On behaviors making them at less risk of dementia [46]. BIBS39 Secondly, as may be seen in many studies, Table 1. Baseline characteristics of the three SIS 3 treatment groups.including the present one, cognitive decline is a slow process in elderly non-demented subjects. For this reason, a short study follow-up may be insufficient to assess strategies, either pharmacological or non-pharmacological, that may have a significant but modest impact on cognitive decline. In our study, the treatment benefit associated with EGb761H only became clinically relevant after several years, a longer duration than that involved in the GEM study and the GuidAge study, the two clinical trials which reported no effect of EGb761H on the incidence of dementia. Another reason to believe that the possible effect of EGb761H may be appreciable in the long- rather that short-term relates to the long evolution of Alzheimer’s disease before the dementia stage is attained. Dementia has been shown to be the end stage of a long evolutive process lasting more than a decade. Several long-term prospective studies have now clearly demonstrated differences on cognitive tests in individuals who ultimately developed dementia aVariable Age (years): mean (SD) Gender (women): n ( ) Education: n ( ) No formal education School certificate or higher Depressive symptoms: n ( ) Baseline MMSE: mean (SD) Memory complaints: n ( ) Number of medications: mean (SD)EGb761H (n = 589) 74.8 (6.6) 435(73.9 )Piracetam (n = 149) 75.7 (6.6) 91 (61.1 )Neither (n = 2874) 75.0 (6.9) 1556 (54.1 )p (3-way)*0.329 ,0.0001 0.p (2-way)*0.128 0.002 0.172 (30.6 ) 391 (69.4 ) 60 (10.4 ) 26.3 (2.9) 283 (63.7 ) 4.2 (2.7)44 (31.2 ) 97 (68.8 ) 26 (17.9 ) 25.7 (3.9) 88 (75.2 ) 4.1 (2.7)1050 (38.4 ) 1685 (61.6 ) 388 (13.8 ) 25.7 (3.5) 984 (58.4 ) 4.0 (2.8) 0.023 ,0.001 ,0.001 0.182 0.012 0.040 0.020 0.*Probability values are determined using the x test or by analysis of variance as appropriate. The three-way determinations compared the distribution of variables between all three treatment groups and the two-way determinations between the EGb761H and piracetam groups only. doi:10.1371/journal.pone.0052755.tGinkgo Biloba and Long-Term Cognitive DeclineTable 2. Means and standard deviations for the three cognitive scores at each follow-up visit for the three treatment groups.Mean (SD) Test Mini Mental State Evaluation Group Neither T0 25.7 (3.5) T1 26.7 (3.1) 25.6 (4.6) 27.1 (2.4) 35.2 (5.0) 35.1 (5.5) 35.7 (4.7) 10.9 (2.6) 10.7 (2.8) 10.9 (2.6) T3 26.2 (3.8) 24.9 (5.2) 26.7 (3.5) 34.9 (5.7) 33.3 (6.6) 35.7 (4.9) 10.8 (2.5) 10.4 (2.8) 11.0 (2.6) T5 26.1 (4.3) 24.4 (6.3) 26.5 (3.9) 39.2 (10.4) 35.7 (11.3) 39.2 (9.3) 10.9 (2.6) 10.8 (2.5) 11.1 (2.4) T8 25.7 (5.0) 22.6 (8.4) 26.1 (4.7) 38.3 (11.2) 34.7 (11.2) 37.8 (9.4) 10.5 (2.7) 10.3 (3.2) 10.9 (2.2) T10 24.6 (6.3) 21.1 (9.2) 25.4 (5.2) 37.0 (11.2) 34.9 (11.3) 37.2 (9.2) 10.6 (2.7) 10.3 (2.7) 10.6 (2.7) T13 24.4 (6.3) 22.0 (7.7) 24.3 (5.7) 37.7 (11.2) 33.8 (13.1) 36.2 (9.4) 10.6 (2.6) 9.5 (3.1) 10.5 (2.5) T15 24.1 (6.5) 22.1 (7.6) 24.3 (6.4) 37.4 (11.0) 34.4 (10.0) 35.9 (10.3) 10.7 (2.5) 10.7 (2.4) 10.5 (2.4) T17 24.0 (6.0) 23.3 (5.7) 23.5 (5.8) 36.2 (11.1) 33.0 (11.2) 33.5 (10.7) 10.4 (2.5) 9.0 (3.4) 10.2 (2.4) T20 24.0 (5.7) 23.8 (6.5) 23.7 (5.4) 34.6 (12.0) 34.8 (12.2) 31.4 (10.4) 10.6 (2.7) 8.7 (3.3) 10.2 (2.4)Piracetam 25.7 (3.9) EGb761H Isaacs set test (30 sec) Neither 26.3 (2.9) 34.4 (5.4)Piracetam 34.6 (5.6) EGb761H Benton Visual Retention Neither Test 35.3 (4.8) 10.1 (2.On behaviors making them at less risk of dementia [46]. Secondly, as may be seen in many studies, Table 1. Baseline characteristics of the three treatment groups.including the present one, cognitive decline is a slow process in elderly non-demented subjects. For this reason, a short study follow-up may be insufficient to assess strategies, either pharmacological or non-pharmacological, that may have a significant but modest impact on cognitive decline. In our study, the treatment benefit associated with EGb761H only became clinically relevant after several years, a longer duration than that involved in the GEM study and the GuidAge study, the two clinical trials which reported no effect of EGb761H on the incidence of dementia. Another reason to believe that the possible effect of EGb761H may be appreciable in the long- rather that short-term relates to the long evolution of Alzheimer’s disease before the dementia stage is attained. Dementia has been shown to be the end stage of a long evolutive process lasting more than a decade. Several long-term prospective studies have now clearly demonstrated differences on cognitive tests in individuals who ultimately developed dementia aVariable Age (years): mean (SD) Gender (women): n ( ) Education: n ( ) No formal education School certificate or higher Depressive symptoms: n ( ) Baseline MMSE: mean (SD) Memory complaints: n ( ) Number of medications: mean (SD)EGb761H (n = 589) 74.8 (6.6) 435(73.9 )Piracetam (n = 149) 75.7 (6.6) 91 (61.1 )Neither (n = 2874) 75.0 (6.9) 1556 (54.1 )p (3-way)*0.329 ,0.0001 0.p (2-way)*0.128 0.002 0.172 (30.6 ) 391 (69.4 ) 60 (10.4 ) 26.3 (2.9) 283 (63.7 ) 4.2 (2.7)44 (31.2 ) 97 (68.8 ) 26 (17.9 ) 25.7 (3.9) 88 (75.2 ) 4.1 (2.7)1050 (38.4 ) 1685 (61.6 ) 388 (13.8 ) 25.7 (3.5) 984 (58.4 ) 4.0 (2.8) 0.023 ,0.001 ,0.001 0.182 0.012 0.040 0.020 0.*Probability values are determined using the x test or by analysis of variance as appropriate. The three-way determinations compared the distribution of variables between all three treatment groups and the two-way determinations between the EGb761H and piracetam groups only. doi:10.1371/journal.pone.0052755.tGinkgo Biloba and Long-Term Cognitive DeclineTable 2. Means and standard deviations for the three cognitive scores at each follow-up visit for the three treatment groups.Mean (SD) Test Mini Mental State Evaluation Group Neither T0 25.7 (3.5) T1 26.7 (3.1) 25.6 (4.6) 27.1 (2.4) 35.2 (5.0) 35.1 (5.5) 35.7 (4.7) 10.9 (2.6) 10.7 (2.8) 10.9 (2.6) T3 26.2 (3.8) 24.9 (5.2) 26.7 (3.5) 34.9 (5.7) 33.3 (6.6) 35.7 (4.9) 10.8 (2.5) 10.4 (2.8) 11.0 (2.6) T5 26.1 (4.3) 24.4 (6.3) 26.5 (3.9) 39.2 (10.4) 35.7 (11.3) 39.2 (9.3) 10.9 (2.6) 10.8 (2.5) 11.1 (2.4) T8 25.7 (5.0) 22.6 (8.4) 26.1 (4.7) 38.3 (11.2) 34.7 (11.2) 37.8 (9.4) 10.5 (2.7) 10.3 (3.2) 10.9 (2.2) T10 24.6 (6.3) 21.1 (9.2) 25.4 (5.2) 37.0 (11.2) 34.9 (11.3) 37.2 (9.2) 10.6 (2.7) 10.3 (2.7) 10.6 (2.7) T13 24.4 (6.3) 22.0 (7.7) 24.3 (5.7) 37.7 (11.2) 33.8 (13.1) 36.2 (9.4) 10.6 (2.6) 9.5 (3.1) 10.5 (2.5) T15 24.1 (6.5) 22.1 (7.6) 24.3 (6.4) 37.4 (11.0) 34.4 (10.0) 35.9 (10.3) 10.7 (2.5) 10.7 (2.4) 10.5 (2.4) T17 24.0 (6.0) 23.3 (5.7) 23.5 (5.8) 36.2 (11.1) 33.0 (11.2) 33.5 (10.7) 10.4 (2.5) 9.0 (3.4) 10.2 (2.4) T20 24.0 (5.7) 23.8 (6.5) 23.7 (5.4) 34.6 (12.0) 34.8 (12.2) 31.4 (10.4) 10.6 (2.7) 8.7 (3.3) 10.2 (2.4)Piracetam 25.7 (3.9) EGb761H Isaacs set test (30 sec) Neither 26.3 (2.9) 34.4 (5.4)Piracetam 34.6 (5.6) EGb761H Benton Visual Retention Neither Test 35.3 (4.8) 10.1 (2.