Continuous Glucose Monitoring Plus an Online Peer Support Community Reinforces Healthy Behaviors in Hispanic Adults With Type 2 Diabetes

A qualitative interpretive description design was used to analyze the experiences of Hispanic, Spanish-speaking adults with type 2 diabetes who engaged in a CGM+OPSC intervention.

Participants wore a professional, blinded CGM for 1 week to establish baseline glycemic levels and then concurrently used personal CGM and a Spanish-language OPSC for 12 weeks. Five bilingual, trained peer facilitators with lived diabetes experience and familiarity with CGM and OPSC use moderated the OPSC. Training involved how to incorporate ADCES7 Self-Care Behaviors into the OPSC. The OPSC included three intervention-related posts each week focusing on 1) description of a personal experiment and goal-setting, 2) check-in and troubleshooting the goal, and 3) final check-in on the goal. Participants were encouraged to engage in the OPSC at least three times per week. Peer facilitators and participants could also initiate posts on any topic.

Participants were identified and recruited by bilingual Hispanic community health workers in an urban area of a Mountain West state. A Hispanic, trilingual (Spanish/English/French) pre-medicine research assistant (RA) contacted and obtained consent from participants. Participants were included if they were ≥21 years of age, self-reported as Hispanic, had type 2 diabetes and internet access, and met the following criteria: not using insulin, able to speak and read Spanish, willing to wear a CGM sensor, willing to engage in an OPSC, and willing to avoid vitamin C >500 mg daily and aspirin >325 mg daily (because of interactions with CGM accuracy). Recruits were excluded if they had type 1 diabetes or illness that prevented them from participating (cognitive impairment, two or more hospitalizations in the past year, alcohol abuse or dependence, or life expectancy <6 months), had used CGM in the previous 6 months, or met the following criteria: using insulin, pregnant or planning to become pregnant, or currently enrolled in another diabetes study.

This study was conducted between December 2019 and July 2020. Two significant unforeseen events occurred during the study: the coronavirus disease 2019 (COVID-19) pandemic shutdown on 13 March 2020 and, less than a week later, a 5.7-magnitude earthquake with more than 2,000 aftershocks in the participation area.

The RA conducted the interviews. The first three interviews were conducted with a Hispanic bilingual (Spanish/English) doctorate-prepared nurse practitioner for training purposes. Interviews used a semi-structured interview guide and were conducted in Spanish. Given that participants experienced both the COVID-19 pandemic shutdown and a major earthquake during the study, additional questions asked specifically how the intervention influenced their health during these events. Interviews were conducted via phone and audio-recorded. Recordings were translated and transcribed and then reviewed to verify accuracy.

The research team conducted an initial reading of the dataset to familiarize themselves with the data and note initial impressions. A codebook was developed by author M.L.L. and revised through ongoing discussions among the study team (13). All data were then coded according to the codebook (13,14). Using Microsoft Excel, responses were then extracted, abstracted, and condensed. Emergent themes were identified through inductive thematic analysis (15). Data content—not code frequency—was used to determine saturation (16). All members of the research team reviewed the data and agreed on emergent themes.

Interviews were completed with 26 participants: 16 females and 10 males. The average age was 54.8 ± 9.5 years, with a diabetes duration of 6.1 ± 6.2 years. Most were married (65.4%), uninsured (61.5%), had a high school education or less (57.7%), and had a household income of <$35,000 (46.2%). Half had previously received diabetes education. Three overarching themes emerged: 1) CGM supports understanding of the relationship between glucose levels and health behaviors, 2) the OPSC reinforces how to make healthy choices, and 3) CGM+OPSC supports behavior change and increases confidence. These themes and their subthemes are described below.

The immediate feedback provided by CGM helped participants better understand how healthy eating, being active, taking medication, stress, and sleep behaviors affected glucose levels. This awareness reinforced healthy behaviors as habits and instilled a sense of confidence in diabetes self-management.

CGM helped participants recognize what types, portions, or combination of foods raised their glucose levels. CGM was preferred over blood glucose monitoring (BGM) because of the cost limitations of BGM in this majority uninsured population, coupled with the limited value a single glucose level could provide. One participant stated, “I noticed that when I ate a lot of tortillas, oh, man, my . . . glucose would skyrocket. And I never used to pay attention to that before, because I didn’t want to prick my finger.” Participants also gained a better understanding of how culturally common foods in combination (e.g., beans, rice, and tortillas) raised glucose. Participants also recognized which foods stabilized glucose levels. For example, one participant stated, “I could see that I was eating a lot of starches, or when I’d eat a lot of vegetables. All of that would help me a lot. Sometimes they [glucose levels] would go up if I was eating a lot of tortillas, and when I’d eat vegetables, it would go back to normal.” Once participants recognized how food affected their glucose levels, CGM was used to support meal planning. For example, when participants recognized that their glucose levels were elevated, they modified their behavior at the next meal: “On certain weekends, when I may treat myself with things that aren’t as healthy, I noticed that it goes up too much. Then I say, ‘No, no, enough. If I just ate this [treat] and it went up so much, then when it’s dinnertime, I won’t eat dinner. I’ll just drink a shake and that’s it.’”

Occasionally, participants actively experimented to see how much their glucose would rise from specific foods: “I knew that carbs were bad for me. I knew that sugar would elevate it. But I didn’t know how much. So, when I started using [CGM] I thought, ‘Okay, how much is one tortilla going to make it go up?’ And—boom—I realized that it goes up by 40 points. I mean, that’s a lot.” Participants described other behavior changes as well, such as decreasing their caffeine intake. A few people reported weight loss as a result of healthier eating behaviors.

CGM helped participants recognize how their glucose levels reacted to being active: “When I exercise, I check to see how my body reacts. Sometimes [my glucose level] kind of goes up a little, but then it goes back down. The levels drop.” Some participants reported increasing their physical activity during the study; faced with elevated glucose levels, participants explored what they could control, and physical activity was used as treatment: “Even with the medication, I realized that my sugar was like a rollercoaster. It would go up and down, and very clearly. So, I started to go for walks. Not light walks, but fast-paced ones.” Beyond formally scheduled exercise, participants also increased their physical activity by parking further away at the grocery store or work or riding a bike.

CGM served for some participants as a medication reminder, improving oral diabetes medication-taking behavior: “My blood sugar level fluctuated in a way that I hadn’t seen before. Each time it would indicate that I hadn’t taken my medication, needed to take the medication, or perhaps I needed to exercise a bit more.” In a few instances, participants were working with their prescribing providers on medication dose adjustments; one participant stated, “I would take 500 [mg of metformin], then I would scan again. It wouldn’t go down that much. But when I would take the 1,000 [mg of metformin], I could see the difference.” Another participant who didn’t perform BGM regularly reported that he thought he could tell when his glucose was elevated. CGM helped him realize that he wasn’t always aware of elevated glucose; as a result, he realized that he needed a higher dose of medication. Another participant recognized the need for a second oral medication he was prescribed but wasn’t taking and started taking the second medication.

In response to elevated glucose levels, some individuals incorporated home remedies, often in the form of a tea that included ingredients such as cinnamon, horsetail, and corn silk. In two instances, participants who made enough changes to their eating and activity behaviors recognized that they could use less medication and reduced their dose of metformin on their own accordingly.

Generally, participants underestimated the impact of stress and various emotions on glucose levels. CGM use allowed participants to identify how their glucose levels responded to stressful situations. One participant stated, “What really shocked me was the stress. Even if it was very light, my blood sugar would still go up. It helped me realize. I mean, theoretically, you already know that. But now, you’re able to see it.” Another participant reported, “During the huge snowstorm, my car was skidding the whole way on the road. It took me almost an hour to get to work. When I got there, I checked my blood sugar and it was really high. It was due to the stress.”

The majority of participants expressed exacerbated stress in response to COVID-19 and the lockdown:

Sometimes higher glucose was attributed to stress alone. For example: “I’m usually 90, 95, 126, 130 [mg/dL]. After I eat cake, I’m 212. But this quarantine—it’s up to 359 now.” For others, stress influenced unhealthy behaviors, and CGM helped to identify and support healthy changes. One person stated, “The pandemic stressed me out. It gave me anxiety. And I started eating and eating. I started realizing that the [CGM, during a scan] would tell me, ‘Beep, beep.’ I mean, ‘What’s up? Are you going to keep eating?’ I was stress-eating, and the [CGM] helped me. It was like a judge that would ask me, ‘What are you eating? That’s bad for you.’”

The earthquake and subsequent aftershocks also influenced participant stress. One person detailed feelings of terror: “There were moments where I would fall apart. I would say, ‘We’re going to be destroyed, and this earthquake is going to liquify us.’” Facing unprecedented stressors, whether from the pandemic or the earthquake, participants varied in how they engaged with CGM. A few ignored their CGM for a few days in favor of pressing matters such as family safety in a home damaged by the earthquake, whereas the majority increased CGM scans to understand how their glucose was affected:

Some participants started to recognize how a full night of sleep affected not only energy levels, but also glucose levels. Being mindful of the link between sleep and glucose levels motivated participants to make behavior changes to improve their sleep quality. One participant described, “When I checked myself in the morning, I could obviously see if I slept well, then my blood sugar would be okay. If I didn’t sleep well, then my blood sugar would be higher. Now I stop using the phone before bed so I can get better sleep.” In addition to deceasing phone activity, some participants also described going to bed earlier: “I realized that when I went to sleep earlier, besides the fact that I would wake up more well-rested, my blood sugar level wasn’t so high.”

When reviewing CGM data, some people compared and contrasted glucose levels during sleeping hours and daytime hours: “When you sleep, gosh, your blood sugar doesn’t go up. It stays – it goes down little by little. It regulates when you’re resting.” This helped them better understand how activities during the daytime resulted in more frequent glucose fluctuations.

Although CGM could help participants understand the relationship between glucose and day-to-day activities, the OPSC reinforced the “how” of living with diabetes. Making healthier choices hinged on personal experiments and goal-setting, collective learning, and social support.

Participants engaged in 12 weekly goal-setting activities using personal experiments posted in the OPSC. A discussion followed each personal experiment; people described how they were choosing to interpret the experiment in a way that made sense in their lives. Participants looked forward to the days the personal experiments were posted. Importantly, participants described that, in many cases, if the personal experiment hadn’t been presented to them, they would not have pursued the challenge (e.g., trying a new snack to see how glucose levels responded).

Personal experiments reinforced a sense of positive peer pressure, as one participant illustrated with respect to physical activity: “I also wanted to start walking because some people said that walking helped them. They were even setting goals to do that. Because of that, I motivated myself and also started to walk.” Personal experiments also motivated those who were contemplating a specific healthy behavior but had not yet engaged.

The e-mail digest served as an extra layer of communication for those who didn’t log into the OPSC. Participants looked forward to learning about and engaging in the personal experiments. One participant described:

Participants used the OPSC in different ways, with some mostly reading posts, whereas others fully engaged in responding to peer facilitators and other participants. Generally, participants could not differentiate peer facilitators from other participants. There were isolated cases in which two male participants stated that they personally didn’t need the OPSC; however, they said they recognized that they could see how others needed it.

Participants valued the varied experiences of others, providing a source of collective learning: “Your doctor guides you and everything, but there’s nothing like the . . . people who actually have experience with this, because there are patients that have had diabetes for years.” As a result, the tips and tricks learned from one another were invaluable and, perhaps most importantly, were rooted in Hispanic culture. Ideas on how to make healthy behaviors work in a pandemic were also shared. As a result, one participant who had never done yoga before started a yoga video in Spanish because of the recommendations and encouragement of others.

Recommendations included culturally appropriate recipes and how to incorporate more vegetables into food choices. There were discussions about carbohydrates, added sugar and sugar substitutes, and how to incorporate healthy behaviors into day-to-day life: “People would share the things they did. I used several recipes people said helped them so their blood sugar levels didn’t go up too much. I really liked the tips. They help us become more aware and have more knowledge of what we eat.”

Collective learning extended beyond diabetes and also addressed COVID-19. Participants described learning that COVID-19 complications were more common in people living with diabetes. Participants also described COVID-19–related misinformation as being common. The OPSC was regarded as highly trustworthy for diabetes content, and as a result, participants viewed it as a credible source of COVID-19 information as well:

Participants received emotional and informational social support from their peers through the OPSC. The OPSC provided a safe place to ask questions, express worries, or vent about frustrations and ultimately feel heard. One powerful comment from a participant showed the value of the OPSC: “More than anything, you feel supported. It’s like going—for example, I was going to a psychologist when they murdered my son. [In the OPSC], I felt good because I was able to vent, and they supported me and told me what I could do.” Another participant stated:

The OPSC was uplifting and a source of positivity around diabetes and helped people cope with diabetes. One participant noted, “No one talks about negative things. No one says things that make you feel bad. On the contrary. Everyone is there to . . . lift your spirits, like to tell you, ‘Work hard.’ So that’s why I’m saying I don’t want to stop using it.” As a result, strong bonds were formed within the OPSC, even within the short 12-week period. One participant commented: “The support group, well, you make friends, coincidentally, and sometimes it’s good to hear comforting words from other people, you know? Because a lot of the time you hear bad comments from your own family, and not from people [in the OPSC].” Another person reported, “[The OPSC] helped me tremendously. When I posted comments, sometimes they would reply that I was making the effort for my health. Others would tell me that what I was doing was very difficult, but that I could do it. I have received several comments, and I really liked it, to be honest.”

For many, there was a reciprocal give-and-take nature to the social support, although it came in different ways. Whereas some participants would exchange similar information (e.g., recipes), others had different approaches to providing support within the OPSC. A participant who was 1 year into her diabetes diagnosis noted, “I don’t have that much knowledge or experience. So, more than anything, I was like the person who would welcome them and encourage them. But in terms of teaching them, no, because I was learning myself.”

Support within the OPSC extended beyond diabetes and helped support individuals through unforeseen stressors such as the pandemic and earthquake. One participant noted, “It’s helpful because we express ourselves there. [We discuss] what I feel anxious about, like going back to work [post-pandemic lockdown].” Participants viewed the OPSC as caring for not just their diabetes, but them as a whole being.

In combination, CGM+OPSC was perceived as highly beneficial to participants. Data from the CGM in isolation were helpful, but the combination of the personal experiments, collective learning, and social support from the OPSC magnified the benefit. Participants said the OPSC helped them better decipher CGM data to meld “what” glucose is doing with “how” to influence glucose with healthy behaviors: “It’s ideal because the [CGM] is one thing and the [OPSC] is a tool that helps you use the [CGM].” We were able to identify a positive feedback loop (Figure 1).

During stressful situations, including the 5.7-magnitude earthquake that immediately followed the COVID-19 pandemic shutdown, participants realized as a result of CGM that their glucose levels were rising. They then looked to the OPSC—an established and trusted community—for support. People appreciated having someone respond to their questions quickly. Peer facilitators and peers engaged in conversations to help one another with emotional and informational support, which in turn helped stress and glucose levels.

The positive feedback loop (Figure 1) reinforces confidence in diabetes management. When participants tried personal experiments described in the OPSC and saw a positive effect on glucose levels, it increased their confidence. Participants who tried a personal experiment and saw unwanted glucose effects could go back to the OPSC to receive emotional support and problem-solving techniques. The OPSC provided backup support to prevent potential reductions in participant confidence. Participants thus felt that they could maintain the new healthy behaviors they learned beyond the study period: healthy eating, increasing exercise, taking medication, improving sleep, and focusing on decreasing stress.

Participants described their own ways and offered examples of maintaining healthy eating behaviors, including reducing portion sizes (e.g., of bread or ice cream), omitting certain foods or beverages altogether (e.g., tortillas, tortilla chips, cakes, oranges, soda, or juice), changing food combinations (e.g., avoiding eating tortillas, rice, and beans together), avoiding eating late at night (e.g., no more apples as 1:00 a.m. snacks), and increasing portions of healthy foods (e.g., vegetables).

Healthy eating was the most commonly reported behavior participants said they planned to maintain. Others pledged medication consistency: “One of the changes is to be on top of my medications. I will keep up with my medication at my scheduled times.” Still others had learned how exercise could support stress management:

The CGM+OPSC intervention reinforced healthy behaviors while instilling confidence: “I don’t think [it will be hard to maintain], because all the tips they posted, everything they told us, well, what we all told each other, I do think it helps a lot. You remember what they told you. All that helps.” Another participant stated, “In the amount of time that I’ve had the [CGM], I learned that I have to be conscious. So, no, it’s not going to be difficult for me to know which foods elevated my glucose levels more, and which did not.” One participant reported that not having continued use of CGM and the OPSC would hinder the ability to maintain healthy behaviors.

Participants wanted to continue CGM+OPSC beyond the study. A public version of the OPSC was available to participants after the study period; however, CGM was not because it requires a prescription. Generally, participants experienced limitations following up with their health care provider (HCP), especially because of pandemic limitations during the study. Participants who did have follow-up with their HCP reported that, most often, the HCP did not have previous knowledge of or experience with CGM or OPSC. One person stated, “I told her about the experiment and everything. She told me that they had just received training at the clinics on it and said, “Next time you come, I want you to tell me how it worked for you to see if I would recommend it to my patients.” Once HCPs saw how participants were doing with the intervention, they were supportive. One person stated, “The last time that I had an appointment, I explained that I was participating in a diabetes study and I showed him the sensor. He told me that was perfect, that it was great.” Two participants said their HCP had praised their ability to reduce their A1C levels.

This study included a largely uninsured population. Some participants reported that CGM augmented their ability to manage their diabetes between infrequent HCP visits: “I get a checkup once a year. I only go get a physical, and I try to maintain my glucose levels through my diet. That’s why the use of this machine helped me a great deal, because I was able to see how my glucose levels were doing.”

To our knowledge, this is the first study that addresses CGM+OPSC in Spanish-speaking adults with type 2 diabetes. Our research adds to the limited body of evidence addressing diabetes technology disparities. We found that this diverse population will use diabetes technology that is optimized for their language and social support needs. As a combined intervention, CGM+OPSC provided the tools necessary to learn, problem-solve, and make and reinforce healthy behavior changes, enhancing self-efficacy.

The combination of CGM+OPSC helped participants understand how to apply CGM data to support healthy behaviors. Participants in this study had not previously received education from HCPs on how to use CGM data to change behavior. The CGM+OPSC intervention provided opportunities for discovery and learning, which increased confidence. Consistent with self-efficacy theory, participants’ self-efficacy was rooted in their own experience (personal experiments and goal-setting), vicarious experiences (role modeling from peer facilitators and others within the OPSC), verbal persuasion (collective learning), and emotional arousal (social support) (17). In a systematic review of diabetes technology, Greenwood et al. (18) found that A1C reduction was optimized when there was a technology-enabled self-management feedback loop that connected people with diabetes and their health care team using two-way communication, analysis of patient-generated health data, tailored education, and individualized feedback. In our study, CGM provided patient-generated health data that peer facilitators, in lieu of HCPs, helped participants contextualize, providing general and individualized tips to support behavior change through OPSC communications. The provided emotional support helped participants cope with diabetes, the central component of the ADCES7 Self-Care Behaviors framework (19). Through this two-way communication and enhanced coping, participants reported self-efficacy toward making healthy behavior changes.

This study addressed language, a crucial component for engaging Hispanic Spanish-speaking populations in using diabetes technologies (4). At the time of writing, there was only one Spanish-enabled CGM application approved in the United States. Although a few diabetes OPSCs exist that are in Spanish and culturally tailored to Hispanic populations, evidence on how usage relates to health outcomes is limited. The ADA recommends that diabetes technology should be based on each person’s needs, desires, skill level, and access (5). This study indicates that people desire, benefit from, and have the skills to use CGM+OPSC; however, CGM is not readily available because of both cost and insurance restrictions. Hispanic participants in this study were mostly uninsured and had low levels of discretionary income and thus would have barriers accessing CGM. To address health disparities, technology developers should consider how language and cost barriers further widen the technology disparity gap for racially and ethnically diverse populations.

There were several limitations of our study. Transferability of findings may be limited to other Hispanic, Spanish-speaking populations. Although our sample size was small, it was adequate for qualitative research, as evidenced by our reaching data saturation.

This study also had several strengths. First, the RA and one of the researchers were members of the Hispanic community studied. Second, the peer facilitators had lived experience with diabetes and CGM use and, importantly, were Hispanic. Finally, trust in research teams is crucially important among health-disparate populations, and yet it is often lacking. As in the rest of the country (20), COVID-19 disproportionately affected Hispanic participants in this study. Our research was further complicated by an unexpected earthquake, exacerbating anxiety at a time of heightened stress. The research team’s ability to respond to participant informational and emotional support needs, as reinforced within the OPSC, enhanced trust and kept participants engaged in the study. This trust reinforced sound COVID-19 recommendations and addressed misinformation (21).

CGM+OPSC is a promising intervention to address health disparities seen in Hispanic populations with diabetes. This exciting and innovative area of research requires further exploration in a trial setting to examine clinical efficacy.

The authors thank the Association of Diabetes Care & Education Specialists for donating paraprofessional training and the diabetes nonprofit organization Beyond Type 1 for hosting the OPSC platform. They also acknowledge the intervention team, including Eugenia Araiza, Lorena Drago, Rebeca Peon, Julissa Rulon, Ruis Santos, and Diana Velo, all of whom were contractors with the University of Utah.

This work was supported by an investigator-initiated grant funded by Abbott Diabetes Care.

D.A.G. is currently employed by Dexcom, although she was not an employee of Dexcom at the time of study conceptualization or data collection. No other potential conflicts of interest relevant to this article were reported.

M.L.L., A.N., E.I., and D.A.G. conceptualized the study. M.L.L., A.S.-B., and B.R.-G. collected and analyzed the data. M.L.L., N.A.A., A.S.-B., A.N., and D.A.G. wrote the manuscript. B.R.-G. and E.I. reviewed and edited the manuscript. M.L.L. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.