In the course of cognitive deterioration leading to Alzheimer's disease (AD) the increase of amyloid beta (Abeta42) in cerebrospinal fluid or plasma might be an initial event. We previously reported about the associations between concomitant medication and plasma Abeta42 levels in the non-demented population cohort of the Vienna transdanube aging study at baseline. In the present study, the longitudinal influence of insulin, gingko biloba, non-steroidal anti-inflammatory drugs (NSAIDs), oral anti-diabetics (sulfonylurea and biguanides), estrogens, fibrates, and statins on plasma Abeta42 are presented. Associated with medial temporal lobe atrophy (MTA), users of insulin showed significantly increased levels of Abeta42. Long-term users of gingko biloba, independent of their MTA, had significantly decreased plasma Abeta42 and the age-dependent increase of plasma Abeta42 was significantly smaller in long-term gingko biloba treated subjects. The use of fibrates also decreased plasma Abeta42 levels. In multiple testing considering interactions between medications, gender, APOE-epsilon4 presence and creatinine, insulin long-term users again showed significantly increased levels; fibrate and gingko biloba users showed a trend to rather decreased plasma Abeta42 levels compared to the non-users (p=0.05-0.08). Neither statins nor NSAIDs showed a significant effect on plasma Abeta42 in this model. Measuring the effect on cognition, no single medication studied was a significant predictor of conversion to AD or mild cognitive impairment (MCI). Whether the use of gingko biloba might prevent the conversion to MCI or AD needs to be proven in prospective, clinical trials.