Text-message responsiveness to blood glucose monitoring reminders is associated with HbA1c benefit in teenagers with Type 1 diabetes

Abstract

Aims: To evaluate an 18-month text-messaging intervention in teenagers with Type 1 diabetes and to assess factors associated with text responsiveness and glycaemic benefit.

Methods: Teenagers with diabetes (N = 147), aged 13-17 years, received two-way text reminders at self-selected times to check blood glucose levels and reply with blood glucose results.

Results: At baseline, the participants (48% boys, 78% white, 63% pump-treated) had a mean ± sd age of 14.9 ± 1.3 years, diabetes duration of 7.1 ± 3.9 years and HbA1c concentration of 69 ± 12 mmol/mol (8.5 ± 1.1%). The mean proportion of days with ≥1 blood glucose response declined over time (0-6 months, 60 ± 26% of days, 7-12 months, 53 ± 31% of days, 13-18 months, 43 ± 33% of days). Over 18 months, 49% responded with ≥1 blood glucose result on ≥50% of days (high responders). Regression analysis controlling for baseline HbA1c revealed no significant change in HbA1c from baseline to 18 months in high responders (P = 0.54) compared with a significant HbA1c increase in low responders (+0.3%, P = 0.01). In participants with baseline HbA1c ≥64 mmol/mol (≥8%), high responders were 2.5 times more likely than low responders to have a clinically significant [≥5.5 mmol/mol (≥0.5%)] HbA1c decrease over 18 months (P < 0.05). In participants with baseline HbA1c <64 mmol/mol(<8%), high responders were 5.7 times more likely than low responders to have an 18-month HbA1c <58 mmol/mol (<7.5%; P < 0.05).

Conclusions: Teenagers with Type 1 diabetes who responded to text reminders on ≥50% of days over 18 months experienced clinically significant glycaemic benefit. There remains a need to tailor interventions to maintain teenager engagement and optimize improvements.

Conflict of interest statement

Competing interests

None declared.

© 2019 Diabetes UK.

Figures

FIGURE 1

Decline in text-message responsiveness over time. The mean ± sd percent of days with ≥1 blood glucose (BG) response declined over time, P<0.0001. Month 0–6, n=147; month 7–12, n=142; month 13–18, n=137.

FIGURE 2

Blood glucose (BG) monitoring and HbA1c over time. (a) Mean ± sd frequency of BG monitoring was higher in high responders than in low responders at baseline, 6 and 12 months. (b) Mean ± sd HbA1c was lower in high responders than in low responders at each time point. *P<0.05 for difference between groups, **P<0.01 for difference between groups.

FIGURE 2

Blood glucose (BG) monitoring and HbA1c over time. (a) Mean ± sd frequency of BG monitoring was higher in high responders than in low responders at baseline, 6 and 12 months. (b) Mean ± sd HbA1c was lower in high responders than in low responders at each time point. *P<0.05 for difference between groups, **P<0.01 for difference between groups.

FIGURE 3

Impact of text messaging on HbA1c. In a linear regression model controlling for baseline HbA1c, text-message responsiveness predicted change in HbA1c over 18 months (P=0.03). In low responders, HbA1c increased by 0.3% from baseline to 18 months (P=0.01). In high responders, there was no significant change in HbA1c from baseline to 18 months (P=0.54).