Aging of human short-wave cone pathways

Abstract

The retinal image is sampled concurrently, and largely independently, by three physiologically and anatomically distinct pathways, each with separate ON and OFF subdivisions. The retinal circuitry giving rise to an ON pathway receiving input from the short-wave-sensitive (S) cones is well understood, but the S-cone OFF circuitry is more controversial. Here, we characterize the temporal properties of putative S-cone ON and OFF pathways in younger and older observers by measuring thresholds for stimuli that produce increases or decreases in S-cone stimulation, while the middle- and long-wave-sensitive cones are unmodulated. We characterize the data in terms of an impulse response function, the theoretical response to a flash of infinitely short duration, from which the response to any temporally varying stimulus may be predicted. Results show that the S-cone response to increments is faster than to decrements, but this difference is significantly greater for older individuals. The impulse response function amplitudes for increment and decrement responses are highly correlated across individuals, whereas the timing is not. This strongly suggests that the amplitude is controlled by neural circuitry that is common to S-cone ON and OFF responses (photoreceptors), whereas the timing is controlled by separate postreceptoral pathways. The slower response of the putative OFF pathway is ascribed to different retinal circuitry, possibly attributable to a sign-inverting amacrine cell not present in the ON pathway. It is significant that this pathway is affected selectively in the elderly by becoming slower, whereas the temporal properties of the S-cone ON response are stable across the life span of an individual.

Fig. 1.

Threshold data defined in terms of S-cone stimulation (S-cone Td) and the IRFs from one young observer (Upper) and one older observer (Lower). (Left) Log contrast thresholds for one pulse in S-cone Td as a function of the SOA. Blue squares and curves denote increment thresholds and model-fitting curves, respectively, obtained from the S-cone ON IRFs shown (Right). Yellow squares and curves denote decrement thresholds and the fit of the S-cone OFF IRFs. Error bars denote ±1 SEM. (Right) IRFs with the relative response plotted as a function of time. Bluish and yellowish curves denote S-cone ON (increment) and S-cone OFF (decrement) IRFs, respectively.

Fig. 2.

Mean peak (maximum) amplitude (Left) and peak time (time at the maximum intensity) (Right) of young and old observer groups’ S-cone ON and OFF IRFs calculated from the model. Error bars denote ±1 SEM.

Fig. 3.

Relations between S-cone ON and OFF IRFs for peak amplitude (Left) and peak time (Right). Triangles and circles denote the data points for younger and older observers, respectively. The lines are fitted by least-squares linear regression for all observers.

Fig. 4.

(Left) Average IRFs for younger and older S-cone ON and OFF IRFs compared with luminance modulation from a previously published study using the same methods (26). The amplitude of luminance modulation is scaled down by a factor of 10 for purposes of illustration. Squares denote points at peak amplitude and peak time and at duration (5% intensity of peak amplitude). Error bars denote ±1 SEM. (Right) Corresponding tCSFs calculated from the inverse Fourier transform. Luminance and chromatic modulation were defined as the amplitude of the flicker sine wave divided by the mean luminance or chromaticity. Peak sensitivity of the luminance function is normalized to unity, and minimum sensitivity is defined as −2 log unit, corresponding to flicker modulation from 0.01 to 1 (32). Squares denote points at 1 Hz and 4 Hz for each function. Error bars denote the possible range of curves defined by ±1 SEM.