, with higher scores indicating more positive relationship with parenting partner. Internal consistency was acceptable, with a Cronbach’s α = 0.90 in the main survey and α = 0.91 in the second survey compared to the original measure (α = 0.81-0.89 across different samples) [30].

Perceived parenting competence was assessed using a modified version of the 16-item Parental Sense of Competence (PSOC) scale [31]. We modified the language of the PSOC to replace the words “mother” and “father” in items with “parent” as well as to replace “child” with “children”. The PSOC is composed of two subscales, efficacy and satisfaction, and combined creates an overall score representing parenting self-esteem. The overall PSOC score is reflective of 16-items with a range from 16 to 96 and higher scores indicate greater sense of overall parenting self-esteem. The satisfaction subscale is comprised of 9 items with scores ranging from 9 to 54. Higher scores indicate higher levels of parenting satisfaction. The efficacy subscale is assessed with 7 items, scores ranging from 7 to 42, and higher scores indicate more parenting efficacy. Internal consistency was acceptable for the overall score and both subscales, with a Cronbach’s α = 0.80 for the overall score, α = 0.77 for the satisfaction subscale, and α = 0.76 for the efficacy subscale in the main survey and α = 0.83 for the overall score, α = 0.77 for the satisfaction subscale, and α = 0.83 for the efficacy subscale in the second survey compared to the original measure (α = 0.79 for the overall score, α = 0.76 for satisfaction subscale, and α = 0.76 for efficacy subscale) [31].

Additional psychological measures

We included several additional psychological measures to provide data for secondary analyses of study data (Table 1). As part of the main survey, participants in both cohorts completed the Ten-Item Personality Inventory (TIPI)[32] which assesses the personality traits extraversion, agreeableness, conscientiousness, emotional stability, and openness to experience. Participants in Cohort 1 completed several additional measures as part of the main and second surveys (Table 1). Participants completed a modified General Help-Seeking Questionnaire (GHSQ) [33,34] which asked individuals to rate how likely they would be to seek help from various sources (e.g., partner or spouse, mental health professional) if they were experiencing sadness, low mood, irritability, and other depressive symptoms. Knowledge of mental health conditions and mental health literacy were assessed using a subset of questions from the Mental Health Literacy Scale (MHLS) [35]. Given the interests of the current study (perinatal mental health versus general mental health/illness), we included items assessing knowledge of generalized anxiety disorder (GAD) and major depressive disorder (MDD), the connection between sleep and mood, information-seeking related to perinatal depression, and general attitudes towards people with perinatal depression (stigma).

Participants completed the Iowa-Netherlands Comparison Orientation Measure (INCOM) [37–39], designed to assess frequency of social comparisons. We modified the measure stem to ask about comparisons with “other mothers” rather than “other people” to assess tendency to make social comparisons to other mothers. Participants completed the short form version of the Self-Compassion Scale (SCS-SF) [40,41], a validated questionnaire measuring self-compassion via 6 subscales. Feelings of loneliness were ascertained using the short-form Revised University of California at Los Angeles (UCLA) Loneliness Scale (ULS-8) [42,43]. Finally, participants completed the Brief COPE [36], which assesses 14 coping strategies: self-distraction, active coping, denial, substance use, use of emotional support, use of instrumental support, behavioral disengagement, venting, positive reframing, planning, humor, acceptance, religion, and self-blame.

General social media use and digital health literacy

Participants reported their use of social media generally, including whether they have accounts on different platforms (i.e., Instagram, Facebook, TikTok), how often they engage with various platforms in different ways (e.g., post a video), and whether they follow social media accounts related to motherhood or parenting, mental health, or another health topic. Participants were also asked whether they subscribe to or regularly listen to podcasts (other than Momwell). We modified two items from the e-Health Literacy Scale (eHEALS) [44] to assess perceived skill and confidence using social media to access health information: “I have the skills I need to evaluate health information I see on social media.” and “I feel confident in using health information from social media to make health decisions for me or my family.” Response options ranged from “Strongly Disagree (1)” to “Strongly Agree (5)”. As a brief screener of health literacy, participants were asked how confident they are filling out medical forms by themselves [45].

Demographic characteristics, health behaviors, and health care utilization

Participants reported demographic characteristics, including country and state/province of residence, urbanicity of their city/town of residence, gender, sex assigned at birth, and sexual orientation [46]. Race and ethnicity were assessed via an open-ended question: “How would you describe your racial and ethnic background?” Participants living in the US and Canada were also asked to indicate their racial and ethnic background by choosing all that apply from a standardized list respective to each country. Participants from the US were also asked whether they consider themselves Hispanic or Latina/o (yes or no). Participants also reported whether they self-identified as someone who is a visible minority [47]. Participants also reported education, employment status, family structure, the number of children living in the home at least part-time, and the ages of such children. In the second survey, participants were asked about current pregnancy (including gestational age), if they have given birth to a baby in the past 12 months and if yes, how many months ago they gave birth. Financial strain was assessed by asking participants to identify how difficult it has been for them, in the past 30 days, to pay for usual household expenses (e.g., food, rent, car payments, medical expenses) [48]. Participants reported their current employment status by selecting one of the following options: working full-time, working part-time, working but currently on parental leave, stay-at-home parent or homemaker, unemployed, student, and other. Given the small number of participants who endorsed each of these categories, we collapsed unemployed, student, and other employment status.

Participants completed the Hunger Vital Sign (HVS) [49], a validated, 2-item food insecurity screening tool based on the U.S Household Food Security Survey Module. Participants reported if, in the last 12 months, they had ever worried about food running out due to financial hardship, and if there was ever not enough money to get more food when food did run out. Participants who answered either of the two items as “often true” or “sometimes true” (vs “never”) were considered to be at risk of experiencing food insecurity.[49] Participants who completed the second survey reported employment status, food security, financial strain, and changes to household structure occurring since completing the main survey. To put any changes in mental health in context, as part of the second survey we asked participants to reported if they had experienced various life events since the first survey (e.g., spouse or partner died, got divorced or separated, became pregnant, gave birth or adopted a child). Life events were adapted from the Holmes-Rahe Life Stress Inventory[50]; we added “gave birth or adopted a child”. For participants who reported one or more potentially stressful life event, they were asked to rate how stressful that time was for them on a Likert scale from “not at all stressful” to “extremely stressful”.

Participants were asked questions about their health, health behaviors, and health care utilization. Participants self-reported their height and weight and whether they have any physical or mental health conditions requiring frequent medical visits (and if so, what conditions). Participants answered questions regarding diet, alcohol consumption, tobacco use, physical activity, and sleep. To assess diet quality, participants completed the 8-item Starting The Conversation (STC) dietary screener [51]. Participants reported how many days per week they consumed at least one drink of any alcoholic beverage, number of drinks per day they consumed alcohol, and frequency of binge drinking. Participants also reported ever use and current use of cigarettes and e-cigarettes. Participants reported frequency and duration of moderate or greater intensity physical activity in the past 4 weeks. To assess quality of sleep in the last 4 weeks, participants completed a subset of questions from the Pittsburgh Sleep Quality Index (PSQI) [52] relating to sleep duration and self-rated sleep quality. Participants first reported via free response how many hours of sleep they get at night. Participants were then asked to rate their quality of sleep with possible answer options of “Very bad”, “Fairly bad”, “Fairly good”, and “Very good”. Mothers were also asked how often in the past 4 weeks a child had disrupted their sleep. Participants reported any physical or mental health care services accessed via in-person visits, telehealth visits, phone call, email, or via online patient portal in the past 6 months. In the second survey, participants were asked about new medical diagnoses, health care utilization, and sleep.

Statistical Analysis

The eligibility screener and surveys were administered in Qualtrics (Qualtrics LLC, Provo, UT). We used Research Electronic Data Capture (REDCap) for participant tracking [53]. Analyses were conducted in SAS 9.4 (SAS Institute, Inc, Cary, NC). We described recruitment and enrollment and then characteristics of participants in our final sample, overall and by study cohort. We compared participant characteristics and exposure to and engagement with Momwell content on different platforms in relation to whether participants completed the second survey using chi-squared tests for categorical variables, Wilcoxon rank sum tests for poor physical and mental health days, and t-tests for other continuous variables.

Principal Results

The overall goal of this study is to understand how mothers and the perinatal population consume and engage with digital psychoeducation content and the peer community and how it impacts their mental health and parenting. In this paper, we described the study design including recruitment procedures, characteristics of the sample, and their exposure to maternal mental health psychoeducational content on different digital platforms.

While we recruited a cohort of nearly 600 perinatal persons and mothers, we encountered challenges during recruitment, particularly related to fraudulent and low-quality responses. Bots and fraudulent respondents are becoming more and more of a problem when recruiting participants online, particularly when using public recruitment links and recruiting for studies that provide participant with a monetary incentive for completing an online survey [21–24]. As some of our recruitment messages were posted publicly, we recognized the potential for attracting the attention of bots or individuals trying to participate under false pretenses – and indeed, we experienced significant interest in our study by suspicious actors (human or bot), with several hundred respondents with duplicate email addresses, more than 1,800 with nonsense email addresses, and over 2,000 with conflicting information in terms of location, digital platform engagement, or recruitment source. Bot and fraud detection features of survey administration platforms such as Qualtrics and REDCap (Research Electronic Data Capture) provide a starting point, but researchers must go further to verify the eligibility and quality of respondents. As recommended [23,24], we employed a multi-faceted, multi-stage strategy. We separated eligibility screening from the survey, and only sent individual survey invitations to individuals who passed our reviews. We compared information collected by multiple methods (e.g., recruitment source reported by participant vs recorded by Qualtrics) and consistency of information provided by participants (e.g., ages of children). Our strategy for identifying low-quality or inattentive respondents included an attention check question, reviewing responses to open-ended survey questions, and checking for straight-lining when completing psychological measures [23,24]. We also phrased eligibility questions to reduce the chance that respondents could guess the “right” (i.e., eligible) response. We followed up with participants with inconsistent information on the survey via phone or email to clarify information. While it is possible that we excluded some eligible individuals, our quality control procedures were designed to prevent bots or fraudulent participants from participating in our study. Researchers who are recruiting samples online using public survey links are encouraged to develop and implement similar procedures for preventing and detecting bot and fraudulent respondents. For additional discussion and specific recommendations, researchers are encouraged to read Walker (2023) and Wang (2023), especially their Table 1 and Tables 2 and 4, respectively [23,24].

Another recruitment challenge related to which members of the Momwell community volunteered to participate in this study. We had originally intended to invite only new followers/subscribers (i.e., started following Momwell in the past 7 days) to complete the optional second survey to explore engagement and concurrent changes in mental health and parenting self-efficacy soon after connecting with Momwell content and community. However, only 42 eligible respondents reported joining Momwell within the past 7 days (of whom 15 completed the main survey) – perhaps a result of algorithmic biases in which followers were likely to see our recruitment messages [54,55], and which followers were interested in sharing their experiences with us. Therefore, we pivoted our design to invite any participants to complete the second survey.

Our sample was more likely to have Bachelor’s degree or higher education (87%) than birthing persons in the United States (36%) [56] and women aged 25-34 years born in Canada (40%) [57]. Thirteen percent of our sample identified as a visible minority, compared to 31% of women 25-64 years in Canada [58]; 48% of birthing persons in the United States identify as a race or ethnicity other than non-Hispanic white [56]. In our sample, food insecurity was reported by 11%, which is similar to food insecurity among households with children and married couples in the United States (11%) [59], but lower than the proportion among women aged 25-44 years in Canada (20%) [60]. Whether this is a function of the demographics of the Momwell community or due to which Momwell followers volunteered to participate in this study is unknown.

In the current study, 18% of participants had elevated depressive symptoms and 24% had elevated anxiety symptoms based on their scores on the PHQ8 and GAD7, respectively. These prevalences are similar to estimated prevalences globally and specifically among perinatal persons and mothers in Canada and the United States. The prevalence of depression varies across populations and whether measured by self-report measures of symptoms (as in the current study) or clinical interviews, but estimates are in the 13-28% for perinatal or post-partum persons worldwide [61–63]. Another systematic review and meta-analysis found a prevalence of depression of 16% among healthy mothers without a previous diagnosis of depression from North America [64]. Meta-analyses have estimated the global prevalence of perinatal anxiety to be 21% [65] or 15% during pregnancy and 15% during the first 6 months post-partum [66]. These authors also noted variability across populations sampled and measure of anxiety [65]. The findings of these studies – along with the findings of the current study – that between 1 in 5 perinatal persons or mothers of young children are experiencing elevated symptoms of depression or anxiety highlights the importance of this area of research.