Neuropsychological verbal tests may lack "hold" properties in occupational studies of neurotoxic effects

In epidemiological studies of neurotoxic effects neuropsychological tests are often applied to assess possible functional changes. In these studies tests presumed to be resistant to neurotoxic effects, "hold tests", are often used to adjust performance in tests presumed to be sensitive to neurotoxic effects, "nonhold tests", in assessment of pre-exposure ability. This conception is based on experience from the examination of patients with organic brain disorders. For this adjustment to be valid hold tests should change comparatively little over time and should not change differently in groups of people with different exposures to neurotoxic agents.

These assumptions were examined in 71 carpenters and 135 painters divided in three subgroups according to level of cumulative exposure to organic solvents. The results were noted from two verbal tests (test A, which involved following verbal instructions and test C, word comprehension), which were performed at conscription (age 18-20). The tests resemble hold tests used in occupational studies. The same tests were performed again at the age of 45-60 together with a conventional synonym test (SRB 1), often used in occupational studies.

In the three tests given at the time of the investigation some differences were found between the carpenters and the subgroups of painters. The painters with low exposure tended to perform better than the carpenters in all three tests and the heavily exposed painters tended to perform less well in the tests. These differences were not present at the age of 18-20. In one of the conscription tests painters with different cumulative exposure to organic solvents developed differently over time. There was little improvement among painters with heavy and intermediate exposure and obvious improvement among painters with low exposure. The results in the other conscription test showed similar tendencies.

The results indicate that the hold tests examined do not meet basic criteria for hold tests used in occupational studies. Thus adjustments for conventional verbal tests seem inappropriate in such studies. It is possible that other presumed hold tests may meet the criteria but our results suggest that such tests also should be evaluated before being used as hold tests in working populations.