The influence of bright and dim light on substrate metabolism, energy expenditure and thermoregulation in insulin-resistant individuals depends on time of day

Abstract

Aims/hypothesis: In our modern society, artificial light is available around the clock and most people expose themselves to electrical light and light-emissive screens during the dark period of the natural light/dark cycle. Such suboptimal lighting conditions have been associated with adverse metabolic effects, and redesigning indoor lighting conditions to mimic the natural light/dark cycle more closely holds promise to improve metabolic health. Our objective was to compare metabolic responses to lighting conditions that resemble the natural light/dark cycle in contrast to suboptimal lighting in individuals at risk of developing metabolic diseases.

Methods: Therefore, we here performed a non-blinded, randomised, controlled, crossover trial in which overweight insulin-resistant volunteers (n = 14) were exposed to two 40 h laboratory sessions with different 24 h lighting protocols while staying in a metabolic chamber under real-life conditions. In the Bright day-Dim evening condition, volunteers were exposed to electric bright light (~1250 lx) during the daytime (08:00-18:00 h) and to dim light (~5 lx) during the evening (18:00-23:00 h). Vice versa, in the Dim day-Bright evening condition, volunteers were exposed to dim light during the daytime and bright light during the evening. Randomisation and allocation to light conditions were carried out by sequential numbering. During both lighting protocols, we performed 24 h indirect calorimetry, and continuous core body and skin temperature measurements, and took frequent blood samples. The primary outcome was plasma glucose focusing on the pre- and postprandial periods of the intervention.

Results: Spending the day in bright light resulted in a greater increase in postprandial triacylglycerol levels following breakfast, but lower glucose levels preceding the dinner meal at 18:00 h, compared with dim light (5.0 ± 0.2 vs 5.2 ± 0.2 mmol/l, n = 13, p=0.02). Dim day-Bright evening reduced the increase in postprandial glucose after dinner compared with Bright day-Dim evening (incremental AUC: 307 ± 55 vs 394 ± 66 mmol/l × min, n = 13, p=0.009). After the Bright day-Dim evening condition the sleeping metabolic rate was identical compared with the baseline night, whereas it dropped after Dim day-Bright evening. Melatonin secretion in the evening was strongly suppressed for Dim day-Bright evening but not for Bright day-Dim evening. Distal skin temperature for Bright day-Dim evening was lower at 18:00 h (28.8 ± 0.3°C vs 29.9 ± 0.4°C, n = 13, p=0.039) and higher at 23:00 h compared with Dim day-Bright evening (30.1 ± 0.3°C vs 28.8 ± 0.3°C, n = 13, p=0.006). Fasting and postprandial plasma insulin levels and the respiratory exchange ratio were not different between the two lighting protocols at any time.

Conclusions/interpretation: Together, these findings suggest that the indoor light environment modulates postprandial substrate handling, energy expenditure and thermoregulation of insulin-resistant volunteers in a time-of-day-dependent manner.

Trial registration: ClinicalTrials.gov NCT03829982.

Funding: We acknowledge the financial support from the Netherlands Cardiovascular Research Initiative: an initiative with support from the Dutch Heart Foundation (CVON2014-02 ENERGISE).

Keywords: Biological clock; Circadian rhythm; Glucose intolerance; Insulin resistance; Light at night; Light exposure; Melatonin; Postprandial metabolism; Sleeping metabolic rate.

Fig. 1

Study scheme. Fasted blood samples were taken at 07:45 h on days 2 and 3 and at 17:45 h on day 2. Postprandial blood samples were taken for 4 h at 30 min intervals after both breakfasts and dinner. Slow stepping exercise for 30 min was performed at 12:30 h preceding lunch and at 15:30 h

Fig. 2

Overview of different skin temperature outcomes averaged over 30 min intervals (n = 13). T_proximal (a), T_distal (b) and DPG (c). The dashed lines with the light bulb indicate the time points when the light settings were changed. Data are presented as mean ± SEM

Fig. 3

Plasma melatonin (n = 14) in the evening of day 2 upon Bright day–Dim evening (a) and Dim day–Bright evening (b). Lines represent individual data. Data points below the detection threshold of 1.9 pg/ml (8.18 pmol/l) are illustrated as 0 values. The horizontal dashed line indicates the DLMO threshold of 10 pg/ml (43.05 pmol/l). To convert melatonin values from pg/ml to pmol/l, please multiply by 4.305

Fig. 4

Postprandial plasma responses for the two meals on day 2 (Breakfast1 [n = 13] and Dinner [n = 14]) and breakfast on day 3 (Breakfast2 [n = 13]). Blood glucose (a) and TG (d) for the first breakfast; blood glucose (b), TG (e) and insulin (g) for the dinner; blood glucose (c), TG (f) and insulin (h) for the second breakfast. Postprandial data were analysed using a generalised linear mixed model with time and light conditions and their interaction as fixed effects. Data are presented as mean ± SEM; ***p<0.001, *p<0.05 (note that there is a single * symbol in Fig. 4b, partially obscured by the y-axis)

Fig. 5

Energy expenditure over the entire time spent in the respiration chamber (a) and SMR of both nights per condition (n = 13; Night1 refers to the first baseline night spent in the respiration chamber without any differences in light intervention; Night2 refers to the second night after the respective light intervention) (b); p values are based on paired t tests; **p<0.01. Data are presented as mean ± SEM