Human cone light sensitivity and melatonin rhythms following 24-hour continuous illumination.

This study investigates the possibility of an endogenous circadian rhythm in retinal cone function in humans. A full-field cone electroretinogram (ERG) was performed every 2 h for 24 h under continuous rod-saturating ambient white light (53 ± 30 lux; pupils dilated) in nine healthy subjects. Distinct circadian variations were superimposed upon a gradual decrease in cone responsiveness to light, demonstrated most reliably in the implicit times of b-wave and oscillatory potentials, and to a lesser extent in amplitude and a-wave implicit times. After mathematical correction of the linear trend, the cone response was found to be greatest around 20:00 h and least around 06:00 h. The phase of the ERG circadian rhythm was not synchronized with the phase of the salivary melatonin rhythm measured the previous evening. Melatonin levels measured under constant light on the day of ERG assessments were suppressed by 53% on average compared to melatonin profiles obtained previously under near-total darkness in seven participants. The progressive decline in cone responsiveness to light over the 24 h may reflect an adaptation of the cone-driven retinal system to constant light, although the mechanism is unclear. The endogenous rhythm of cone responsiveness to light may be used as an additional index of central or retinal circadian clock time.

Human retinal light sensitivity and melatonin rhythms following four days in near darkness.

The rods in the retina are responsible for night vision, whereas the cone system enables day vision. We studied whether rod function in humans exhibits an endogenous circadian rhythm and if changes occur in conditions of prolonged darkness. Seven healthy subjects (mean age+/-SD: 25.6+/-12.3 yr) completed a 4.5-day protocol during which they were kept in complete darkness (days 1 and 4) and near darkness (<0.1 lux red light, days 2 and 3). Electroretinography (ERG) and saliva collections were done at intervals of at least 3 h for 27 h on days 1 and 4. Full-field ERGs were recorded over 10 low-intensity green light flashes known to test predominantly rod function. As a circadian marker, salivary melatonin concentration was measured by radioimmunoassay. The ERG data showed that rod responsiveness to light progressively diminished in darkness (significantly lower a- and b-wave amplitudes, longer b-wave implicit time). The decrease in amplitude (b-wave) from day 1 to day 4 averaged 22+/-14%. After correction for the darkness-related linear trend, the circadian variations in ERG indices were weak and usually non-significant, with slightly higher responsiveness to light during the day than night. Rod sensitivity (by K index) tended to decrease. Strikingly, the overall amount of melatonin secretion (area under 24 h curve) also decreased from day 1 to day 4 by 33.1+/-18.9% (p=.017). The drift of the melatonin rhythm phase was within the normal range, less than 56 min over three days. There was no significant correlation between the changes in ERG responses and melatonin. In conclusion, scotopic retinal response to (low-intensity) light and the amount of melatonin secreted are diminished when humans are kept in continuous darkness. Both processes may have a common underlying mechanism implicating a variety of neurochemicals known to be involved in the regulation of both photoreceptor and pineal gland function.

Influence of timed nutrient diet on depression and light sensitivity in seasonal affective disorder.

Seasonal Affective Disorder (SAD) patients crave and eat more carbohydrates (CHO) in fall-winter when depressed, especially in the evenings, and feel energetic thereafter. Evening CHO-rich meals can phase delay circadian rhythms, and glucose increases retinal response to light. We studied timed CHO- or protein-rich (PROT) diet as a putative therapy for SAD. Unmedicated, DSM-IV-diagnosed depressed women with SAD (n=22, 19-63 yrs) in the follicular phase of the menstrual cycle (present in 19) were randomized to nine days of eating approximately 1600 kcal of either CHO before 12:00 h (n=9), CHO after 18:00 h (n=6), or PROT after 18:00 h (n=7); only water was allowed for the rest of the day. Measurements included the depression questionnaire SIGH-SAD (with 21-item Hamilton depression subscale), Eating Behavior Questionnaire (DEBQ), percentage fat (by bioimpedancemetry), clinical biochemistry (glucose, cholesterol, triglycerides, TSH, T4, cortisol), and electroretinogram (ERG). No differential effects of diet were found on any of the studied parameters (except DEBQ). Clinically, participants improved slightly; the 21-HDRS score (mean+/-SD) decreased from 19.6+/-6.4 to 14.4+/-7.4 (p=.004). Percent change correlated significantly with menstrual day at diet onset (mood improved the first week after menstruation onset), change in available sunshine (more sunlight, better mood), and initial percentage fat (fatter patients improved more). Scotopic ERG amplitude was diminished after treatment (p=.025, three groups combined), probably due to greater exposure to sunshine in 14/22 subjects (partial correlation analysis significant). Keeping in mind the limitations of this ambulatory study (i.e., inability to control outdoor light exposure, small number of participants, and briefness of intervention), it is suggested that the 25% clinical improvement (of the order of magnitude of placebo) is not related to nutrient diet or its timing, but rather to natural changes during the menstrual cycle, available sunshine, and ease of dieting for fatter patients.