The analysis of self-monitoring of blood glucose (SMBG) in the community-based observational Fremantle Diabetes Study (1) becomes even more interesting when combined with three other studies of SMBG in type 2 diabetes: the Italian Qualità ed Esito in Diabetologia (QuED) Project (2), analyses of the Kaiser Permanente Northern California Medical Care Program (3), and the German ROSSO Study (4).
All four studies concur that patients using SMBG are younger at diagnosis by 3–4 years (1,2,4). Patients present with higher A1C (mean +0.9%) (4). Even during continuous use of SMBG, mean A1C levels are slightly higher (difference 0.2–0.3%) (2–4) or slightly lower (−0.3%) (1) than in patients not using SMBG (mean of all four studies +0.2–0.3%).
Where, then, is the assumed beneficial impact of SMBG on blood glucose control? Though one cannot see it in cross-sectional analyses, it is evident in longitudinal studies. In the ROSSO Study, mean A1C is different at diagnosis between later SMBG users and permanent nonusers by 0.9%. In later years, the initial large difference almost disappeared; i.e., metabolic control improved significantly more in patients using SMBG than in nonusers (4). Similarly, in the Kaiser Permanente cohort, there was an improvement of A1C by ∼0.6% after initiation of SMBG, whereas A1C deteriorated by 0.2% in nonusers. These opposing changes were also observed after adjustments for change of type of antidiabetic medication or other potential confounders (3).
This concordance of observational studies on three different continents is remarkable. In the real world, SMBG appears to be preferentially used by younger patients who exhibit worse than average metabolic control, and the initiation of SMBG is followed by improved metabolic control.