Abstract
Objective This study examined the dynamic, real-time associations between partner involvement in diabetes self-care and continuous glucose monitor (CGM) metrics in adults with type 2 diabetes. Methods For 1 week, 63 participants wore Dexcom G4 CGMs and provided momentary reports of partner involvement in diabetes self-care five times per day. Dynamic structural equation models were used to estimate the reciprocal lagged effects of partner involvement on next-hour CGM metrics (and vice versa). Results Partner involvement predicted improved next-hour glucose control for five of six CGM metrics in analyses adjusted for time-varying covariates. The hour after partner involvement, the model predicted a 26.34 mg/dl decrease in glucose level (standardized beta = -0.19), 30% greater odds of meeting target time in target range (beta = 0.07), 48% higher odds of target time below target range (TBR; beta = 0.04; the only nonsignificant effect), 47% greater odds of target time above target range (beta = 0.11), a 4.20 unit decrease in glucose standard deviation (beta = -0.19), and a 0.01 unit decrease in glucose coefficient of variation (beta = -0.08; all p values < .05). There was less consistent support for the reverse pathway, with only two metrics significantly related to next-hour partner involvement: glucose level (beta = 0.15) and TBR (beta = 0.21), such that having higher levels and meeting target TBR were significantly predictive of next-hour partner involvement. Conclusions This is the first study showing that partner involvement in daily diabetes management predicts short-term glucose control. More research is needed to understand how partners influence glycemic control and evaluate interventions that promote their involvement in diabetes care.
