Melatonin, a pineal hormone synthesized at night, is critical for the synchronization of circadian and seasonal rhythms, being a key regulator of energy metabolism in many animal species. Although studies in humans are lacking, several reports, mainly on hibernating animals, demonstrated that melatonin supplementation and a short photoperiod increase brown adipose tissue (BAT) mass. The present proof-of-concept study is the first, to our knowledge, to evaluate BAT in patients with melatonin deficiency (radiotherapy or surgical removal of pineal gland) before and after daily melatonin (3 mg) replacement for 3 months. All four studied patients presented increased BAT volume and activity measured by positron emission tomography-MRI. We also found an improvement in total cholesterol and triglyceride blood levels without significant effects on body weight, liver fat, and HDL and LDL levels. Albeit not statistically significant, fasting insulin levels and HOMA of insulin resistance decreased in all four patients. The present results show that oral melatonin replacement increases BAT volume and activity and improves blood lipid levels in patients with melatonin deficiency, suggesting that melatonin is a possible BAT activator. Future studies are warranted because hypomelatoninemia is usually present in aging and appears as a result of light-at-night exposure and/or the use of β-blocker drugs.
Introduction
Melatonin, a pineal hormone synthesized and released at night that plays a critical role in the synchronization of circadian and seasonal rhythms, has been studied as a key regulator of energy metabolism in many animal species for a long time (1,2). Pinealectomized rats show increased body weight gain and metabolic disturbances that are prevented by daily melatonin supplementation at night, without decreasing energy intake (1–5). This suggests that melatonin also regulates energy metabolism by its action on energy expenditure, possibly as related to the activation of brown adipose tissue (BAT). Indeed, melatonin has been shown to increase BAT recruitment, measured as increased BAT mass, in experimental models, such as hibernating animals (3). BAT has long been recognized as a thermogenic tissue in mammals, but its significance in humans was considered to be minor and limited to newborns. Recently, positron emission tomography (PET) with 2-deoxy-2-[fluorine-18] fluoro-d-glucose (FDG) showed that human adults present active BAT, especially after cold exposure (6,7). This finding quickly led to an exponential increase in BAT research, since activation of BAT leads to increased energy expenditure that could, at least theoretically, be a possible tool for the treatment of obesity and type 2 diabetes (8). Many compounds, including melatonin, have been studied to determine their ability of BAT recruitment and activation, although none of these studies have been carried out in humans (3,9). Melatonin has been used in the treatment of several pathologies, including sleep disturbances, neurodegenerative diseases, cardiovascular diseases, and cancer (2). As far as human endocrine and metabolic diseases are concerned, the importance of regular melatonin daily secretion in determining adequate energy balance and glucose metabolism is seen in several articles that showed a negative correlation between the levels of nocturnal melatonin secretion and BMI, insulin resistance (IR), and type 2 diabetes incidence (10,11). In addition, a few articles demonstrated that chronic therapeutic daily melatonin administration can induce a reduction of fat mass and an increase of lean mass besides being able to counteract antipsychotic drug–induced metabolic adverse effects, including overweight (12–14). However, the exact mechanism by which melatonin exerts its metabolic effects in humans is not known.
In the present study, we sought to determine whether melatonin therapy for patients with melatonin deficiency as a result of pinealectomy could increase BAT activation evaluated by FDG-PET after cold exposure as a proof of concept of a potential influence of melatonin in BAT in humans. Since pineal tumors are very rare, only four patients could be selected. Metabolic parameters were also measured, since a recent meta-analysis suggested a role of melatonin in improving total cholesterol and triglyceride levels (15). Given that high exposure to light at night (LAN) is associated with a reduction in or even suppression of melatonin levels, commonly leading to individuals having functional melatonin deficiency, an investigation in this particular population would also be relevant (16). However, the evaluation of melatonin replacement in patients after pinealectomy is very important, given that despite being widely accepted in classical endocrinology that absolute hormone deficiency should be treated with hormonal replacement, data about melatonin replacement in these patients are scarce.
Research Design and Methods
Ethical Statement
This protocol has been approved by the Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo ethical committee and is registered in the Brazilian Unified Research Platform (Plataforma Brasil) under the number 30460114.5.0000.0068. All patients signed an informed consent approved by the ethical committee.
Recruitment and Selection of Patients
We searched for patients referred to surgical removal of the pineal gland or to radiotherapy in the pineal area because of pineal tumors at Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, a reference center for complex and rare diseases in Brazil. We were able to find records of 18 patients in the past 12 years. Five patients could not be contacted because of a lack of an updated address and phone number, three patients died, four patients refused to participate, one did not comply with the protocol, and one was excluded because of a sympathetic lesion caused by an infiltrating tumor. The remaining four patients were studied, and all were male. The characteristics of the patients are described in the Supplementary Data. We had some concern about patient 3 because of the presence of panhypopituitarism that could bias the analysis, as it is well known that cortisol levels present a circadian rhythm that could influence BAT (17). Despite this, we did not exclude the patient because of the low number of subjects. Before melatonin treatment, all subjects had salivary melatonin measured every 3 h on a 27-h schedule (from 7 p.m. until 10 p.m. the next day), and the absence or very low levels of melatonin were confirmed by ELISA (IBL International, Hamburg, Germany). Salivary melatonin was also collected just before the end of melatonin treatment.
Melatonin Replacement
Patients received 108 tablets of melatonin 3 mg (Aché, Guarulhos, Brazil) and were advised to take one tablet 30 min before sleep every day for 3 months. Blister packs were brought back to evaluate adherence.
FDG-PET/MRI
After overnight fasting, the patients were admitted to the Nuclear Medicine Institute. They were exposed to a cold, air-conditioned room (18°C) wearing light clothes and a cooling vest (Polar Products, Stow, OH) for 2 h. After 1 h of cold exposure, the FDG radioisotope was injected, and PET-MRI was performed 1 h later in accordance with protocols used in other studies in the field (18,19).
BAT volume (mL) and activity (volume × mean standardized uptake value [SUV]) with an SUV threshold of 1.5 and 2.0 were assessed by PET-MRI, and liver fat was analyzed by MRI. The software AMIDE (Amide’s a Medical Imaging Data Examiner; http://amide.sourceforge.net/packages.html) was used to analyze BAT volume and SUV by a nuclear medicine specialist. Adipose tissue was automatically segmented by function approximation techniques magnetic resonance attenuation correction (FAT-MRAC) or liver imaging with volume acceleration-flexible (LAVA-Flex) sequence using an SUV signal over 1.5 and 2.0 as the threshold for detection of adipose tissue. The activity of the renal pelvis was then subtracted after manually defining the area of interest to exclude urinary activity in regions of possibly erroneously identified adipose tissue. Volumes and activities inside the adipose tissue segment were then calculated. The above procedure was done before and after melatonin replacement.
Blood Test Analysis
We collected blood samples immediately before the PET-MRI protocol both before and after melatonin replacement. Total cholesterol, HDL cholesterol, LDL cholesterol, fasting triglycerides, fasting insulin, and fasting glucose levels were determined and analyzed in the Central Laboratory, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo. A difference in any of them was then considered a secondary end point.
Statistical Analysis
Data were analyzed by paired t test with one-tailed distribution since an increase in BAT was expected as well as a decrease in triglyceride and total cholesterol levels. For small sample sizes, normality tests have little power to reject the null hypothesis, and, therefore, small samples most often pass normality tests.
End Points
The primary end points were an increase in BAT volume and/or activity by PET-MRI. The secondary end points were differences in the blood levels of any of the metabolites studied.
Results
All four patients completed the protocol, with an adherence of >90%. The study protocol also analyzed sleep patterns, quality of life, and dysautonomia tests, but those data will not be presented here since there are several end points still to be measured with a longer duration of the treatment.
The individual patterns of BAT volume and activity are shown in Fig. 1A and B, respectively, considering an SUV threshold of 2.0. Both BAT volume and BAT activity statistically increased in all patients after melatonin replacement (P = 0.0179 and 0.0139, respectively). BAT volume and activity considering an SUV threshold of 1.5 are presented in the Supplementary Data and have a very similar pattern of increase (P = 0.0199 and 0.0159, respectively).
Patient 1 had a notably clear pattern of response (Fig. 2) with volume and activity increase of 2.7 and 2.39 times, respectively, after melatonin replacement. Nevertheless, the results could be potentially affected by external factors, such as outdoor temperature and photoperiod, but a correlation between outdoor temperature and BAT volume and/or activity was not found (Supplementary Data).
Despite the positive responses observed in BAT, the patients’ body weight did not change (Table 1) after melatonin replacement therapy. Although significant differences in weight in such a small sample were not expected, three patients presented a small weight gain. Maybe a larger sample and/or a longer supplementation could lead to a different result in body weight.
. | Metabolic parameters . | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | HDL-C (mg/dL) . | LDL-C (mg/dL) . | Total cholesterol (mg/dL) . | Triglyceride (mg/dL) . | Fasting glucose (mg/dL) . | Fasting insulin (µU/mL) . | HOMA-IR . | Body weight
(kg) . | ||||||||
Patient | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After |
1 | 28 | 42 | 69 | 54 | 120 | 109 | 153 | 54 | 91 | 54 | 16.1 | 4.0 | 3.6 | 0.5 | 54.2 | 51.2 |
2 | 32 | 45 | 113 | 92 | 162 | 155 | 94 | 90 | 78 | 77 | 13.4 | 10.6 | 2.6 | 2.0 | 55.7 | 56.7 |
3 | 29 | 26 | 91 | 108 | 223 | 193 | 552 | 479 | 76 | 81 | 22.2 | 19.8 | 4.2 | 4.0 | 76.0 | 79.2 |
4 | 37 | 38 | 111 | 72 | 177 | 143 | 338 | 249 | 88 | 98 | 24.4 | 8.6 | 5.2 | 2.1 | 98.5 | 99.0 |
Mean | 31.5 | 37.5 | 96 | 81.5 | 170 | 150 | 284 | 218 | 83.25 | 77.5 | 19.0 | 10.75 | 3.85 | 2.15 | 71.1 | 71.5 |
P value (paired t test, one-tailed) | 0.11 | 0.15 | 0.0279* | 0.0268* | 0.31 | 0.05 | 0.058 | 0.76 |
. | Metabolic parameters . | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
. | HDL-C (mg/dL) . | LDL-C (mg/dL) . | Total cholesterol (mg/dL) . | Triglyceride (mg/dL) . | Fasting glucose (mg/dL) . | Fasting insulin (µU/mL) . | HOMA-IR . | Body weight
(kg) . | ||||||||
Patient | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After | Before | After |
1 | 28 | 42 | 69 | 54 | 120 | 109 | 153 | 54 | 91 | 54 | 16.1 | 4.0 | 3.6 | 0.5 | 54.2 | 51.2 |
2 | 32 | 45 | 113 | 92 | 162 | 155 | 94 | 90 | 78 | 77 | 13.4 | 10.6 | 2.6 | 2.0 | 55.7 | 56.7 |
3 | 29 | 26 | 91 | 108 | 223 | 193 | 552 | 479 | 76 | 81 | 22.2 | 19.8 | 4.2 | 4.0 | 76.0 | 79.2 |
4 | 37 | 38 | 111 | 72 | 177 | 143 | 338 | 249 | 88 | 98 | 24.4 | 8.6 | 5.2 | 2.1 | 98.5 | 99.0 |
Mean | 31.5 | 37.5 | 96 | 81.5 | 170 | 150 | 284 | 218 | 83.25 | 77.5 | 19.0 | 10.75 | 3.85 | 2.15 | 71.1 | 71.5 |
P value (paired t test, one-tailed) | 0.11 | 0.15 | 0.0279* | 0.0268* | 0.31 | 0.05 | 0.058 | 0.76 |
HDL-C, HDL cholesterol; LCL-C, LDL cholesterol.
*Statistically significant.
Although such a small sample was analyzed, melatonin replacement caused a significant decrease in total cholesterol (P = 0.0204) and triglyceride (P = 0.0104) levels (Table 1). With regard to glycemic parameters, no differences were observed; however, fasting insulin and HOMA-IR decreased in all subjects after melatonin replacement (Table 1). Hepatic fat was also evaluated by MRI, and no difference was found (Supplementary Data).
Discussion
The present study shows a significant increase in BAT volume and activity after melatonin replacement therapy as measured by PET-MRI in four patients with melatonin deficiency as a result of pinealectomy. Despite the very small sample size, the increase was observed in all patients, suggesting that melatonin may be a potential BAT recruiter that should be taken into consideration in further studies in humans, at least in situations in which a reduction in melatonin production occurs. Many experimental studies associated melatonin treatment with an increase in BAT mass as well as melatonin deficiency with a decrease in BAT mass, but these data were restricted to animal studies, mainly hibernating animals (3), and were obtained before the identification of active BAT in human adults (6–8). Today, BAT is being studied in view of its possible relation with obesity, type 2 diabetes, and other metabolic diseases, and potential BAT activators and recruiters have been proposed (20) as adjuvant treatment of these diseases. Some authors hypothesized that an increase in BAT not only would lead to increased energy expenditure after cold exposure but also could be related to diet-induced thermogenesis (21,22).
The present observed increase in volume suggests that melatonin replacement is able to recruit BAT, with a greater activation in response to the acute cold challenge. Melatonin replacement to pinealectomized animals decreases body weight gain with a very small decrease in food intake, with the increased energy expenditure being the most plausible mechanism for the final body weight reduction (5). Taking these experimental data and the present results into account, it is possible to speculate that the same may be true in humans (1), even though energy expenditure was not addressed.
Because several clinical and social conditions that present hypomelatoninemia, such as LAN, diabetes, and neurological disorders (2), are associated with increased body fat, a reduction in BAT as a result of the hypomelatoninemic condition could be a possible explanation for this correlation (23), but we do not have any direct evidence of that. Importantly, we do not believe that melatonin could be an independent antiobesity drug, but rather, a deficiency in melatonin could lead to weight gain in the long-term (5), and its replacement or supplementation could prevent weight gain in some obesogenic situations, such as reduced BAT activity after the use of antipsychotic drugs (13,14).
It is interesting to note that patient 2 had a high BAT activity even before melatonin replacement, suggesting that BAT recruitment and/or activation can occur in the absence of melatonin. However, the further increase in activity after replacement in this patient suggests an important role played by melatonin even when BAT is already reasonably recruited.
The observed reduction in triglyceride and total cholesterol levels is in agreement with the conclusion of a recent meta-analysis (15). However, none of the studies included in the meta-analysis evaluated patients with melatonin deficiency, and the present results with such a small sample suggest that the clinical benefit to lipid levels would be much higher in patients with melatonin deficiency. Such a suggested positive correlation between the therapeutic effect of melatonin and the previous endogenous amount of melatonin production was already demonstrated in sleep and daily activity/rest rhythm clinical studies (24). Notably, two of our four patients had an increase in HDL cholesterol of ∼10 mg/dL, with clinical significance. Differently from other studies, liver fat evaluated by MRI was not affected by melatonin supplementation (25,26). Although no difference was observed in the glycemic parameters (which is somehow expected because of the small number of patients), fasting insulin and HOMA-IR were reduced in all four patients after melatonin replacement (despite the weight gain in three of them), with a clinically significant reduction in patients 1 and 4 (Table 1). In our opinion, it is possible to speculate that a reduction in IR could likely be observed in a larger sample and can possibly be related to an increase in BAT activity and recruitment after melatonin treatment because BAT activation is known to improve whole-body glucose homeostasis in humans (18).
This small pilot study is innovative for several reasons. It is the first study in our knowledge to objectively analyze BAT activity after melatonin treatment in humans, despite a huge amount of data in animals. Although the clinical significance of this finding cannot be clearly defined (since energy expenditure was not addressed, the sample is small and the duration is shorter than required to evaluate body weight or glycemic responses), melatonin and its agonists could be considered potential BAT-recruiting agents for future research. Moreover, we studied patients with melatonin deficiency as a result of pinealectomy, individuals who have received very little attention in the previous literature. There is virtually no study (with the exception of a few individual case reports) that evaluated melatonin replacement in these patients (27). Since melatonin is a known chronobiotic in humans (2), we strongly expect that its replacement therapy could improve sleep and quality of life of our present patients. Many decades of research in pinealectomized animals have shown important metabolic derangements, such as weight gain, IR, and decreased thermoregulatory responses, as well as remarkable effects of melatonin replacement in reversing these alterations (1,2). However, no data are available in humans that evaluated either the baseline metabolic characteristics of patients who had undergone pinealectomy or the effects of melatonin replacement therapy. Besides that, we decided to study patients with pinealectomy to mirror the evidence obtained in animal models, in which the metabolic effect of melatonin replacement is clearly evident. The positive results of the present small study warrant a placebo-controlled study to evaluate sleep, cardiovascular disease, and energy metabolism in a larger sample of patients with melatonin deficiency, potentially indicating a systematic melatonin replacement treatment of those patients in the future. In addition, once the putative therapeutic effects of melatonin are shown in this melatonin-deficient population, future studies should also look at several clinical and social situations leading to the so-called hypomelatoninemia condition (2), such as shift work, excessive LAN exposure, β-blocker use, neurologic disorders, and aging (28–30). All these conditions are associated with weight gain and metabolic diseases, and it remains to be determined whether melatonin supplementation has an impact on body weight and metabolic parameters also in these patients. We decided not to study control subjects since they are regular producers of melatonin, and the use of melatonin for 3 months in healthy individuals could have low adherence because of putative side effects, such as headache, drowsiness, or daytime somnolence (31).
In conclusion, in this small sample of patients who had undergone pinealectomy with confirmed very low melatonin levels, replacement with 3 mg of melatonin increased BAT volume and activity analyzed by PET-MRI as well as reduced total cholesterol and triglyceride levels. Melatonin could be considered a possible BAT recruitment agent in humans with the potential for future therapeutic use. Further research on melatonin supplementation in a larger sample of patients with pinealectomy and other clinical situations associated with melatonin reduction is warranted since the paucity of data in the literature is evident.
Article Information
Acknowledgments. The authors thank the late Dr. Alfredo Halpern (Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo) for the original idea of the protocol. The authors also thank Dr. Licio Augusto Velloso (Universidade Estadual de Campinas) for lending the water vest for our research.
Funding. This study was funded by Fundação de Amparo à Pesquisa do Estado de São Paulo grant 09/52920-0.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. B.H. conducted all steps of the research. B.H. wrote the manuscript. B.H., M.C.M., and M.E.d.M. performed the statistical analysis of all data. B.H., M.C.M., and J.C.-N. designed the experiments and analyzed the data. M.C.M. and J.C.-N. reviewed the manuscript and gave opinions and ideas. C.B. and I.P.B. recruited the patients and helped with the collection of melatonin salivary samples. M.S.L., C.G.C., M.T.S., and C.A.B. were responsible for the PET-MRI procedures and analyses. F.G.d.A. analyzed salivary melatonin samples and gave important opinions and ideas about the procedures. J.C.-N. is responsible for the thematic project in which this trial belongs and received funding. J.C.-N. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.