Limits...
Predictors of Cognitive Decline in Older Adult Type 2 Diabetes from the Veterans Affairs Diabetes Trial

View Article: PubMed Central - PubMed

ABSTRACT

Aims: Cognitive decline disproportionately affects older adult type 2 diabetes. We tested whether randomized intensive (INT) glucose-lowering reduces the rate(s) of cognitive decline in adults with advanced type 2 diabetes (mean: age, 60 years; diabetes duration, 11 years) from the Veterans Affairs Diabetes Trial.

Methods: A battery of neuropsychological tests [digit span, digit symbol substitution (DSym), and Trails-making Test-Part B (TMT-B)] was administered at baseline in ~1700 participants and repeated at year 5. Thirty-seven risk factors were evaluated as predictors of cognitive decline in multivariable regression analyses.

Results: The mean age-adjusted DSym or TMT-B declined significantly in all study participants (P < 0.001). Randomized INT glucose-lowering did not significantly alter the rate of cognitive decline. The final model of risk factors associated with 5-year decline in age-adjusted TMT-B included as significant predictors: longer baseline diabetes duration (beta = −0.028; P = 0.0057), lower baseline diastolic blood pressure (BP; beta = 0.028; P = 0.002), and baseline calcium channel blocker medication use (beta = −0.639; P < 0.001). Higher baseline pulse pressure was significantly associated with decline in age-adjusted TMT-B suggesting a role for both higher systolic and lower diastolic BPs. Baseline thiazide diuretic use (beta = −0.549; P = 0.015) was an additional significant predictor of 5-year decline in age-adjusted digit symbol score. Post-baseline systolic BP-lowering was significantly associated (P < 0.001) with decline in TMT-B performance. There was a significant inverse association between post-baseline plasma triglyceride-lowering (P = 0.045) and decline in digit symbol substitution task performance.

Conclusion: A 5-year period of randomized INT glucose-lowering did not significantly reduce the rate of cognitive decline in older-aged adults with type 2 diabetes. Systolic and diastolic BPs as well as plasma triglycerides appeared as modifiable risk factors of the rate of cognitive decline in older adult type 2 diabetes.

No MeSH data available.


Box–whisker plot of the distribution of 5-year change in (A) digit span, (B) Trails making Test-Part B, or (C) Digit Symbol age-adjusted test score in standard vs. intensive glucose-lowering treatment groups. (A–C)P-value(s) from T-test comparing mean scaled score in standard vs. intensive glucose-lowering treatment group. Boxes represent first and third quartiles, horizontal line denotes median value, whiskers represent minimum and maximum values. (A)N = 1174; (B)N = 1155; (C)N = 1114.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5015004&req=5

Figure 1: Box–whisker plot of the distribution of 5-year change in (A) digit span, (B) Trails making Test-Part B, or (C) Digit Symbol age-adjusted test score in standard vs. intensive glucose-lowering treatment groups. (A–C)P-value(s) from T-test comparing mean scaled score in standard vs. intensive glucose-lowering treatment group. Boxes represent first and third quartiles, horizontal line denotes median value, whiskers represent minimum and maximum values. (A)N = 1174; (B)N = 1155; (C)N = 1114.

Mentions: We next computed the mean (and 95% confidence intervals) for the scaled DS, digit symbol, and Trails B test scores obtained at baseline and repeated at the 5-year study visit by a subset of all participants (Table 2). There was no significant difference (−0.003) in the mean scaled DS score (between the baseline and the 5-year interval measurement) in participants who completed the DS at both study intervals (Table 2). The mean scaled digit symbol score declined significantly (−0.92; P < 0.001, Table 2) and the mean scaled TMT-B declined significantly (−0.488; P < 0.001, Table 2) between the baseline and 5-year interval measurements. These results are consistent with Yeung et al. (11) who reported that executive function (Trails B) and processing speed (digit symbol) were affected by type 2 diabetes to a significantly greater extent than was recall. The distribution of change in scaled cognitive test score in all study participants performance by randomized treatment assignment group is illustrated in Figures 1A–C. There were no significant differences in the mean change (over the 5-year period) in scaled DS (Figure 1A, P = 0.83), scaled digit symbol (Figure 1B, P = 0.97) or scaled TMT-B (Figure 1C, P = 0.55) test score between participants randomized to intensive (INT) vs. STD glycemic treatment.


Predictors of Cognitive Decline in Older Adult Type 2 Diabetes from the Veterans Affairs Diabetes Trial
Box–whisker plot of the distribution of 5-year change in (A) digit span, (B) Trails making Test-Part B, or (C) Digit Symbol age-adjusted test score in standard vs. intensive glucose-lowering treatment groups. (A–C)P-value(s) from T-test comparing mean scaled score in standard vs. intensive glucose-lowering treatment group. Boxes represent first and third quartiles, horizontal line denotes median value, whiskers represent minimum and maximum values. (A)N = 1174; (B)N = 1155; (C)N = 1114.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5015004&req=5

Figure 1: Box–whisker plot of the distribution of 5-year change in (A) digit span, (B) Trails making Test-Part B, or (C) Digit Symbol age-adjusted test score in standard vs. intensive glucose-lowering treatment groups. (A–C)P-value(s) from T-test comparing mean scaled score in standard vs. intensive glucose-lowering treatment group. Boxes represent first and third quartiles, horizontal line denotes median value, whiskers represent minimum and maximum values. (A)N = 1174; (B)N = 1155; (C)N = 1114.
Mentions: We next computed the mean (and 95% confidence intervals) for the scaled DS, digit symbol, and Trails B test scores obtained at baseline and repeated at the 5-year study visit by a subset of all participants (Table 2). There was no significant difference (−0.003) in the mean scaled DS score (between the baseline and the 5-year interval measurement) in participants who completed the DS at both study intervals (Table 2). The mean scaled digit symbol score declined significantly (−0.92; P < 0.001, Table 2) and the mean scaled TMT-B declined significantly (−0.488; P < 0.001, Table 2) between the baseline and 5-year interval measurements. These results are consistent with Yeung et al. (11) who reported that executive function (Trails B) and processing speed (digit symbol) were affected by type 2 diabetes to a significantly greater extent than was recall. The distribution of change in scaled cognitive test score in all study participants performance by randomized treatment assignment group is illustrated in Figures 1A–C. There were no significant differences in the mean change (over the 5-year period) in scaled DS (Figure 1A, P = 0.83), scaled digit symbol (Figure 1B, P = 0.97) or scaled TMT-B (Figure 1C, P = 0.55) test score between participants randomized to intensive (INT) vs. STD glycemic treatment.

View Article: PubMed Central - PubMed

ABSTRACT

Aims: Cognitive decline disproportionately affects older adult type 2 diabetes. We tested whether randomized intensive (INT) glucose-lowering reduces the rate(s) of cognitive decline in adults with advanced type 2 diabetes (mean: age, 60&thinsp;years; diabetes duration, 11&thinsp;years) from the Veterans Affairs Diabetes Trial.

Methods: A battery of neuropsychological tests [digit span, digit symbol substitution (DSym), and Trails-making Test-Part B (TMT-B)] was administered at baseline in ~1700 participants and repeated at year 5. Thirty-seven risk factors were evaluated as predictors of cognitive decline in multivariable regression analyses.

Results: The mean age-adjusted DSym or TMT-B declined significantly in all study participants (P&thinsp;&lt;&thinsp;0.001). Randomized INT glucose-lowering did not significantly alter the rate of cognitive decline. The final model of risk factors associated with 5-year decline in age-adjusted TMT-B included as significant predictors: longer baseline diabetes duration (beta&thinsp;=&thinsp;&minus;0.028; P&thinsp;=&thinsp;0.0057), lower baseline diastolic blood pressure (BP; beta&thinsp;=&thinsp;0.028; P&thinsp;=&thinsp;0.002), and baseline calcium channel blocker medication use (beta&thinsp;=&thinsp;&minus;0.639; P&thinsp;&lt;&thinsp;0.001). Higher baseline pulse pressure was significantly associated with decline in age-adjusted TMT-B suggesting a role for both higher systolic and lower diastolic BPs. Baseline thiazide diuretic use (beta&thinsp;=&thinsp;&minus;0.549; P&thinsp;=&thinsp;0.015) was an additional significant predictor of 5-year decline in age-adjusted digit symbol score. Post-baseline systolic BP-lowering was significantly associated (P&thinsp;&lt;&thinsp;0.001) with decline in TMT-B performance. There was a significant inverse association between post-baseline plasma triglyceride-lowering (P&thinsp;=&thinsp;0.045) and decline in digit symbol substitution task performance.

Conclusion: A 5-year period of randomized INT glucose-lowering did not significantly reduce the rate of cognitive decline in older-aged adults with type 2 diabetes. Systolic and diastolic BPs as well as plasma triglycerides appeared as modifiable risk factors of the rate of cognitive decline in older adult type 2 diabetes.

No MeSH data available.