Methods: A randomized pilot study of 17 patients with diabetes was completed to assess feasibility and to obtain preliminary effect on clinical outcomes such as A1C, depression, and anxiety. Patients were enrolled if they reported being Latinx, preferred Spanish as the medium of communication in health care settings, were over 21 years, had been diagnosed with T2D, and their most recent A1C was 7 or higher. Participants were randomly assigned to a control or Health Coaching group. The latter received a baseline visit with a Health Coach (HC), up to 14 HC calls, and a mid-point and final in-person visit.
Results: The first 17 pilot participants were 65% female (n=11) and 35.3% male (n=6) with an average age of 51.7 (SD=12.9). While all participants were Hispanic/Latino, 1 participant indicated being African American/Black with the rest identifying as Caucasian. Of the 7 individuals who reported on education, 2 had Bachelor degrees, 3 had High School Diplomas and 2 graduated from technical school. Two thirds of the participants (n=13, 76.5%) indicated that they were employed and earning less than $20,000 per year. Approximately half had a partner (n=8, 47.1%) and 41.2% was single (n=7). There were no significant differences in A1C levels between the groups at baseline. The control group experienced little change in A1C baseline (M=10.40; SD =1.22) to midpoint (M=10.37; SD = 1.18) to endpoint (M=10.14; SD = 1.04. The HC group, however, experienced clinically significant decreases in A1C from baseline (M=10.90; SD=2.02) to midpoint (M=9.42; SD =1.93) to endpoint (M=8.96; SD =2.26).
In paired t-tests examining A1C levels within the HC group, there were significant reductions from baseline (M=10.9; SD =2.02) to endpoint to M=8.96 (SD = 2.26) (t=6.17, p<.001). Mean depression scores dropped from M=14.00 (SD=11.3) at baseline to M=1.50 (SD=1.70) at endpoint, (t=4.34, p<.001, effect size dz= 1.2). Anxiety scores significantly dropped from M=10.00 (SD=4.24) at baseline to M=7.00 (SD=5.65) at endpoint (t=4.22, p<.001, effect size dz= 1.1).
Conclusion: While the number of patients (n = 17) was small, observed differences in our primary outcomes of A1C levels from pre to post intervention were clinically significant and statistically significant in paired t-tests between time points. A language-concordant health coaching program to manage diabetes among limited English proficiency Latinx in the U.S. may improve health outcomes and is ready for large scale RCT testing.