There is substantial research on the effects of physical exercise on cognitive functions. However, less attention has been paid on the requirements of training intensity and length to enhance cognitive abilities in the elderly. To the investigators knowledge no studies have evaluated the effects of extensive endurance exercise training on cognitive functions by studying elderly marathon runners and bicyclists. On the basis of the scientific literature published so far it is not known whether the beneficial impact of endurance exercise training depends on the intensity of training.

The investigators therefore designed a cohort study with adequate power in order to evaluate the effects of intensive endurance exercise training on cognition. This trial, an Austrian prospective cohort study in cognitive function of elderly marathon-runners (APSOEM) is being conducted and will compare neuropsychological performance outcomes of elderly marathon runners or bicyclists with controls matched concerning age, education years, occupation, and verbal intelligence.

Design of the trial The data reported in this study are derived from the cognitive testing and determination of humoral growth factors at the time of the survey 1 (recruitment phase). To determine whether extensive endurance training enhances cognitive abilities in the elderly, we plan to compare longitudinally the outcomes of various neuropsychological tests and humoral markers in elderly marathon runners or bicyclists with controls matched for age, sex, and years of education at the time of the survey 2 (5-year follow-up) and at the time of the survey 3 (10-year follow-up). Runners participating in the 2008 Wachau half marathon (21,2 km) and the Vienna City marathon (42,5 km) as well as bicyclists participating at the Corinthian marathon (180km) were recruited with the assistance of the organizers and via personal contacts. Runners or bicyclists were eligible for inclusion in the study if (a) they had participated in at least one of these 3 marathons in the preceding two years, (b) were still in continuous training during the recruitment phase (at least 2 hours/week) and (c) were over the age of 60. Exclusion criteria were (a) present or past exposure to neurotoxic substances (b) if they did not speak German as their native language (verbal intelligence) (c) diseases that markedly affect CNS functions: for example, cerebrovascular stroke, brain tumor, depression, Alzheimer's disease, epilepsy, multiple sclerosis, Parkinson's disease, etc. (d) manifest cardiovascular disease, (e) chronic alcoholism (daily alcohol intake > 60 g or diagnosed history of alcoholism) and (f) unwillingness to give informed consent. Controls were subsequently contacted via personal contact and three additional advertisements (two in an Austrian newspaper ("Krone") and one in an Austrian bicyclist journal ("Bicyclist Sports"). The controls were matched according to age, sex and years of education. Detailed information about education, smoking habits, family history, exercise training and medication was obtained. All participants had an extensive medical evaluation carried out by an experienced internal specialist including blood withdrawal when entering the study. Since it has been shown that humoral growth factor levels can vary with the circadian rhythm , all examinations were started between 10 am and 10.30 am. In cases where the examination revealed neurological abnormalities, participants were additionally examined by a neurologist. Besides the physical examination with blood withdrawal described above, the study protocol included an ergometry, neuropsychological testing, the use of a set of questionnaires and magnetic resonance imaging. All participants underwent the study protocol tests in the above-mentioned same chronological order.

The study was approved by the ethics committee of the medical faculty of the University of Vienna (number EK 401/2005). All subjects gave written informed consent before entering the study. Procedures followed were in accordance with institutional guidelines.

Ergometry Individual working capacity was calculated as a percentage of the predicted (=100% work load) Watt value (derived from the tabulation, standardized for sex, age, and body surface [28]). Briefly, the workload was increased every two minutes in steps of 25 W, beginning with 25 W and going on until the point of exhaustion (Ergoline, Ergometrics 900). The individual physical working capacity (PWC) was expressed as the individual maximal power (Watt)max in percent of a reference value (Wattref): PWCind = 100 x Wattmax/Wattref [28].

Neuropsychological testing and questionnaires The Vienna Neuropsychological Test Battery (VNTB)as well as the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) were selected in order to assess cognitive functions, such as visuo-construction, concentration/attention, language, memory and executive functioning domains. These cognitive abilities are known to be commonly affected by Alzheimer's disease and other dementias. The two batteries of tests have been found to be sensitive in the evaluation of mild cognitive impairment.

Each test run began with a screening of global cognitive functions via the Mini Mental State Examination (MMSE) and the Clock Drawing Test. Thereafter participants were subjected to a cognitive testing assessing visuo-constructional abilities, attention, language functions, memory and executive functions. Details of cognitive testing are described in the Appendix.

After finishing the test batteries, subjects were asked to fill out several self-rating scales and forms to assess premorbid intelligence levels, subjective memory functions, psychological and physiological well-being , depression, and activities of daily life.

Laboratory procedures Blood counts, all clinical chemistry tests, vitamin B12 and folic acid, thyroid hormones and HbA1c determinations were performed according to standard routine laboratory testing procedures.

As preanalytical factors crucially affect levels of growth factors, the collection and processing of specimens was carried out under strictly standardized conditions.

Whole blood for DNA extraction and serum samples for growth factor measurements were stored at -80°C within the MedUni Vienna Biobank facility. Post-storage isolation of DNA was done on a Corbett X-tractor Gene CAS 1820 semi-automated nucleic acid extraction system (Qiagen, Hilden, Germany) using a Macherey Nagel Nucleospin 96 blood kit (Macherey Nagel, Dueren, Germany). Quantification of genomic DNA by Warburg-Christian method (260/280 nm) on a Nanodrop spectrophotometer (PEQLAB Biotechnologie GmbH, Erlangen, Germany) revealed an average nucleic acid concentration of 25 ng/µL. ApoE genotyping of 1:10 (v/v in Buffer BE, Macherey Nagel) diluted aliquots was performed on a ABI TaqMan® 7900HT Real Time(RT)-PCR thermocycler (Applied Biosystems, Rotkreuz, Switzerland). For this, pre-designed TaqMan SNP-Genotyping assays to distinguish the ApoE ε4 allele from ε2 and ε3 at amino acid position 112 (ApoE rs429358, Assay ID C_3084793_20, Applied Biosystems) and the ApoE ε2 allele from ε3 and ε4 at amino acid position 158 (rs7412, Assay ID C_904973_10, Applied Biosystems) were purchased. RT-PCR was accomplished in a total reaction volume of 5µL using TaqMan® Genotyping Master Mix (Applied Biosystems) in a 384-well format according to the standard protocol supplied by the manufacturer. Thermal conditions: enzyme activation for 10 minutes at 95°C, followed by 45 cycles of alternating denaturation (15 seconds, 95°C) and primer annealing/elongation (1 minute, 60°C). Allelic discrimination was achieved using SDS 2.3 software (Applied Biosystems).

Insulin-like growth factor 1 was measured with the LIAISON® IGF-1 test on a fully automated LIAISON chemiluminescence analyzer (both from Diasorin, Saluggia, Italy).

Measurement of BDNF concentration was done manually using RayBio Human BDNF ELISA Kit (Ray Biotech, Inc, Norcross, USA) according to the standard protocol supplied by the manufacturer.

Magnetic resonance imaging

Magnetic resonance imaging was performed on a 1.5 T superconducting magnet (Siemens Symphony 1,5 T, Siemens Co., Erlangen) using a standard head coil, as previously described [40]. The standardized imaging protocol included:

1. axial FLAIR (fluid attenuated inversion recovery): TR 696 msec, TE 24 msec, 5mm slice thickness, distance factor 20%, FOV (field of view) 210 mm x 100 mm, number of slices 20.
2. axial T2* flash 2d: TR 477, TE 12 msec, 5mm slice thickness, distance factor 20%, FOV 210 x 100, number of slices 20. (c) axial T1 TSE (turbo spin echo sequence) TR 477 msec, TE 12 msec, 5mm slice thickness, distance factor 20%, FOV 210 x 100 mm, number of slices 20. (d) coronal T2 TSE: TR 4480 msec, TE 94 msec, high resolution (perpendicular to hippocampus), 2mm slice thickness, distance factor 20%, FOV 220 x 100, number of slices 24. (e) coronal 3D MPRAGE: TR 1420 msec, TE 3,2 msec, slice thickness (partition) 3 mm, FOV 240, number of slices 36. Quantitative, morphometric imaging data were not acquired.

Data management and statistical analyses A separate database was maintained by the biobank. All results were sent to the trial office of the Occupational Medicine Unit, where they were matched and appended to the participant´s records on an Access 2000 database.

All statistical analyses were performed using SPSS 17.0. Depending on the scale properties of the data, mean and standard deviation or frequencies and percentages are provided. Univariate group differences were evaluated by means of t-tests/Mann & Whitney U-Test or Fisher exact tests, and MANOVA models were used to evaluate multivariate group differences with Wilk's-being reported. Multiple correlations between the sets of psychological parameters and the biomarkers BDNF and IGF were performed exploratively. All statistical analyses were conducted at a significance level of 5%; due to the large number of comparisons and to reduce the risk of an inflation of the error type I the significance level will be adjusted to 2.5‰ (Bonferroni-Holm adjustment of error type I) when discussing the results. For the follow-up evaluation, sample sizes of n=41 patients per group are aimed at to achieve the detection of group differences in the percentage of cognitive impairment; this will make it possible to prove differences in percentages for median up to large effects (H>.50) at a level of significance of 5% and a power of 80%.

Inclusion Criteria:

• Current Marathon runners or marathon bicyclists and controls from the greater Vienna area

Exclusion Criteria:

• Current of past neurotoxic exposure
• Not German as native language
• Diseases, that might affect central nervous system (stroke, meningitis, meningeoma, hydrocephalus,..)
• Manifest cardiovascular disease
• Alcoholism

Marathon runners, bicyclists and controls from the greater Vienna area

Of 136 screened individuals (63 athletes, 73 controls), 56 athletes and 58 controls have been enrolled since they met the inclusion criteria.