Training to estimate blood glucose and to form associations with initial hunger

Nutrition & Metabolism

Table 2 Estimated versus measured blood glucose at the final laboratory session (week 7).

	N	Estimated blood glucose	Measured blood glucose	Difference (Estimated - Measured)	Estimation error (%)
All Trained	64	84.9 ± 7.8¹	87.2 ± 7.9²	-2.3 ± 4.7³	4.1 ± 3.1 (4.7 ± 3.6)⁴
Hungry Trained⁵	18	78.1 ± 6.7	80.1 ± 6.3	-2.0 ± 2.5³	2.6 ± 1.9 (3.2 ± 2.4)
Not-hungry Trained⁶	46	87.6 ± 6.5	90.0 ± 6.6⁷	-2.4 ± 5.3³	4.8 ± 3.2 (5.4 ± 3.6)
All Controls	72	78.5 ± 11.6	89.8 ± 10.5⁸	-11.3 ± 14.8^9,10	15.4 ± 10.4 (17.1 ± 11.5)
Hungry Controls⁵	42¹¹	75.9 ± 9.8	89.2 ± 10.2⁷	-13.3 ± 11.9^12,13	14.9 ± 9.8 (16.7 ± 11.0)
Not-hungry Controls⁶	30	82.2 ± 12.9	90.6 ± 10.9	-8.4 ± 17.9^14,15	16.1 ± 11.3 (17.8 ± 12.4)

¹Mean ± SD, mg/dL. Subjects stated to be either hungry or not hungry and they estimated their blood glucose at the hospital laboratory before breakfast.
²Measurements performed by hospital autoanalyzer.
³Estimated less measured blood glucose, significant at p < 0.01.
⁴Absolute value of difference between estimated and measured blood glucose and, inside parenthesis, % of measurement.
⁵Subjects who declared feeling hungry at the laboratory investigative session. The agreement limits (mean difference ± 2SD) were -7.0 to +3.0 mg/dL and -41.3 to +18.6 mg/dL in trained and control groups, respectively.
⁶Subjects reporting to be not hungry at the laboratory investigative session. The agreement limits were -12.9 to +8.2 mg/dL and -45.0 to +28.0 mg/dL in trained and control groups, respectively.
⁷p < 0.01 vs trained hungry subjects in the respective column.
⁸p = 0.08, not significant, vs all 64 trained subjects.
⁹F = 10.6, p = 0.0001 on the difference between estimated and measured blood glucose.
¹⁰t-test p = 0.0001 vs all trained subjects.
¹¹p = 0.001 vs number of hungry subjects in the trained group.
¹²F = 24.6, p = 0.0001 on the difference between estimated and measured blood glucose.
¹³t test p = 0.0001 vs trained hungry subjects.
¹⁴F = 11.9, p = 0.0001 on the difference between estimated and measured blood glucose.
¹⁵t test p = 0.07 vs trained not-hungry subjects.

ISSN: 1743-7075