Acta Psychiatr Scand 2017: 1–12 All rights reserved DOI: 10.1111/acps.12772 © 2017 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA PSYCHIATRICA SCANDINAVICA Solar insolation in springtime influences age of onset of bipolar I disorder Bauer M, Glenn T, Alda M, Aleksandrovich MA, Andreassen OA, Angelopoulos E, Ardau R, Ayhan Y, Baethge C, Bharathram SR, Bauer R, Baune BT, Becerra-Palars C, Bellivier F, Belmaker RH, Berk M, Bersudsky Y, Bicakci S ß , Birabwa-Oketcho H, Bjella TD, Bossini L, Cabrera J, Cheung EYW, Del Zompo M, Dodd S, Donix M, Etain B, Fagiolini A, Fountoulakis KN, Frye MA, Gonzalez-Pinto A, Gottlieb JF, Grof P, Harima H, Henry C, Isomets€ a ET, Janno S, Kapczinski F, Kardell M, Khaldi S, Kliwicki S, K€ onig B, Kot TL, Krogh R, Kunz M, Lafer B, Landen M, Larsen ER, Lewitzka U, Licht RW, Lopez-Jaramillo C, MacQueen G, Manchia M, Marsh W, Martinez-Cengotitabengoa M, Melle I, Meza-Urz ua F, Yee Ming M, Monteith S, Morken G, Mosca E, Munoz R, Mythri SV, Nacef F, Nadella RK, Nery FG, Nielsen RE, O’Donovan C, Omrani A, Osher Y, Østermark Sørensen H, Ouali U, Pica Ruiz Y, Pilhatsch M, Pinna M, da Ponte FDR, Quiroz D, Ramesar R, Rasgon N, Reddy MS, Reif A, Ritter P, Rybakowski JK, Sagduyu K, ^ Scippa AM, Severus E, Simhandl C, Stein DJ, Strejilevich S, Subramaniam M, Sulaiman AH, Suominen K, Tagata H, Tatebayashi Y, Tondo L, Torrent C, Vaaler AE, Veeh J, Vieta E, Viswanath B, Yoldi-Negrete M, Zetin M, Zgueb Y, Whybrow PC. Solar insolation in springtime influences age of onset of bipolar I disorder. Objective: To confirm prior findings that the larger the maximum monthly increase in solar insolation in springtime, the younger the age of onset of bipolar disorder. Method: Data were collected from 5536 patients at 50 sites in 32 countries on six continents. Onset occurred at 456 locations in 57 countries. Variables included solar insolation, birth-cohort, family history, polarity of first episode and country physician density. Results: There was a significant, inverse association between the maximum monthly increase in solar insolation at the onset location, and the age of onset. This effect was reduced in those without a family history of mood disorders and with a first episode of mania rather than depression. The maximum monthly increase occurred in springtime. The youngest birth-cohort had the youngest age of onset. All prior relationships were confirmed using both the entire sample, and only the youngest birth-cohort (all estimated coefficients P < 0.001). Conclusion: A large increase in springtime solar insolation may impact the onset of bipolar disorder, especially with a family history of mood disorders. Recent societal changes that affect light exposure (LED lighting, mobile devices backlit with LEDs) may influence adaptability to a springtime circadian challenge. M. Bauer1,* , T. Glenn2, M. Alda3, M. A. Aleksandrovich4, O. A. Andreassen5, E. Angelopoulos6, R. Ardau7, Y. Ayhan8, C. Baethge9, S. R. Bharathram10, R. Bauer1, B. T. Baune11, C. Becerra-Palars12, F. Bellivier13 , R. H. Belmaker14, M. Berk15,16, Y. Bersudsky14, S ß . Bicakci8, H. Birabwa-Oketcho17, T. D. Bjella5, L. Bossini18, J. Cabrera19, E. Y. W. Cheung20, M. Del Zompo7, S. Dodd15,21, M. Donix1, B. Etain13 , A. Fagiolini18, K. N. Fountoulakis22, M. A. Frye23, A. Gonzalez-Pinto24, J. F. Gottlieb25, P. Grof26, H. Harima27, C. Henry28, E. T. Isomets€ a29,30, 31 S. Janno , F. Kapczinski32, M. Kardell33, S. Khaldi34,†, S. Kliwicki35, B. K€ onig36, T. L. Kot37, 38 R. Krogh , M. Kunz32, B. Lafer39, M. Landen40,41, E. R. Larsen38, U. Lewitzka1, R. W. Licht42,43, C. Lopez-Jaramillo44, G. MacQueen45, M. Manchia46, W. Marsh47, M. Martinez-Cengotitabengoa24, ua12, M. Yee I. Melle5, F. Meza-Urz 48 49 Ming , S. Monteith , G. Morken50,51, E. Mosca7, R. Munoz52, S. V. Mythri53, F. Nacef54, R. K. Nadella10, F. G. Nery39, R. E. Nielsen42,43, C. O’Donovan3, A. Omrani55, Y. Osher14, H. Østermark Sørensen42, U. Ouali54, Y. Pica Ruiz56, M. Pilhatsch1, M. Pinna57, F. D. R. da Ponte32, D. Quiroz58, R. Ramesar59, N. Rasgon60, M. S. Reddy53, A. Reif61, P. Ritter1, J. K. Rybakowski35, ^. M. Scippa63, K. Sagduyu62, A 1 E. Severus , C. Simhandl36, D. J. Stein64, S. Strejilevich65, M. Subramaniam66, A. H. Sulaiman67, K. Suominen68, H. Tagata27, Y. Tatebayashi69, L. Tondo70,71, C. Torrent72, A. E. Vaaler50,51, J. Veeh61, E. Vieta72 , B. Viswanath10, M. Yoldi-Negrete73, M. Zetin74, Y. Zgueb54, P. C. Whybrow75 1 Bauer et al. Significant outcomes • There was a strong, inverse association between the maximum monthly increase in solar insolation in • springtime and the age of onset of bipolar I disorder using a global sample. The effect was reduced in those without a family history of mood disorders. There was a large birth-cohort effect, with the youngest group having the youngest onset. Major societal changes that may affect vulnerability to a circadian challenge need investigation: exposure to LED lighting, mobile devices backlit with LEDs, and the 24-h society. Limitations • The data collection process was not standardized. Data on family history or age of onset may be • • 1 unreliable. There was no individual data on behaviours or exposures that affect circadian rhythms. The sample was not demographically representative of the country populations. Department of Psychiatry and Psychotherapy, University Hospital Carl Gustav Carus, Technische Universit€at Dresden, Dresden, Germany, 2ChronoRecord Association, Fullerton, CA, USA, 3Department of Psychiatry, Dalhousie University, Halifax, NS, Canada, 4Soviet Psychoneurological Hospital, Urai, Russia, 5NORMENT – K.G. Jebsen Centre for Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital & Institute of Clinical Medicine, University of Oslo, Oslo, Norway, 6Department of Psychiatry, Medical School, Eginition Hospital, National and Capodistrian University of Athens, Athens, Greece, 7Section of Neurosciences and Clinical Pharmacology, Department of Biomedical Sciences, University of Cagliari, Sardinia, Italy, 8Department of Psychiatry, Hacettepe University Faculty of Medicine, Ankara, Turkey, 9Department of Psychiatry and Psychotherapy, University of Cologne Medical School, Cologne, Germany, 10Department of Psychiatry, NIMHANS, Bangalore, India, 11Department of Psychiatry, School of Medicine, University of Adelaide, Adelaide, SA, Australia, 12National Institute of Psychiatry ‘“Ramon de la Fuente Mu~niz”, Mexico City, Mexico, 13Psychiatry and Addiction Medicine, Assistance Publique – H^opitaux de Paris, FondaMental Foundation, INSERM UMR-S1144, Denis Diderot University, Rene Descartes University,Paris, France, 14 Department of Psychiatry, Faculty of Health Sciences, Beer Sheva Mental Health Center, Ben Gurion University of the Negev, Beer Sheva, Israel, 15IMPACT Strategic Research Centre, School of Medicine, Barwon Health, Deakin University, Geelong, Vic., Australia, 16Department of Psychiatry, Orygen, the National Centre for Excellence in Youth Mental Health, the Centre for Youth Mental Health and the Florey Institute for Neuroscience and Mental Health, University of Melbourne, Parkville, Vic., Australia, 17Butabika Hospital, Kampala, Uganda, 18Department of Molecular Medicine and Department of Mental Health (DAI), University of Siena and University of Siena Medical Center (AOUS), Siena, Italy, 19 Mood Disorders Clinic, Dr. Jose Horwitz Psychiatric Institute, Santiago de Chile, Chile , 20Department of General Adult Psychiatry, Castle Peak Hospital, Tuen Mun, Hong Kong, 21 Department of Psychiatry, University of Melbourneo, Parkville, Vic, Australia, 22Division of Neurosciences, 3rd Department of Psychiatry, School of Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece, 23Department of Psychiatry & Psychology, Mayo Clinic Depression Center, Mayo Clinic, Rochester, MN, USA, 24Department of Psychiatry, University Hospital of Alava, University of the Basque Country, CIBERSAM, Vitoria, Spain, 25Department of Psychiatry, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA, 26Mood Disorders Center of Ottawa, University of Toronto, Toronto, ON, Canada, 27Department of Psychiatry, Tokyo Metropolitan Matsuzawa Hospital, Setagaya, Tokyo, Japan, 28AP-HP, Hopitaux Universitaires Henri Mondor and INSERM U955 (IMRB) and Universite Paris Est and Institut Pasteur, Unite Perception et Memoire, Paris, France, 29Department of Psychiatry, University of Helsinki and Helsinki University Hospital, Helsinki, Finland, 30National Institute for Health and Welfare, Helsinki, Finland, 31Department of Psychiatry, University of Tartu, Tartu, Estonia, 32Department of Psychiatry, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil, 33 Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, 34Private practice, Tunis, Tunisia, 35Department of Adult Psychiatry, Poznan University of Medical Sciences, Poznan, Poland, 36BIPOLAR Zentrum Wiener Neustadt, Wiener Neustadt, Austria, 37Khanty-Mansiysk Clinical Psychoneurological Hospital, Khanty-Mansiysk, Russia, 38Department of Affective Disorders, Q, Mood Disorders Research Unit, Aarhus University Hospital, Aarhus, Denmark, 39Bipolar Disorder Research Program, Department of Psychiatry, University of S~ao Paulo Medical School, S~ao Paulo, Brazil, 40Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, the Sahlgrenska Academy, University of Gothenburg, Gothenburg and M€olndal, Sweden, 41 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden, 42Unit for Psychiatric Research, Aalborg University Hospital, Psychiatry, Aalborg, Denmark, 43Department of Clinical Medicine, Aalborg University, Aalborg, Denmark, 44Mood Disorders Program, Hospital Universitario San Vicente Fundacion, Research Group in Psychiatry, Department of Psychiatry, Faculty of Medicine, Universidad de Antioquia, Medellín, Colombia, 45Department of Psychiatry, Faculty of Medicine, University of Calgary, Calgary, AB, Canada, 46Section of Psychiatry, Department of Medical Science and Public Health, University of Cagliari, Cagliari, Italy, 47Department of Psychiatry, University of Massachusetts, Worcester, MA, USA, 48Department of General Psychiatry, Mood Disorders Unit, Institute of Mental Health, Singapore City, Singapore, 49Traverse City Campus, Michigan State University College of Human Medicine, Traverse City, MI, USA, 50Department of Mental Health, Norwegian University of Science and Technology – NTNU, Trondheim, Norway, 51Department of Psychiatry, St Olavs’ University Hospital, Trondheim, Norway, 52Department of Psychiatry, University of California San Diego, San Diego, CA, USA, 53Asha Bipolar Clinic, Asha Hospital, Hyderabad, Telangana, India, 54Razi Hospital, Faculty of Medicine, University of Tunis-El Manar, Tunis, Tunisia, 55Tunisian  del Pedregal”, Mexico City, Mexico, 57Lucio Bini Mood Disorder Center, Cagliari, Italy, Bipolar Forum, Erable Medical Cabinet 324, Tunis, Tunisia, 56Hospital “Angeles 58 Deparment of Psychiatry, Diego Portales University, Santiago de Chile, Chile, 59UCT/MRC Human Genetics Research Unit, Division of Human Genetics, Institute of Infectious Disease and Molecular Medicine, University of Cape Town, Cape Town, South Africa, 60Department of Psychiatry and Behavioral Sciences, Stanford School of Medicine, Palo Alto, CA, USA, 61Department of Psychiatry, Psychosomatic Medicine and Psychotherapy, University Hospital Frankfurt, Johann Wolfgang Goethe-Universit€at Frankfurt am Main, Frankfurt am Main, Germany, 62Department of Psychiatry, University of Missouri Kansas City School of Medicine, Kansas City, MO, USA, 63Department of Neuroscience and Mental Health, Federal University of Bahia, Salvador, Brazil, 64Department of Psychiatry, MRC Unit on Risk & Resilience in Mental Disorders, University of Cape Town, Cape Town, South Africa, 65Bipolar Disorder Program, Neuroscience Institute, Favaloro University, Buenos Aires, Argentina, 66Research Division, Institute of Mental Health, Singapore City, Singapore, 67Department of Psychological Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia, 68Department of Social Services and Health Care, Psychiatry, City of Helsinki, Helsinki, Finland, 69Schizophrenia & Affective Disorders Research Project, Tokyo Metropolitan Institute of Medical Science, Seatagaya, Tokyo, Japan, 70 McLean Hospital-Harvard Medical School, Boston, MA, USA, 71Mood Disorder Lucio Bini Centers, Cagliari e Roma, Italy, 72Clinical Institute of Neuroscience, Hospital Clinic, University of Barcelona, IDIBAPS, CIBERSAM,Barcelona, Catalonia, Spain, 73Consejo Nacional de Ciencia y Tecnología - Instituto Nacional de Psiquiatría Ramon de la Fuente Mu~niz, Ciudad de Mexico, Mexico, 74Department of Psychology, Chapman University, Orange, CA, USA and 75Department of Psychiatry and Biobehavioral Sciences, Semel Institute for Neuroscience and Human Behavior University of California Los Angeles (UCLA), Los Angeles, CA, USA 2 Solar insolation and onset of bipolar disorder Key words: bipolar disorder; circadian rhythm; solar insolation; epidemiology Michael Bauer, Department of Psychiatry and Psychotherapy, University Hospital Carl Gustav Carus, Technische Universit€at Dresden, Fetscherstr. 74, 01307 Dresden, Germany. E-mail: michael.bauer@uniklinikum-dresden.de † Deceased in 2016. Accepted for publication June 16, 2017 Introduction There is considerable evidence of circadian rhythm dysfunction in patients with bipolar disorder including disturbances in the sleep/wake cycle, activity patterns, melatonin secretion, as well as suggestive associations with clock gene polymorphisms and epigenetic alterations (1–6). Clinical symptoms that are frequently reported include sleep timing disturbances (7), irregular daily schedules (8), and an evening preference (9). These symptoms of circadian disruption may occur during episodes, while euthymic (7, 10–13), and in those at high risk for bipolar disorder (7, 12, 14, 15). Even small changes to circadian rhythms such as the shift to daylight savings time may have adverse mental health consequences (16). In the future, circadian symptoms and clock gene polymorphisms may help define endophenotypes of bipolar disorder, including early onset (17–19). Some of the current treatments for bipolar disorder act directly or indirectly on circadian mechanisms, which may contribute to therapeutic effects (2). Lithium modulates the expression of central and peripheral clock genes (20–23), and the amplitude and timing of effects may differ between responders and non-responders (24). Lithium also has phase-delaying properties and may resynchronize overly fast circadian rhythms (2, 21). Other treatments for bipolar disorder that may modify circadian systems include valproate (2, 25) light and dark therapy and sleep deprivation (1, 26), and blue light blocking glasses (27). The importance of the circadian system to human health is becoming clearer from recent research, as summarized briefly. Over the course of evolution, life on Earth has adapted to the Sun (28). Humans have endogenous circadian timing cycles for nearly every physiological, metabolic, and behavioural system allowing for anticipation of light and dark, and adaptation to seasonal changes and environmental challenge (29–31). The cycle length of endogenous circadian rhythms is not exactly 24 h and must be regularly synchronized to the natural 24-h light–dark cycle that arises from one rotation of the Earth on its axis (32). Sunlight is the primary and most potent signal that entrains human circadian systems to the natural environment. Specialized non-visual receptors, photosensitive retinal ganglion cells (pRGC), express the melanopsin photopigment that is sensitive to short-wavelength blue light (33). The pRGC detect environmental fluctuations in light and project primarily to non-visual centers of the brain including the suprachiasmatic nucleus (SCN) in the hypothalamus (34). The SCN is the master pacemaker over a system that includes circadian clock genes expressed in the SCN, and throughout the rest of the brain, peripheral tissues and cells (35, 36). The signals from the pRGC entrain the SCN (37), which integrates the external signals with signals from internal activities, and in turn entrains the peripheral circadian clocks (35, 36). Timekeeping for optimal diurnal physiological processes and mental health requires both synchronization of the circadian systems with the natural environment, and internal system-wide circadian coordination. Aims of the study As the onset of bipolar disorder is highly variable, it is important to understand the factors that may influence it, including environmental (38, 39). We previously found a large, significant, inverse relation between the maximum monthly increase in solar insolation (incoming solar radiation striking the Earth’s surface) and the age of onset of bipolar I disorder (40, 41). This effect was reduced in those without a family history of mood disorders and was smaller for those with a first episode of mania rather than depression. The aim of this analysis was to confirm that the relations found previously were not sample specific, by repeating the analyses using significantly more data from geographically dispersed countries. Methods The data were collected by researchers at 50 collection sites in 32 countries. In the Northern Hemisphere, the collection sites were Aalborg, Denmark; Aarhus, Denmark; Ankara, Turkey; Athens, Greece; Bangalore, India; Barcelona, 3 Bauer et al. Spain; Beer Sheva, Israel; Cagliari, Italy (2 sites); Calgary, Canada; Dresden, Germany; Halifax, Canada; Helsinki, Finland; Hong Kong; Hyderabad, India; Kampala, Uganda; Kansas City, KS, USA; Khanti-Mansiysk, Russia; Kuala Lumpur, Malaysia; Los Angeles, CA, USA; Medellı́n, Colombia; Mexico City, Mexico; Oslo, Norway; Ottawa, Canada; Palo Alto, CA, USA; Paris, France; Poznan, Poland; Rochester, MN, USA; San Diego, CA, USA; Siena, Italy; Singapore; Stockholm/Gothenburg, Sweden; Tartu, Estonia; Thessaloniki, Greece; Tokyo, Japan; Trondheim, Norway; Tunis, Tunisia; Vitoria, Spain; Wiener Neustadt, Austria; Worcester, MA, USA, and W€ urzburg, Germany. In the Southern Hemisphere, the collection sites were: Adelaide, Australia; Buenos Aires, Argentina; Cape Town, South Africa; Melbourne/Geelong, Australia; Porto Alegre, Brazil; Salvador, Brazil; Santiago, Chile (2 sites); and S~ ao Paulo, Brazil. This analysis combined newly collected data with data collected and analyzed previously. Approval for this study was obtained from local institutional review boards according to local requirements. All patients had a diagnosis of bipolar disorder according to DSM-IV criteria from a psychiatrist, with age of onset defined as the first occurrence of an episode of mania, hypomania, or depression. Other patient data in this analysis were sex, family history of a mood disorder in any first degree relative, and polarity of the first episode. Data were obtained retrospectively by direct questioning, reviewing records, or both. Solar insolation Solar insolation is defined as the amount of electromagnetic energy from the Sun received on Earth for a given surface area at a given time, expressed in kilowatt h/square meter/day (kWh/m2/day) (42). Several factors determine the intensity of solar insolation including the angle at which the Sun’s rays strike the Earth’s surface, time of day, latitude, atmospheric conditions, surface reflection, and the Earth’s tilt. The tilt of Earth’s axis relative to the plane of its orbit around the Sun results in the seasonal changes in solar insolation and day length (43). The pattern of monthly changes in solar insolation varies by latitude, with very little monthly change near the equator and larger monthly changes as one nears the poles. However, locations at the same latitude may have different patterns due to local conditions such as altitude, cloud cover, and proximity to bodies of water. All solar insolation data were obtained from the National Aeronautics and Space Administration 4 (NASA) Surface Meteorological and Solar Energy (SSE) database version 6.0, which is based on global data collected by satellite for 22 years between 1983 and 2005 (42). The average monthly solar insolation data are available with a spatial resolution of 1° 9 1° latitude/longitude. The actual onset locations were grouped into reference onset locations, which represent all locations in the 1° 9 1° grid of latitude and longitude. The number of reference onset locations from a collection site varied greatly, influenced by country size, migration patterns, and cultural factors. The reference onset locations were used in all analyses. All solar insolation data were shifted by 6 months for onset locations in the Southern Hemisphere to compare with data from the Northern Hemisphere. Solar insolation is an appropriate variable for a sample with multiple birth-cohorts as the incoming global average, annual mean solar insolation has remained essentially unchanged over the last 2000 years (44). The monthly change in solar insolation was calculated as the difference between the current month minus the previous month. The maximum monthly increase in solar insolation was defined as the largest monthly increase over the year. The interaction between the maximum monthly increase in solar insolation 9 family history and the maximum monthly increase in solar insolation 9 polarity of first episode were also analyzed. Birth-cohort The birth-cohort was included in all analyses as an older age of onset of bipolar disorder in older cohorts was reported in many studies (40, 41, 45– 47). Three birth-cohort groups were created, for those born before 1940, born between 1940 and 1959, and born 1960 or later, consistent with prior research (40, 41, 45–47). In this sample, 34.6% of the patients with bipolar I disorder were born before 1960. Country specific variables As the onset of bipolar disorder spans several decades, an older mean age of onset would be expected in a country with an older median age (48, 49), and country median age was included in the prior analyses. The country median age also provides information about country socioeconomic characteristics (50). However, in the current sample, there was a 31-year difference in median age between the oldest country (Japan 46.9 years) and the youngest (Uganda 15.7 years) and more socioeconomic variation among countries. Solar insolation and onset of bipolar disorder Additional socioeconomic measures were obtained to explain country specific differences in age of onset: physician density of any specialty per 1000 population, GDP per capita, total health expenditures as a per cent of GDP, and the Gini index of income inequality (50). Bipolar I disorder Only data from patients with a diagnosis of bipolar I disorder were included to be consistent with our prior studies, and because there was a large imbalance in the per cent of patients with a diagnosis of bipolar I disorder at the collection sites, varying from 99% to 23%. Also, there was a potential for bias related to age of onset for those who received a diagnosis before the criteria were expanded to include bipolar II disorder. Statistics Estimates of the effects of solar insolation on the age of onset were calculated using generalized estimating equations (GEE) to account for the correlated data and unbalanced number of data points within each onset location (cluster). The GEE model uses a population-based or marginal approach to estimate the effect across the entire population rather than within a cluster (51). An exchangeable correlation matrix was selected for the GEE models, which is appropriate for a large number of clusters including many with a single observation (52). Models were estimated for all patients, and after excluding the patients born before 1960 as in our prior studies. In all GEE models, the dependent variable was the age of onset. Sidak’s adjustment for multiple comparisons was used to make pair-wise comparisons between the birth-cohorts. A significance level of 0.01 was used for all evaluations. The corrected quasi-likelihood independence model criterion was used to assist with model evaluation (53). SPSS version 24 (IBM, Armonk, NY, USA) was used for all analyses. available for 5055 of the 5536 patients (91.3%), and of the 5055 patients, the polarity was mania in 2543 (50.3%) and depression in 2512 (49.7%). The unadjusted mean age of onset was 25.4  10.6 years. The onset of bipolar disorder for the 5536 patients occurred in 456 unique onset locations in 57 countries. The average number of patients in each onset location was 12.1, with 240 of the 5536 patients (4.3%) in an onset location of one. Of the 5536 patients, 4283 (77.4%) had onset in the Northern Hemisphere and 1253 (22.6%) in the Southern Hemisphere (Table 1). The number of patients, onset locations, and onset countries in this study was considerably larger than in our prior studies (Table 2). Solar insolation The largest maximum monthly increase in solar insolation occurred in the northern latitudes such as the Nordic countries, Russia, Estonia, and Canada, and in warm dry areas in Chile, USA, Mexico, Greece, and South Africa. The smallest changes occurred near the equator in Uganda, Colombia, Malaysia, and Brazil (Table 3). The maximum monthly increase in solar insolation occurred in springtime in both hemispheres. Table 1. Latitude of patient onset locations Degrees latitude (north and south)* Number of patients Number of onset locations 473 413 315 1957 1654 464 259 1 5536 30 54 34 120 157 46 14 1 456 0–9 10–19 20–29 30–39 40–49 50–59 60–69 70–79 Total *1253 in the Southern Hemisphere. Table 2. Number of patients by study Results Patients and onset locations Study date Data were collected for 7392 patients with bipolar disorder. Of these, 5536 had a diagnosis of bipolar I disorder and were included in the analysis. Of the 5536 patients, 3221 (58.2%) were female, and 2314 (41.8%) were male. Family history was available for 4698 of the 5536 patients (84.9%), and of the 4698 patients, 2567 (54.6%) had a positive family history. The polarity of the first episode was 2012 (40) 2014 (41) 2017 Per cent increase between 2012 and 2017 studies (%) Per cent increase between 2014 and 2017 studies (%) Number of patients with bipolar I disorder Number of onset locations Number of onset countries 2414 4037 5536 129 180 318 456 153 24 43 57 138 37 43 33 5 Bauer et al. Excluding the locations near the equator that have little change to solar insolation throughout the year, the maximum increase occurred between February and March at 41.1% of onset locations, between March and April at 35.4% of onset locations, and between April and May at 13.4% of onset locations. Model estimates The best model to assess the relation between solar insolation and the age of onset included the interaction of the maximum monthly increase in solar insolation 9 family history, the birth-cohort and the physician density, as shown in Table 4. There was a significant inverse relation between the Table 3. Some examples of the maximum monthly increase in solar insolation at onset locations Onset location Kampala, Uganda Medellín, Colombia Hong Kong Kuala Lumpur, Malaysia Salvador, Brazil Bangalore, India S~ao Paulo, Brazil Tokyo, Japan Singapore Hyderabad, India Mexico City, Mexico Boston, MA, USA Rochester, MN, USA Porto Alegre, Brazil Nova Scotia, Canada Adelaide, Australia Thessaloniki, Greece Tunis, Tunisia Melbourne, Australia Barcelona, Spain Paris, France Ankara, Turkey San Diego, CA, USA Buenos Aries, Argentina Cagliari, Sardinia, Italy Dresden, Germany Bordeaux, France Calgary, Canada Beer Sheva, Israel Valparaiso, Chile Tartu, Estonia Athens, Greece Los Angeles, CA, USA Santiago, Chile Helsinki, Finland Cape Town, South Africa Talca, Chile Oslo, Norway Khanti-Mansiysk, Russia Stockholm, Sweden Trondheim, Norway 6 Maximum monthly increase in solar insolation (kWh/m2/day) 0.3 0.3 0.6 0.6 0.6 0.7 0.7 0.7 0.7 0.8 0.9 1.0 1.0 1.0 1.1 1.1 1.1 1.1 1.1 1.2 1.2 1.2 1.2 1.2 1.3 1.3 1.3 1.4 1.4 1.4 1.4 1.5 1.5 1.5 1.5 1.5 1.6 1.6 1.6 1.6 1.7 Latitude 0.3 6.3 22.5 3.2 12.9 12.9 23.5 35.7 1.3 17.4 19.4 42.2 44.0 30.0 45.1 34.9 40.6 36.8 37.5 41.4 48.9 39.9 32.4 34.6 39.2 51.1 44.8 51.1 31.2 33.0 58.4 38.0 34.0 33.3 60.2 33.9 35.4 59.9 61.0 59.3 63.4 N N N N S N S N N N N N N S N S N N S N N N N S N N N N N S N N N S N S S N N N N maximum monthly increase in solar insolation and the age of onset, labeled Model 1. For every 0.1 kWh/m2/day increase in the maximum monthly increase in solar insolation, there was approximately a 0.57-year (6.8 months) decrease in the age of onset. Alternatively, comparing the largest (1.7) to the smallest (0.3) maximum monthly increase in solar insolation, Model 1 suggests an 8-year decrease in the age of onset (1.4 range in maximum monthly increase in solar insolation * 5.702 estimated coefficient). This effect was reduced by about 30% if there was no family history. The inverse relation was also found when including the interaction of the maximum monthly increase in solar insolation 9 polarity of first episode, the birth-cohort and the physician density, labeled Model 2, with the effect about 20% smaller for a first episode of mania (5.3 months). The results were similar when models 1 and 2 were estimated excluding the patients born before 1960 and the birth-cohort (Models 3 and 4 in Table 5). Of the 5536 patients, 287 (5.1%) were born before 1940, 1631 (29.5%) were born between 1940 and 1959, and 3618 (65.4%) were born in 1960 or later. The birth-cohort was significantly associated with age of onset (P < 0.001). In Model 1, compared to the youngest birth-cohort born in 1960 or later, those born before 1940 had an onset 15.7 years older, and those born between 1940 and 1959 had an onset 7.7 years older. The model estimates were improved when physician density was used to explain socioeconomic differences rather than the country median age (40, 41). Because the model has changed, these results cannot be directly compared with our prior studies. The other economic variables were not significant or the models were not as good. The collection site was considered to be an adequate proxy for the onset location at some sites: Barcelona, Cape Town, Helsinki, Melbourne/Geelong, Salvador, Stockholm/Gothenburg, and W€ urzburg. The best models were estimated excluding all data from these collection sites and results remained significant (P < 0.001). Compared with other solar insolation variables, models including the maximum monthly increase in solar insolation remained the best. Discussion Despite increasing the number of patients, onset locations, and countries, the findings from our prior studies were confirmed. The maximum monthly increase in solar insolation was inversely associated with the age of onset of bipolar disorder. The effect was reduced in those without a Solar insolation and onset of bipolar disorder Table 4. Estimated coefficients of parameters explaining age of onset of bipolar I disorder 99% Confidence interval Parameters Model 1* N = 4698 Maximum monthly increase in solar insolation No family history 9 maximum monthly increase in solar insolation Model 2† N = 5055 Maximum monthly increase in solar insolation First episode manic 9 maximum monthly increase in solar insolation Coefficient estimate Standard error Lower Upper Coefficient significance Wald chi-square P 5.702 1.733 0.932 0.247 8.102 1.097 3.301 2.369 37.423 49.288 <0.001 <0.001 5.491 1.037 1.008 0.269 8.088 0.345 2.895 1.729 29.685 14.901 <0.001 <0.001 *Dependent variable: Age of Onset. Model: intercept, physicians per 1000 onset country population, maximum monthly increase in solar insolation, no family history 9 maximum monthly increase in solar insolation and birth-cohort group. †Dependent variable: Age of Onset. Model: intercept, physicians per 1000 onset country population, maximum monthly increase in solar insolation, first episode manic 9 maximum monthly increase in solar insolation and birth-cohort group. Table 5. Estimated coefficients of parameters explaining age of onset for patients with bipolar I disorder born in 1960 or later 99% Confidence interval Parameters Coefficient estimate Model 3* N = 3101 Maximum monthly increase in solar insolation No family history 9 maximum monthly increase in solar insolation Model 4† N = 3308 Maximum monthly increase in solar insolation First episode manic 9 maximum monthly increase in solar insolation Standard error Lower Upper Coefficient significance Wald chi-square P 4.676 1.240 0.838 0.255 6.835 0.584 2.517 1.897 31.118 23.694 <0.001 <0.001 4.626 1.524 0.831 0.279 6.766 0.805 2.487 2.242 31.023 29.854 <0.001 <0.001 *Dependent variable: Age of Onset. Model: intercept, physicians per 1000 onset country population, maximum monthly increase in solar insolation, and no family history 9 maximum monthly increase in solar insolation. †Dependent variable: Age of Onset. Model: intercept physicians per 1000 onset country population, maximum monthly increase in solar insolation, and first episode manic 9 maximum monthly increase in solar insolation. family history of a mood disorder and with an onset of mania rather than depression. This confirmation using a larger and more diverse sample with collection and onset locations on six continents, and when only including the youngest birthcohort, suggests that these findings are not due to chance. As with the prior studies, the maximum monthly increase in solar insolation occurred in springtime. The effect was related to the size of the maximum monthly increase, regardless if starting from a low level of solar insolation at very northern latitude, or from a medium level at a mid-latitude desert. From a clinical perspective, physicians should recognize the potential for a younger age of onset in locations with a large increase in sunlight in springtime, and the potential for an older onset in areas with little seasonal change. The interaction with family history suggests that a genetic predisposition to bipolar disorder involves circadian dysregulation (18). The findings also emphasize the importance of obtaining a family history from all patients, especially as the age of onset was younger in the youngest birth-cohort. Early onset bipolar disorder is associated both with family history and with poorer outcomes (54–57). These models were different from our prior analyses due to the inclusion of physician density. Using data from such diverse countries, it is not surprising that having an important socioeconomic variable in addition to the maximum monthly increase in solar insolation improved the model for age of onset. The physician workforce directly impacts health outcomes (58, 59) and varies greatly between high- and low-income countries and within high-income regions, and especially for mental health (60, 61). Recent and remarkable societal changes relating to light exposure may be contributing to the increased vulnerability to a springtime circadian challenge in the youngest birth-cohort and may be of particular concern for future generations. These include the conversion from incandescent to LED (light-emitting diode) lighting, the use of mobile 7 Bauer et al. devices backlit with LEDs, and the rise of a 24-h society. The Nobel Prize in physics in 2014 was awarded to Isamu Akasaki, Hiroshi Amano, and Shuji Nakamura for the invention of the blue LED in the 1990s, which enabled a new method for semiconductors to create white light (62). LEDs are energy-efficient, requiring 85% less energy than incandescent light bulbs (63), and are long-lasting. Conversion to LEDs has been extremely rapid, with 69% of all light bulbs sold worldwide expected to be LEDs by 2020 (63). However, differences in the properties of LEDs may have physiological impact. Incandescent lights have a dominant wavelength of 574 nm, close to the peak sensitivity of photopic photoreception of 555 nm (64, 65). In contrast, white LEDs have a dominant wavelength of 482 nm, in the blue light range, close to the peak sensitivity for circadian photoreception (~480 nm for melanopsin and ~460 nm for melatonin suppression) (33, 64, 66). Compared to the past, many people today experience darker days and brighter nights (28). In the daytime, people are spending more time indoors and all indoor lighting (incandescent and LED) is vastly dimmer and has a different spectral pattern than the Sun. Unlike with vision, people are not aware if they do not receive enough daytime light for circadian needs (65). In the night-time, indoor lighting is a necessity, but long-term effects of exposure to LEDs after sunset on the circadian system are unknown. Sufficient darkness during the dark phase of the light–dark cycle is needed for melatonin secretion and optimal entrainment (67). Additionally, about 80% of the world’s population lives where the night sky brightness is above the threshold for light pollution (68), and high-intensity LED streetlights pose a health threat (69, 70). International lighting industry standards were optimized for vision. The complexity of non-visual physiological responses and the need to balance benefits and harmful effects pose many challenges for developing new standards (29, 70, 71). The intensity, spectrum, duration, and timing of all lighting, and prior light and dark exposure may all impact circadian entrainment (29,71–72). Digital devices, such as smartphones, eReaders, tablets, video games, computer screens, and TV sets, are backlit with LED light to enhance the daytime brightness and contrast (73, 74). While well suited for the screen size of mobile devices, LEDs emit bright blue light. In the USA, 72% of adolescents used a cell phone in the hour before bedtime (75), and many adolescents in Belgium used a cell phone after lights out (76). Recent investigations report an association between evening use of technology (smartphones, computers, 8 and eReaders) and reduced melatonin secretion in healthy adults and adolescents (77–79), and decreased sleep in children and adults (75, 77, 80, 81). In a systematic review of 67 international studies of children and adolescents, increasing screen time was associated with an adverse sleep outcome (82). The heavy use of mobile technology by children and adolescents is of particular concern. Early pubertal children have increased sensitivity to evening light as measured by melatonin suppression (83), and transmission of blue light in the young is much greater than in the old due to aging of the crystalline lens and loss of pupil area (65, 84). Daily patterns of light exposure reflect individual preferences and societal requirements (29, 35). Recently, competitiveness, consumer demand, and globalization are creating a 24-h society (85). The continuous provision of goods and services offers considerable customer convenience, but requires that many people work non-traditional and irregular work schedules, including late night and early morning. Yet people with bipolar disorder may be especially vulnerable to the circadian disruption experienced by shift workers (86, 87). The importance of regularity in daily patterns of exposure to circadian stimuli may be greatest for individuals who have suboptimal circadian function (88). The increase in irregular work schedules is occurring worldwide (89, 90), with 35% of employed persons in the USA having flexible work hours in 2012 (91). Limitations The process of data gathering was not standardized across all collection sites, although based on the DSM-IV criteria. Family history data were not validated and may be unreliable. There may be recall bias with self-reported age of onset in relation to episodes early in life or of less severity. The sample was not demographically representative of the country populations. However, the unadjusted mean age of onset in the sample of 25.4  10.6 years was similar to that in other international studies: 25.7 years for bipolar I disorder (54) and 25.6 years for any bipolar disorder (92). Although 79% of the samples were from the Northern Hemisphere, about 87.5% of the world’s population lives there (93). There was no individual data on sun exposure, sun-related behaviours, shift work, technology use, skin type, serum vitamin D levels (94), or retinal abnormalities (95, 96). There was also no individual data on perinatal light exposure, which may impact future circadian resilience (97, 98). Other functions stimulated by Solar insolation and onset of bipolar disorder blue light including direct enhancement of cognition (99) were not considered. Societal changes unrelated to light were not considered, including the tumultuous events of the twentieth century that impacted the older birth-cohorts (46). There was a downward bias in the age of onset of the youngest birth-cohort due to the absence of individuals with late onset bipolar disorder, ascertainment bias, and the potential for earlier mortality in those with early onset (48, 100, 101). Data from the Southern Hemisphere was shifted by 6 months, which discounts the cultural dimensions of seasonality. The model results show an association but cannot show causality. However, the impacts of circadian disruption on the course bipolar disorder needs to be studied, whether or not the association is causative (102). In conclusion, a large increase in the maximum monthly solar insolation in springtime may influence the onset of bipolar disorder, especially for patients with a family history of mood disorders. The larger the maximum monthly increase in solar insolation, the younger the onset of bipolar disorder. With the youngest age of onset in the youngest birth-cohort, recent societal changes that may impact adaptability to a circadian challenge need investigation. These include LED lighting, mobile technology backlit with LEDs, and 24-h lifestyles. While country conditions such as physician density are beyond the individual’s ability to control, individual behaviour directly impacts light exposure and darkness at night. Perhaps, treatment recommendations for bipolar disorder may include optimum daytime and night-time light exposure, including from technology. Acknowledgements Michael Berk is supported by an NHMRC Senior Principal Research Fellowship (1059660). Ole A Andreassen, Thomas DBjella and Ingrid Melle are supported by Research Council of Norway (223273) and KG Jebsen Stiftelsen. Ravi Nadella has received funding from the Accelerator program for Discovery in Brain disorders using Stem cells (ADBS), jointly funded by the Department of Biotechnology, Government of India, and the Pratiksha trust. Biju Viswanath has received funding by Department of Science and Technology INSPIRE scheme, Government of India. Mikael Landen was supported by grants from the Swedish Research Council (K2014-62X-14647-12-51 and K2010-61P-21568-01-4), the Swedish foundation for Strategic Research (KF10-0039), and the Swedish Federal Government under the LUA/ALF agreement (ALF 20130032, ALFGBG-142041). We thank Haydeh Olofsson for valuable data management support. Declarations of interest The authors declare that they have no conflict of interests. References 1. Abreu T, Bragancßa M. The bipolarity of light and dark: a review on bipolar disorder and circadian cycles. J Affect Disord 2015;185:219–229. 2. Bellivier F, Geoffroy PA, Etain B, Scott J. Sleep- and circadian rhythm-associated pathways as therapeutic targets in bipolar disorder. Expert Opin Ther Targets 2015;19:747–763. 3. Bengesser SA, Reininghaus EZ, Lackner N et al. Is the molecular clock ticking differently in bipolar disorder? Methylation analysis of the clock gene ARNTL. World J Biol Psychiatry 2016 https://doi.org/10.1080/ 15622975.2016.1231421. [Epub ahead of print]. 4. Bradley AJ, Webb-Mitchell R, Hazu A et al. Sleep and circadian rhythm disturbance in bipolar disorder. Psychol Med 2017;47:1678–1689. 5. Etain B, Jamain S, Milhiet V, et al. Association between circadian genes, bipolar disorders and chronotypes. Chronobiol Int 2014;31:807–814. 6. McCarthy MJ, Nievergelt CM, Kelsoe JR, Welsh DK. A survey of genomic studies supports association of circadian clock genes with bipolar disorder spectrum illnesses and lithium response. PLoS One 2012;7:e32091. 7. Seleem MA, Merranko JA, Goldstein TR et al. The longitudinal course of sleep timing and circadian preferences in adults with bipolar disorder. Bipolar Disord 2015;17: 392–402. 8. Frank E, Swartz HA, Kupfer DJ. Interpersonal and social rhythm therapy: managing the chaos of bipolar disorder. Biol Psychiatry 2000;48:593–604. 9. Melo MC, Abreu RL, Linhares Neto VB, De Bruin PF, De Bruin VM. Chronotype and circadian rhythm in bipolar disorder: a systematic review. Sleep Med Rev 2016. https://doi.org/10.1016/j.smrv.2016.06.007. pii: S10870792(16)30059-4. 10. Gonzalez R. The relationship between bipolar disorder and biological rhythms. J Clin Psychiatry 2014;75:e323– e331. 11. Mondin TC, Cardoso TA, Souza LD et al. Mood disorders and biological rhythms in young adults: a large population-based study. J Psychiatr Res 2017;84:98–104. 12. Ng TH, Chung KF, Ho FY, Yeung WF, Yung KP, Lam TH. Sleep-wake disturbance in interepisode bipolar disorder and high-risk individuals: a systematic review and meta-analysis. Sleep Med Rev 2015;20:46–58. 13. Takaesu Y, Inoue Y, Murakoshi A et al. Prevalence of circadian rhythm sleep-wake disorders and associated factors in euthymic patients with bipolar disorder. PLoS One 2016;11:e0159578. 14. Alloy LB, Boland EM, Ng TH, Whitehouse WG, Abramson LY. Low social rhythm regularity predicts first onset of bipolar spectrum disorders among at-risk individuals with reward hypersensitivity. J Abnorm Psychol 2015;124:944–952. 15. Melo MC, Garcia RF, Linhares Neto VB et al. Sleep and circadian alterations in people at risk for bipolar disorder: a systematic review. J Psychiatr Res 2016;83: 211–219. 16. Berk M, Dodd S, Hallam K, Berk L, Gleeson J, Henry M. Small shifts in diurnal rhythms are associated with an increase in suicide: the effect of daylight saving. Sleep Biol Rhythms 2008;6:22–25. 17. Kennedy KP, Cullen KR, Deyoung CG, Klimes-Dougan B. The genetics of early-onset bipolar disorder: a systematic review. J Affect Disord 2015;184:1–12. 9 Bauer et al. 18. Mansour HA, Monk TH, Nimgaonkar VL. Circadian genes and bipolar disorder. Ann Med 2005;37:196–205. 19. McCarthy MJ, Welsh DK. Cellular circadian clocks in mood disorders. J Biol Rhythms 2012;27:339–352. 20. Geoffroy PA, Etain B, Lajnef M, et al. Circadian genes and lithium response in bipolar disorders: associations with PPARGC1A (PGC-1a) and RORA. Genes Brain Behav 2016;15:660–668. 21. Geoffroy PA, Etain B. Lithium and circadian rhythms. In: Mahli GS, Masson M, Bellivier F, eds. The science and practice of lithium therapy. Switzerland: Springer International Publishing; 2017:111–124. 22. McCarthy MJ, Wei H, Landgraf D, Le Roux MJ, Welsh DK. Disinhibition of the extracellular-signal-regulated kinase restores the amplification of circadian rhythms by lithium in cells from bipolar disorder patients. Eur Neuropsychopharmacol 2016;26:1310–1319. 23. Yin L, Wang J, Klein PS, Lazar MA. Nuclear receptor Rev-erbalpha is a critical lithium-sensitive component of the circadian clock. Science 2006;311:1002–1005. 24. Geoffroy PA, Curis E, Courtin C et al. Lithium response in bipolar disorders and core clock genes expression. World J Biol Psychiatry 2017 Jan;18:1–34. 25. Dallaspezia S, Benedetti F. Melatonin, circadian rhythms, and the clock genes in bipolar disorder. Curr Psychiatry Rep 2009;11:488–493. 26. Wirz-Justice A, Benedetti F, Berger M et al. Chronotherapeutics (light and wake therapy) in affective disorders. Psychol Med 2005;35:939–944. 27. Henriksen TE, Skrede S, Fasmer OB et al. Blue-blocking glasses as additive treatment for mania: a randomized placebo-controlled trial. Bipolar Disord 2016;18:221–232. 28. Blask D, Brainard G, Gibbons R, Lockley S, Stevens R, Motta M. Report 4 of the council on science and public health (a-12) American Medical Association, 2012. Available at: http://ida.darksky.org/assets/documents/AMA_ 2012_report.pdf. Accessed 10, April 2017. 29. Lucas RJ, Peirson SN, Berson DM et al. Measuring and using light in the melanopsin age. Trends Neurosci 2014;37:1–9. 30. Reddy AB, O’Neill JS. Healthy clocks, healthy body, healthy mind. Trends Cell Biol 2010;20:36–44. 31. Silver R, Lesauter J. Circadian and homeostatic factors in arousal. Ann N Y Acad Sci 2008;1129:263–274. 32. Duffy JF, Czeisler CA. Effect of light on human circadian physiology. Sleep Med Clin 2009;4:165–177. 33. Berson DM, Dunn FA, Takao M. Phototransduction by retinal ganglion cells that set the circadian clock. Science 2002;295:1070–1073. 34. van Gelder RN. Vision: melanopsin and the pharmacology of photons. Curr Biol 2016;26:R804–R806. 35. Rosenwasser AM, Turek FW. Neurobiology of circadian rhythm regulation. Sleep Med Clin 2015;10:403–412. 36. Silver R, Kriegsfeld LJ. Circadian rhythms have broad implications for understanding brain and behavior. Eur J Neurosci 2014;39:1866–1880. 37. Hatori M, Panda S. The emerging roles of melanopsin in behavioral adaptation to light. Trends Mol Med 2010;16:435–446. 38. Childs B, Valle D. Genetics, biology and disease. Annu Rev Genomics Hum Genet 2000;1:1–9. 39. Halldorsdottir T, Binder EB. Gene 9 environment interactions: from molecular mechanisms to behavior. Annu Rev Psychol 2017;68:215–241. 40. Bauer M, Glenn T, Alda M et al. Impact of sunlight on the age of onset of bipolar disorder. Bipolar Disord 2012;14:654–663. 10 41. Bauer M, Glenn T, Alda M et al. Relationship between sunlight and the age of onset of bipolar disorder: an international multisite study. J Affect Disord 2014;167:104– 111. 42. NASA. NASA Langley Research Center atmospheric science data center surface meteorological and solar energy surface meteorology and solar energy (SSE) release 6.0 methodology, Version 3.1. 2012. Available at: http://eosweb.larc.nasa.gov/sse/. Accessed 10, April 2017. 43. NASA. NASA Earth Observatory. Climate and Earth’s energy budget. 2009. Available at: http://earthobservatory. nasa.gov/Features/EnergyBalance/page1.php. Accessed 10, April 2017. 44. NRC (National Research Council) Committee on Surface Temperature Reconstructions for the Last 2,000 Years. Surface temperature reconstructions for the last 2,000 years. National Academies Press, US, 2007. 45. Chengappa KN, Kupfer DJ, Frank E et al. Relationship of birth cohort and early age at onset of illness in a bipolar disorder case registry. Am J Psychiatry 2003;160:1636– 1642. 46. Bauer M, Glenn T, Alda M et al. Influence of birth cohort on age of onset cluster analysis in bipolar I disorder. Eur Psychiatry 2015;30:99–105. 47. Da Silva Magalh~ aes PV, Gomes FA, Kunz M, Kapczinski F. Birth-cohort and dual diagnosis effects on age-at-onset in Brazilian patients with bipolar I disorder. Acta Psychiatr Scand 2009;120:492–495. 48. Chen WJ, Faraone SV, Orav EJ, Tsuang MT. Estimating age at onset distributions: the bias from prevalent cases and its impact on risk estimation. Genet Epidemiol 1993;10:43–59. 49. Heimbuch RC, Matthysse S, Kidd KK. Estimating age-ofonset distributions for disorders with variable onset. Am J Hum Genet 1980;32:564–574. 50. CIA. World Factbook. Washington, DC. 2016. Available at: https://www.cia.gov/library/publications/the-worldfactbook/. Accessed 10, April 2017. 51. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121– 130. 52. Stedman MR, Gagnon DR, Lew RA, Solomon DH, Brookhart MA. An evaluation of statistical approaches for analyzing physician-randomized quality improvement interventions. Contemp Clin Trials 2008;29:687– 695. 53. Pan W. Akaike’s information criterion in generalized estimating equations. Biometrics 2001;57:120–125. 54. Baldessarini RJ, Tondo L, Vazquez GH et al. Age at onset versus family history and clinical outcomes in 1,665 international bipolar-I disorder patients. World Psychiatry 2012;11:40–46. 55. Post RM, Leverich GS, Kupka RW et al. Early-onset bipolar disorder and treatment delay are risk factors for poor outcome in adulthood. J Clin Psychiatry 2010;71:864–872. 56. Post RM, Leverich GS, Kupka R et al. Increased parental history of bipolar disorder in the United States: association with early age of onset. Acta Psychiatr Scand 2014;129:375–382. 57. Suominen K, Mantere O, Valtonen H et al. Early age at onset of bipolar disorder is associated with more severe clinical features but delayed treatment seeking. Bipolar Disord 2007;9:698–705. 58. Anand S, B€ arnighausen T. Human resources and health outcomes: cross-country econometric study. Lancet 2004;364:1603–1609. Solar insolation and onset of bipolar disorder 59. Cooper RA. States with more physicians have better-quality health care. Health Aff (Millwood) 2009;28:w91– w102. 60. Bruckner TA, Scheffler RM, Shen G et al. The mental health workforce gap in low-and middle-income countries: a needs-based approach. Bull World Health Organ 2011;89:184–194. 61. WHO. Health workforce requirements for universal health coverage and the Sustainable Development Goals. 2016. Available at: http://www.who.int/hrh/resources/ health-observer17/en/. Accessed 10, April 2017. 62. Nobel Prize. The Nobel Prize in physics 2014 - popular information. Available at: http://www.nobelprize.org/no bel_prizes/physics/laureates/2014/popular.html. Accessed 10, April 2017. 63. Goldman Sachs. The low carbon economy. 2015. Available at: http://www.goldmansachs.com/our-thinking/pa ges/report-the-low-carbon-economy.html. Accessed 10, April 2017. 64. Bellia L, Bisegna F, Spada G. Lighting in indoor environments: visual and non-visual effects of light sources with different spectral power distributions. Build Environ 2011;46:1984–1992. 65. Turner PL, Mainster MA. Circadian photoreception: ageing and the eye’s important role in systemic health. Br J Ophthalmol 2008;92:1439–1444. 66. Brainard GC, Hanifin JP, Greeson JM et al. Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor. J Neurosci 2001;21:6405– 6412. 67. Turner PL, Van Someren EJ, Mainster MA. The role of environmental light in sleep and health: effects of ocular aging and cataract surgery. Sleep Med Rev 2010;14:269–280. 68. Falchi F, Cinzano P, Duriscoe D et al. The new world atlas of artificial night sky brightness. Sci Adv 2016;2: e1600377. 69. Falchi F, Cinzano P, Elvidge CD, Keith DM, Haim A. Limiting the impact of light pollution on human health, environment and stellar visibility. J Environ Manage 2011;92:2714–2722. 70. AMA. (American Medical Association). AMA adopts guidance to reduce harm from high intensity street lights. 2016. Available at: https://www.ama-assn.org/ama-adoptsguidance-reduce-harm-high-intensity-street-lights. Accessed 10, April 2017. 71. Andersen M, Mardaljevic J, Lockley SW. A framework for predicting the non-visual effects of daylight–Part I: photobiology-based model. Light Res Technol 2012;44:37–53. 72. Pilorz V, Tam SK, Hughes S et al. Melanopsin regulates both sleep-promoting and arousal-promoting responses to light. PLoS Biol 2016;14:e1002482. 73. Gringras P, Middleton B, Skene DJ, Revell VL. Bigger, brighter, bluer-better? current light-emitting devices–adverse sleep properties and preventative strategies. Front Public Health 2015;3:233. 74. Tosini G, Ferguson I, Tsubota K. Effects of blue light on the circadian system and eye physiology. Mol Vis 2016;22:61–72. 75. Gradisar M, Wolfson AR, Harvey AG, Hale L, Rosenberg R, Czeisler CA. The sleep and technology use of Americans: findings from the National Sleep Foundation’s 2011 Sleep in America poll. J Clin Sleep Med 2013;9:1291– 1299. 76. Van den Bulck J. Adolescent use of mobile phones for calling and for sending text messages after lights out: 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. results from a prospective cohort study with a one-year follow-up. Sleep 2007;30:1220–1223. Chang AM, Aeschbach D, Duffy JF, Czeisler CA. Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. Proc Natl Acad Sci U S A 2015;112:1232–1237. Cajochen C, Frey S, Anders D et al. Evening exposure to a light-emitting diodes (LED)-backlit computer screen affects circadian physiology and cognitive performance. J Appl Physiol (1985) 2011;110:1432– 1438. Figueiro M, Overington D. Self-luminous devices and melatonin suppression in adolescents. Light Res Technol 2016;48:966–975. Carter B, Rees P, Hale L, Bhattacharjee D, Paradkar MS. Association between portable screen-based media device access or use and sleep outcomes: a systematic review and meta-analysis. JAMA Pediatr 2016;170:1202– 1208. Exelmans L, Van den Bulck J. Bedtime mobile phone use and sleep in adults. Soc Sci Med 2016;148:93–101. Hale L, Guan S. Screen time and sleep among schoolaged children and adolescents: a systematic literature review. Sleep Med Rev 2015;21:50–58. Crowley SJ, Cain SW, Burns AC, Acebo C, Carskadon MA. Increased sensitivity of the circadian system to light in early/mid-puberty. J Clin Endocrinol Metab 2015;100:4067–4073. Kessel L, Lundeman JH, Herbst K, Andersen TV, Larsen M. Age-related changes in the transmission properties of the human lens and their relevance to circadian entrainment. J Cataract Refract Surg 2010;36:308– 312. Rajaratnam SM, Arendt J. Health in a 24-h society. Lancet 2001;358:999–1005. Gamble KL, Motsinger-Reif AA, Hida A et al. Shift work in nurses: contribution of phenotypes and genotypes to adaptation. PLoS One 2011;6:e18395. Meyrer R, Demling J, Kornhuber J, Nowak M. Effects of night shifts in bipolar disorders and extreme morningness. Bipolar Disord 2009;11:897–899. van Someren EJ, Riemersma-Van DER, Lek RF. Live to the rhythm, slave to the rhythm. Sleep Med Rev 2007;11:465–484. Kuroda S, Yamamoto I. The era of the 24-hour society?: assessing changes in work timing using a Japanese time use survey. Appl Econ Lett 2012;19:1035–1038. UK POST (Parliamentary Office of Science and Technology). 24-Hour Society, November 2005. Available at: http://researchbriefings.parliament.uk/Re searchBriefing/Summary/POST-PN-250. Accessed 10, April 2017. BLS (Bureau of Labor Statistics, US). Economic News Release. Employed persons with flexible work hours by age, sex, educational attainment, and disability status, May 2012. Available at: https://www.bls.gov/news.relea se/dissup.t12.htm. Accessed 10, April 2017. Morselli PL, Elgie R, GAMIAN-Europe. GAMIANEurope/BEAM survey I–global analysis of a patient questionnaire circulated to 3450 members of 12 European advocacy groups operating in the field of mood disorders. Bipolar Disord 2003;5:265–278. Kummu M, Varis O. The world by latitudes: a global analysis of human population, development level and environment across the north–south axis over the past half century. Appl Geogr 2011;31:495–507. 11 Bauer et al. 94. Berk M, Jacka FN, Williams LJ, Ng F, Dodd S, Pasco JA. Is this D vitamin to worry about? Vitamin D insufficiency in an inpatient sample. Aust N Z J Psychiatry 2008;42:874–878. 95. Fountoulakis KN. Retinal response anomalies in patients with mental illness and high risk relatives. Biol Psychiatry 2010;68:e3. 96. Kalenderoglu A, Sevgi-Karadag A, Celik M, Egilmez OB, Han-Almis B, Ozen ME. Can the retinal ganglion cell layer (GCL) volume be a new marker to detect neurodegeneration in bipolar disorder? Compr Psychiatry 2016;67:66–72. 97. Bauer M, Glenn T, Alda M et al. Influence of light exposure during early life on the age of onset of bipolar disorder. J Psychiatr Res 2015;64:1–8. 12 98. Ciarleglio CM, Axley JC, Strauss BR, Gamble KL, McMahon DG. Perinatal photoperiod imprints the circadian clock. Nat Neurosci 2011;14:25–27. 99. Vandewalle G, Collignon O, Hull JT et al. Blue light stimulates cognitive brain activity in visually blind individuals. J Cogn Neurosci 2013;25:2072–2085. € 100. Osby U, Brandt L, Correia N, Ekbom A, Sparen P. Excess mortality in bipolar and unipolar disorder in Sweden. Arch Gen Psychiatry 2001;58:844–850. 101. Tsai SY, Kuo CJ, Chen CC, Lee HC. Risk factors for completed suicide in bipolar disorder. J Clin Psychiatry 2002;63:469–476. 102. Bechtel W. Circadian rhythms and mood disorders: are the phenomena and mechanisms causally related? Front Psychiatry 2015;6:118.