Color perception is controlled, at least in part, by retinal dopaminergic neurons . We propose that slowed color processing and color naming in ADHD reflects a specific problem in blue-yellow color perception, which arises from hypo-functioning retinal dopaminergic mechanisms. In the absence of any evidence to the contrary, changes in central and retinal dopamine are believed to occur together. Thus, hypofunctioning of the central dopaminergic system associated with ADHD  will be accompanied by hypo-functioning retinal dopamine. The abnormalities in retinal dopaminergic tone will give rise to subtle, but detrimental effects on the on several aspects of visual function, particularly on the short-wavelength chromatic pathway that is responsible for blue-yellow color perception. Normalization of central dopaminergic functioning via pharmacological intervention with psychostimulant medication will normalize retinal dopamine, which in turn will normalize blue-yellow color perception and performance on tasks requiring speeded color naming.
To present the detailed hypothesis, we first provide the necessary background on color perception and the role of retinal dopamine, then argue how hypo-dopaminergic functioning in ADHD will give rise to impairments in the tritan color mechanism, which in turn will influence speeded color naming on tasks involving a substantial proportion of blue and yellow stimuli.
Color perception and the role of retinal dopamine
Color perception is based on the three cone photoreceptor types maximally sensitive to long, middle, and short wavelengths in the perceived light spectrum that constitute two functionally and anatomically distinct systems at the retina and lateral geniculate nucleus; a 'red-green' system with a foveal specialization in which long and middle wavelength cone signals are antagonistic, and a 'blue-yellow' pathway in which short wavelength cones are opposed by a combined signals from long and middle wavelength cones without such a foveal overrepresentation . Normal development of the blue-yellow (tritan) color mechanism appears to lag behind that of the red-green mechanism, which is functional in human infants by 2 months of age [43–46].
Most color perception defects (i.e., dyschromatopsias or 'color blindness') are congenital and arise from altered sensitivity defects of the L and M cones. By contrast to these red-green color vision deficits, which respectively affect 2% and 6% of the male population, congenital defects in S-cone sensitivity are rare (about 0.01%) but affect both sexes equally . Acquired dyschromatopsias arising from exposure to environmental pollutants  usually impair blue-yellow color discrimination. For example, pronounced effects on color perception, often dose-dependent and involving the short-wavelength (blue-yellow) mechanism, are reported following both acute and chronic exposure to organic solvents and elemental mercury [49, 50]. Moreover, occupational exposure to organic solvents during pregnancy is associated with increased risk of color vision and acuity impairment in the offspring [51, 52].
Color perception problems, particularly involving the blue-yellow (tritan) mechanism, have also been linked with alterations in retinal dopamine: dopamine is a major neurotransmitter in the mammalian retina [41, 53]. Dopamine receptors, DRD1 and DRD4 that seems to be associated with a subsensitive postsynaptic receptor if coded by the 7-repeat allele  are both found in the retina [53–55]. Retinal dopaminergic neurons are involved in controlling the coupling of horizontal and amacrine cell lateral systems, the organization of the ganglion cell and the bipolar cell receptive fields and modulation of the physiological activity of photoreceptors. Thus, the retinal dopamine system influences light adaptation as well as other visual functions, including color perception, contrast sensitivity, and spatial and temporal processing [41, 53]. The hypothesis proposed focuses on the role of retinal dopamine in color perception.
Alterations in the level of retinal dopamine are reflected particularly in deficits in the short-wavelength chromatic pathway that is responsible for blue-yellow color discrimination, a system that appears to be especially vulnerable to the effects of disorders and drugs [41, 50, 56–58]. For example, discrimination along the blue-yellow axis (compared to the red-green axis) is particularly impaired in various disorders involving altered dopaminergic mechanisms. Thus, specific blue-yellow color vision disturbances are found in Tourette Syndrome , Parkinson's disease [60–63], and Huntington's disease . Changes of retinal dopamine levels arising from cocaine-withdrawal [65–67] and normal aging [41, 68] have also been associated with blue-yellow color vision losses.
The fundamental mechanisms causing a specific retinal impairment of color discrimination along the blue/yellow axis in dopaminergic disorders and acquired dyschromatopsias remain unclear. Short wavelength sensitive cones may be more fragile than long and medium wavelength sensitive cones or their relative scarcity and anatomical distribution may be responsible for the greater vulnerability of the blue-yellow perception by alterations of the dopaminergic system [41, 47, 53, 69]. Accordingly, abnormalities in dopamine production, transport, uptake, or receptor sensitivity could result in impairments in visual processing including color – particularly the color blue. Moreover a selective impairment of the blue-yellow vision system suggests a retinal location of the disturbance rather than a central one [41, 47].
Hypo-dopaminergic functioning in ADHD influences color perception
Hypo-dopaminergic functioning has been postulated in ADHD [4, 70] and abnormal levels and density of the dopamine transporter in the brain have been reported in adults with ADHD . Moreover, ADHD has been associated with anomalous alleles of the D1 and DRD4 receptors [72–74]. Thus, dopaminergically-related impairments in visual functioning, and particularly in blue-yellow color perception, are plausible in ADHD.
We propose that the observed impairments in individuals with ADHD on tasks requiring speeded color processing of blue-yellow stimuli might be attributable in part to hypo-functioning of both central and retinal dopamine. Critical to this hypothesis is the premise that alterations in central and retinal dopamine occur in parallel (and that pharmacologically-induced increases in central dopamine also increases retinal dopamine). To our knowledge there is no direct evidence that this is the case. Rather, we draw inferences from the following evidence: 1) cerebrospinal concentrations of a metabolite of CNS dopamine, homovanillic acid, correlates positively with electroretinogram blue-cone amplitude ; 2) in Parkinson's disease, death of dopaminergic neurons in the CNS also extend to the retina, resulting in impaired visual functions including blue-yellow color perception [61–63]; 3) the visual deficits in Parkinson's disease are mostly reversed by treatment with the dopamine precursor L-DOPA ; 4) methylphenidate, which is the primary treatment modality for ADHD, blocks the re-uptake mechanism of the dopamine transporter, increasing the amount of extracellular dopamine able to bind to its receptors . Thus, changes in the dopamine system, regardless of whether brought about experimentally (lesions, pharmacologically) or naturally (as in ageing or in clinical conditions), leads to predictable changes in retinal function [41, 53].
According to this retinal dopaminergic hypothesis, deficits in blue-yellow color perception result in poor performance on many of the standard neuropsychological tasks (e.g., Stroop, RAN, Wisconsin Card Sorting Task) that include a substantial proportion of blue-yellow stimuli. Thus poor task performance may reflect subtle impairments in color vision as well as, or instead of, impairments in higher-order cognitive function. Consistent with this prediction are the recent findings that individual differences in color perception influence performance on the classic Stroop color-word task and that incongruent opponent color pairs (e.g., the word BLUE in yellow ink) decrease the strength of Stroop interference compared to non-opponent color pairs (e.g., BLUE in red ink) . A neural network simulation of the data confirmed that the difference in magnitude of Stroop interference between incongruent color-word pairs involving opponent versus nonopponent colors was attributable to sensory processing of the physical color of the stimuli (particularly the color yellow), which occurs at the level of retinal ganglion cells . Also, it has been demonstrated that visual function deficiencies associated with normal aging (reduced acuity, contrast sensitivity, and color weakness) accounted for a substantial amount of the variance in Stroop performance . The preceding findings suggest that impairments in the early perceptual-encoding stage of stimulus color contribute to slow performance on neuropsychological tasks requiring speeded naming of color, as well as conceptual and attentional factors .
Stimulant-induced increases in the availability of dopamine would be expected to be reflected in retinal dopaminergic tone and thus have therapeutic effects on the short-wavelength mechanism. Thus, methylphenidate would be expected to improve visual functioning, including the speed of color processing and naming. Indeed, there is preliminary evidence of beneficial effects of methylphenidate on color naming in children with ADHD [8, 23].
Alterations in short-wavelengths mechanism may be gender-related. Notably, estrogen has been found to have a modulatory influence on dopamine activity  and a few studies have revealed significant variations in dopaminergic tone and dopamine receptor density that are sex-specific [79, 80]. Correspondingly, gender-related differences observed in visual-cortical fMRI BOLD response to blue light (higher BOLD signal change in males), but not to red light, have been reported that may be related to variations in dopamine function and/or the effects of estrogen on dopamine . Furthermore, visual pattern reversal evoked potentials have been found to vary with menstrual phase in females and display faster conduction times during the period of peak estrogen levels . Given the incidence of ADHD is estimated to be three times greater in males than females, the relationship between estrogen and dopamine may provide an important area for further investigation.
The limited data available on developmental changes in retinal and central dopaminergic mechanisms, color perception, and color naming, suggest that the retinal dopamine hypothesis will likely hold for children, adolescents, and adults with ADHD. For example, animal research indicates that dopaminergic neurons are among the first neurochemical systems to appear in the developing retina and that the neural retina and dopaminergic system interact closely in a two-way manner throughout developmental period . It is only as the animal passes through maturity towards senescence that the number of retinal dopamine neurons decrease . Thus, assuming that hyperdopaminergic function in ADHD occurs early in pre- or post-natal life, parallel hypofunctioning of the retinal dopamine system is expected to have a detrimental effect on the development of blue-yellow color vision and these early deficits are likely to persist.
On the other hand, the increasing efficiency of the mature (adult) brain permits the development and use of compensatory strategies. Thus, adults with ADHD, although impaired in color perception relative to healthy peers, may be able to call upon compensatory attentional strategies to enhance performance on tasks requiring rapid perception of blue and yellow stimuli, thereby exhibiting better performance than children or adolescents with ADHD.