Rates of acute intracerebral hemorrhage (ICH) increase in winter months but the magnitude of risk is unknown. We aimed to quantify the association of ambient temperature with the risk of ICH in the Intensive Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial (INTERACT2) participants on an hourly timescale.
INTERACT2 was an international, open, blinded endpoint, randomized controlled trial of patients with spontaneous ICH (<6h of onset) and elevated systolic blood pressure (SBP, 150–220 mmHg) assigned to intensive (target SBP <140 mmHg) or guideline-recommended (SBP <180 mmHg) BP treatment. We linked individual level hourly temperature to baseline data of 1997 participants, and performed case-crossover analyses using a distributed lag non-linear model with 24h lag period to assess the association of ambient temperature and risk of ICH. Results were presented as overall cumulative odds ratios (ORs) and 95% CI.
Low ambient temperature (≤10°C) was associated with increased risks of ICH: overall cumulative OR was 1.37 (0.99–1.91) for 10°C, 1.92 (1.31–2.81) for 0°C, 3.13 (1.89–5.19) for -10°C, and 5.76 (2.30–14.42) for -20°C, as compared with a reference temperature of 20°C.There was no clear relation of low temperature beyond three hours after exposure. Results were consistent in sensitivity analyses.
Citation: Zheng D, Arima H, Sato S, Gasparrini A, Heeley E, Delcourt C, et al. (2016) Low Ambient Temperature and Intracerebral Hemorrhage: The INTERACT2 Study. PLoS ONE 11(2): e0149040. https://doi.org/10.1371/journal.pone.0149040
Editor: Xiaoying Wang, Massachusetts General Hospital, UNITED STATES
Received: December 1, 2015; Accepted: January 26, 2016; Published: February 9, 2016
Copyright: © 2016 Zheng et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: An anonymized version of the dataset used for the analysis in this article will be made available to researchers upon request through The George Institute for Global Health research office. Requests for the dataset can be made by contacting Mr Peter Dolnik, Director of Research Services, The George Institute for Global Health, GPO Box 5389, Sydney NSW 2001 Australia. T: +61 2 9657 0369 E: email@example.com
Funding: The second Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage Trial study was supported by Program (571281) and Project (512402 and 1004170) grants from the National Health and Medical Research Council (NHMRC) of Australia (URL: https://www.nhmrc.gov.au). AG was supported by Medical Research Council-UK (Grant ID: MR/M022625/1) (URL:https://www.mrc.ac.uk/). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Acute stroke due to spontaneous (non-traumatic) intracerebral hemorrhage (ICH) is a major global health issue which causes death and permanent disability in several million people worldwide each year . While marked seasonal and temporal patterns to the occurrence of ICH are recognised, with peak incidence in the winter and an association with cold temperature [2–5], this information is primarily derived from daily meteorological data without consideration of ambient temperature close to the time of onset of ICH. Moreover, the time lag method, which has shown an association of ambient temperature with acute myocardial infarction [6,7], has not been used in the study of ICH where influences on fluctuations in blood pressure (BP) appear especially important . Thus, the association of ambient temperature with ICH onset has not been well quantified on an hourly time-scale. Such information could help our understanding of the determinants of ICH in high risk populations that have marked seasonal and geographical trends in rates , and guide public health strategies to optimize preventative strategies in winter.
The primary aim of this study was to quantify the transient increase in risk of ICH associated with declining ambient temperature at an hourly resolution. Secondary aims were to determine the time sequence of low temperature triggering effects on ICH occurrence and whether there is any variation in risks across several pre-defined patient characteristics.
Materials and Methods
The INTERACT 2 study protocol was approved by the ethics committees for each centers and written informed consent was obtained from all patients or relevant surrogates. A full list of centers that participated in the trial is shown in Acknowledgments. Patient records/information was anonymized and de-identified prior to analysis.
We conducted post-hoc analysis using the baseline patient data from the second (main phase) Intensive Blood Pressure Reduction in Acute Cerebral Haemorrhage Trial (INTERACT2). The design and main results of INTERACT2 have been described elsewhere.[10,11]. In brief, INTERACT2 was an international, multicenter, open, blinded endpoint, randomized controlled trial . A total of 2839 patients with CT-confirmed spontaneous ICH, elevated systolic BP (SBP, 150-220mmHg), and capacity to commence BP lowering treatment (<6h of onset assessed by ‘time last seen normal’), were enrolled from 144 hospitals in 21 countries between 7th October 2008 and 30th August 2012. Patients were excluded if they had a structural cerebral cause for the ICH, deep coma (defined as a score of 3–5 on the Glasgow coma scale [GCS] , where scores range from 3 to 15, with lower scores indicating reduced levels of consciousness), massive hematoma with a poor prognosis, or if early surgery to evacuate the hematoma was planned. Eligible patients were randomly assigned to receive intensive (target SBP <140mmHg within 1h) or guideline recommended (SBP <180mmHg) BP treatment. In order to minimize assessor bias, the primary outcome of ‘death or major disability’ (as measured by the modified Rankin scale) at 90 days was evaluated by clinicians who were blind to the randomized treatment. The date of the final patient follow up was 31 December 2012.
We obtained hourly ambient temperature data primarily from Weatherbank Inc (Edmond, US) which directly sources weather observations from worldwide government agencies that represent the official source of data in different countries. Data from the Australian Bureau of Meteorology was also used. For the main analyses, we linked baseline data of 1997 individual patients (70%) to hourly temperature data from monitoring stations less than 100km from the patients’ hospital of enrolment.
Baseline characteristics of the included patients were compared to those excluded in models using chi-square or Wilcoxon rank sum tests. We applied a time-stratified, case-crossover analysis to the data with the preceding 24h period of the reported ICH onset time selected as the case period for each ICH event [14–16]. Control periods of exposure were defined as the comparable 24h before the ICH in the other weeks of the same month and year. This method of selecting control periods has been shown to effectively control for time invariant confounders and seasonality, whilst also avoiding the problems of time trend and overlap bias associated with other methods of referent selection [16,17]. Thus, each participant was represented with a matched set of data for 1 case period and 3–4 control periods.
We assessed the risk of ICH in a distributed lag non-linear model (DLNM) which was originally developed to estimate the non-linear and delayed effects of temperature (or air pollution) on mortality or morbidity [18,19]. We applied recent developments in the method that extended DLNMs beyond time series data, and made it applicable in a case-crossover study design . The DLNM involves a bi-dimensional space of functions that describes simultaneously the shape of the relationship along both the space of the predictor and the lag dimension of its occurrence. We modeled the exposure-response function with a natural cubic spline with internal knots placed at quartiles of ambient temperature and a reference temperature of 20°C (taken as the average optimal temperature of comfort for humans) , and the lag-response function was modeled with a natural cubic spline with two equally spaced internal knots in log-scale. A bi-dimensional plot was constructed to demonstrate the entire relationship between temperature and ICH risk whilst also taking into account the time sequence of effect. Lag-response curves were plotted to describe the evolving temporal change in ICH risk in a 24 h lag period after exposure to specific temperatures in comparison to a reference optimal temperature of 20°C. Furthermore, the effect of temperature was summarized in an overall cumulative exposure-response figure which shows the net increased ICH risk over an entire lag period of 24 h in association with temperature exposure, accounting for any harvesting or lagged effects. Finally, our results were also presented as overall cumulative odds ratios (ORs) and 95% confidence intervals (CI) that were computed by summation of the lag-specific risk contributions for a given pattern of temperature exposures during the 24 hours lag period. The statistical equation describing the DLNM model and derivation of the overall cumulative odds ratio is outlined elsewhere .
In order to check the robustness of the study results, several sensitivity analyses were performed; by inclusion of all patients with known hourly temperature data within 200km of the site of enrolment, which totalled 2346 (83%); and by extending the maximum lag length from 24 to 72 h. All analyses were performed using SAS 9.2 (SAS Institute Inc., Cary, NC) and R project for statistical computing V.3.1.0 using the ‘dlnm’ and ‘survival’ packages.
Table 1 shows the baseline demographic and clinical characteristics of patients with available ambient temperature data (‘included patients’, n = 1997) and those without such data (‘excluded patients’, n = 832). Those without this data were significantly younger, had less prior use of antithrombotic therapy, higher diastolic BP, and more likely to have lobar ICH and larger hematoma volumes. The weather parameters of the 79 cities included in analyses are outlined in the S1 Table.
Fig 1 shows the bi-dimensional exposure-lag-response surface, depicting changes across temperature and lags with a trend of higher ORs at low ambient temperatures and shorter time lags. The lag-exposure plots in Fig 2 indicates an elevated risk of ICH in association with exposure to low ambient temperature (≤10°C) between lag 0 to lag 3 and that a harvesting effect seems to exist at longer lags. The main results of our study are shown in Fig 3, which reflects the relationship between low temperature exposure and ICH risk. A total number of 1568 (79%) ICH cases occurred below the optimal reference temperature. The overall cumulative ORs for ICH progressively increased from 1.37 (0.99–1.91) for 10°C, 1.92 (1.31–2.81) for 0°C, 3.13 (1.89–5.19) for -10°C, and 5.76 (2.30–14.42) for -20°C.
The exposure-response function was modeled with a natural cubic spline with internal knots placed at quartiles of ambient temperature (and a reference temperature of 20°C), and the lag-response function was modeled with a natural cubic spline with two equally spaced internal knots in log-scale.
CI indicates confidence interval. Case (n) indicates the number of ICH events in each temperature exposure category.
Fig 4 shows a temporal pattern to the occurrence of ICH, with peak frequencies in the early morning (06:00–10:00) and evening (20:00–23:00), but no clear relationship with the level of SBP among patients at the time of entry into the trial.
ICH indicates intracerebral hemorrhage; SBP, systolic blood pressure.
Results of sensitivity analyses including all patients with known hourly temperature data within 200km of the site of enrolment and extending the maximum lag length from 24 to 72h all produced similar results (S1–S3 Figs).
To our knowledge, this is the first large-scale international study to have investigated the association between ambient temperature and ICH risk at an hourly resolution and to capture the time sequence of the detrimental effect of low ambient temperature. Using distributed lag non-linear model fitted to a case-crossover model involving 1997 well-characterized ICH patients, we found a clear association of low ambient temperature with ICH occurrence within a few hours of exposure.
Several but not all studies have reported an association between low ambient temperature and ICH [2–5,22–24]. This discrepancy may be due to the use of daily temperature parameters that may not be truly representative of the temperature at the time of ICH onset. Furthermore, previous studies involved subjects from single regions, some of which have relatively small variations in temperature throughout the year. Our study using hourly temperature data in a large international study population corroborates the findings of a Taiwanese study with known hourly temperature data at ICH onset which showed a trend of increased ICH occurrence towards the lower temperature and is also consistent with prior investigations and clinical observation of a peak ICH occurrence in the winter due in part to the trigger effect of low ambient temperature on the risk of ICH onset at particular times of the day [2–5,25]. Mechanisms underlying the increased ICH risk in colder climate have not been resolved but exposure to low temperatures has been shown to stress the sympathetic nervous system leading to acute increases in heart rate, peripheral resistance, and thereby BP [26,27]. Although acute increases in BP associated with exposure to low ambient temperature is the most plausible trigger of ICH, we were unable to find any correlation between admission SBP level and ambient temperature, nor any differences in risks between those with and without a history of hypertension. This could be because the admission systolic BP of participants may not be representative of their BP level close to the time of ICH onset, or that ICH patients with non-elevated BP were excluded from our study.
Previous studies have reported upon a temperature effect for up to 14 days for the occurrence of myocardial infarction [6,7]. However, our study results show no association of low temperature beyond several hours prior to ICH onset and this finding is further supported by sensitivity analysis with extended lag period of 72 h. The differences in these risk time periods may be attributable to the distinctions in pathological mechanisms of onset for myocardial infarction (being plaque rupture from atherothrombosis) and ICH (being mainly due to the rupture of small deep perforating intracerebral vessels).
Key strengths of this study include the large sample of a wide range of patients with early and rigorous standardised evaluations of clinical and imaging findings after acute ICH. Also we were able to attain reliable hourly temperature weather data and link this information to the baseline patient characteristics to conduct precise evaluations on the effects of low ambient temperature. Our case-crossover model with control periods stratified by calendar month inherently adjusted for potential confounders, such as slow time varying individual characteristics, time of the day, day of the week, and seasonality. Furthermore, we were also able to avoid problems of time trend and overlap bias associated with other methods of control selection [16,17].
There are several limitations to our study. First, the study population is subject to varying levels of selection bias that may limit the generalizability of the results. Being a clinical trial population, patients with a poor prognosis due to massive hematoma or deep coma, and in whom early surgery was planned, were excluded. Therefore, future studies on a general ICH population are necessary to validate our findings. Another issue is that we were unable to obtain (≤100km) hourly temperature data for 30% of participants (‘excluded patients’), of which the majority were from second to third tier cities in China (Baotou, Xuzhou, Wenzhou and Xining). Differences in the provision of healthcare and medical development, and diet and lifestyles, may have resulted in significant differences in characteristics such as ‘time from onset to randomization’ and use of medications (e.g. antihypertensives and antithrombotics) and contribute to variability in severity of hospital presentation and prognosis, thus limiting the generalizability of our findings. However, sensitivity analyses that included approximated temperature of Baotou (83%) showed a slight increase in ICH risk in association with low temperature in comparison to the primary analyses. A further issue relates to the precision of the exposure temperature data, as some of our weather data for the main analyses originated from monitoring stations up to 100 kilometres away from participating hospitals. However, this may likely have generated a non-differential misclassification that biased the result toward the null, and therefore produced conservative estimates of the true risk. We were also unable to account for potential confounding from individual changes in behaviour and cold protection. We may have been able to partly mitigate this problem through use of the case-crossover design and control selection strategy as the pattern of activity and clothing habits of individuals are likely to remain consistent at across times of the day, days of the week, and within a one-month calendar period. It is also interesting to note that a Eurowinter group study reported no associations between indoor heating, clothing protection, and even sweating, with reductions in cerebrovascular disease related mortality . While we were unable to collect information on other potential confounders, such as the patients’ emotional state that could have a bearing on BP variability, the presence of viral infections, usage of sympathomimetics and of other weather parameters or air pollution, these are also assumed to be constant in case-crossover analysis. Finally, as the study was post-hoc, we cannot exclude the play of chance and reliably establish a causal relationship between low ambient temperature and ICH occurrence.
In conclusion, the present findings based on 1997 patients from a large global study show an increased risk of ICH within a few hours of exposure to very low ambient temperature. The risk of ICH in high-risk subjects might be reduced by more stringent monitoring and management of BP levels during cold seasons, targeted personal advice and environmental heating interventions, triggered by forecasts of very low temperature.
S1 Fig. Overall cumulative odds ratio plot for temperature effects using approximated weather data.
CI indicates confidence interval.
S2 Fig. Overall cumulative odds ratio plot for temperature effects with extended 72h lag period.
CI indicates confidence interval.
S3 Fig. Lag-exposure plots of intracerebral hemorrhage for specific temperatures with extended 72h lag period.
CI indicates confidence interval.
Executive Committee: C.S. Anderson (principal investigator), J. Chalmers (chair), H. Arima, S. Davis, E. Heeley, Y. Huang, P. Lavados, B. Neal, M.W. Parsons, R. Lindley, L. Morgenstern, T. Robinson, C. Stapf, C. Tzourio, J.G. Wang. National Leaders: China—Steering Committee: Y. Huang (chair), S. Chen, X.Y. Chen, L. Cui, Z. Liu, C. Lu, J. Wang, S. Wu, E. Xu, Q. Yang, C. Zhang, J. Zhang. Europe—Austria: R. Beer, E. Schmutzhard; Belgium: P. Redondo; Finland: M. Kaste, L. Soinne, T. Tatlisumak; France: C. Stapf; Germany: K. Wartenberg; Italy: S. Ricci; The Netherlands: K. Klijn; Portugal: E. Azevedo; Spain: A. Chamorro; Switzerland: M. Arnold, U. Fischer; India: S. Kaul, J. Pandian, H. Boyini, S. Singh; North America–A.A. Rabinstein; South America—Argentina–C. Estol; Brazil–G. Silva; Chile–P. Lavados, V.V. Olavarria; and United Kingdom–T.G. Robinson. Data Safety and Monitoring Committee: R.J. Simes (chair), M.-G. Bousser, G. Hankey, K. Jamrozik (deceased), S.C. Johnston, and S. Li. Project Office Operations Committee: E. Heeley (study director), C.S. Anderson, K. Bailey, J. Chalmers, T. Cheung, C. Delcourt, S. Chintapatla, E. Ducasse, T. Erho, J. Hata, B. Holder, E. Knight, R. Lindley, M. Leroux, T. Sassé, E. Odgers, R. Walsh, and Z. Wolfowicz. Endpoint Adjudication Committee: C.S. Anderson, G. Chen, C. Delcourt, S. Fuentes, R. Lindley, B. Peng, H.-M. Schneble, and M.-X. Wang. Statistical Analysis: H. Arima, L. Billot, S. Heritier, Q. Li, and M. Woodward. CT Analyses: C. Delcourt (chair), S. Abimbola, S. Anderson, E. Chan, G. Cheng, P. Chmielnik, J. Hata, S. Leighton, J.-Y. Liu, B. Rasmussen, A. Saxena, and S. Tripathy. Data Management and Programming: M. Armenis, M.A. Baig, B. Naidu, G. Starzec, and S. Steley. Coordinating Centers: International (The George Institute for Global Health, Sydney, Australia)–C.S. Anderson, E. Heeley, M. Leroux, C. Delcourt, T. Sassé, E. Knight, K. Bailey, T. Cheung, E. Odgers, E. Ducasse, B. Holder, Z. Wolfowicz, R. Walsh, S. Chintapatla, T. Erho; Argentina, Buenos Aires (STAT Research)–C. Estol, A. Moles, A. Ruiz, M. Zimmermann; Brazil, Fortaleza (Medicamenta MRS)–J. Marinho, S. Alves, R. Angelim, J. Araujo, L. Kawakami; Chile, Santiago (Clínica Alemana, Universidad del Desarrollo)–P. Lavados, V.V. Olavarria, C. Bustos, F. Gonzalez, P. Munoz Venturelli; China, Beijing (The George Institute China incorporating George Clinical, and Peking University First Hospital)–Y. Huang, X. Chen, Y. Huang, R. Jia, N. Li, S. Qu, Y. Shu, A. Song, J. Sun, J. Xiao, and Y. Zhao; China, Shanghai (The Centre for Epidemiological Studies and Clinical Trials, The Shanghai Institute of Hypertension, Rui Jin Hospital, Shanghai Jiaotong University School of Medicine)—J.G. Wang, Q. Huang; Europe, Paris (Unité de Recherche Clinique, APHP—Hôpital Lariboisière)–C. Stapf, E. Vicaut, A. Chamam, M.-C. Viaud, C. Dert, U. Fiedler, V. Jovis, S. Kabla, S. Marchand, A. Pena, V. Rochaud; India, Hyderabad (The George Institute India)–K. Mallikarjuna H. Boyini, N. Hasan; Norway, Oslo (Oslo University Hospital)–E. Berge, E.C. Sandset, A.S. Forårsveen; United Kingdom (Department of Cardiovascular Sciences, University of Leicester)–T. Robinson, D. Richardson, T. Kumar, S. Lewin; United Kingdom (London, Imperial Clinical Trials Unit)—N. Poulter, J. Field, A. Anjum, A. Wilson. Principal Investigators and Coordinators (according to country and center): Argentina—Clínica Instituto Medico Adrogue: H. Perelmuter, A.M. Agarie; Hospital Central de Mendoza: A.G. Barboza, L.A. Recchia, I.F. Miranda, S.G. Rauek, R.J. Duplessis; Australia—Austin Hospital: H. Dewey, L. Walker, S. Petrolo; Box Hill: C. Bladin; Gosford Hospital: J. Sturm, D. Crimmins, D. Griffiths, A. Schutz, V. Zenteno; John Hunter Hospital: M.W. Parsons, F. Miteff, N. Spratt, E. Kerr, C.R. Levi; Monash Medical Centre: T.G. Phan, H. Ma, L. Sanders, C. Moran, K. Wong; Royal Brisbane and Women's Hospital: S. Read, R. Henderson, A. Wong, R. Hull, G. Skinner; Royal Melbourne Hospital: S. Davis, P. Hand, B. Yan, H. Tu, B. Campbell; Royal Prince Alfred Hospital): C.S. Anderson, C. Delcourt; Sir Charles Gairdner Hospital: D.J. Blacker; Western Hospital: T. Wijeratne, M. Pathirage, M. Jasinararchchi, Z. Matkovic, S. Celestino; Austria—Allgenmeines Krankenhaus Linz: F. Gruber, M.R. Vosko, E. Diabl, S. Rathmaier; Innsbruck Medical University—Department of Neurology: R. Beer, E. Schmutzhard, B. Pfausler, R. Helbok; Medical University of Graz, Department of Neurology: F. Fazekas, R. Fischer, B. Poltrum, B. Zechner, U. Trummer; Belgium—Cliniques De L'Europe (Europe Clinic): M.P. Rutgers; UCL St Luc: A. Peeters, A. Dusart, M.-C. Duray, C. Parmentier, S. Ferrao-Santos; Universitair Ziekenhuis Brussel: R. Brouns, S. De Raedt, A. De Smedt, R.-J. VanHooff, J. De Keyser; Brazil—Hospital das Clínicas de Porto Alegre: S.C.O. Martins, A.G. de Almeida, R. Broudani, N.F. Titton; Hospital Quinta D'Or: G.R. de Freitas, F.M. Cardoso, L.M. Giesel, N.A. Lima Junior; Hospital Santa Marcelina: A.C. Ferraz de Almeida, R.B. Gomes, T.S. Borges dos Santos, E.M. Veloso Soares, O.L.A. Neto; Universidade Federal de São Paulo: G.S. Silva, D.L. Gomes, F.A. de Carvalho, M. Miranda, A. Marques; Universidade Federal do Paraná: V.F. Zétola, G. de Matia, M.C. Lange; Chile—Clinica Alemana de Santiago: J. Montes, A. Reccius, P. Munoz Venturelli, V.V. Olavarria, A. Soto; Clínica Alemana de Temuco, Chile: R. Rivas, C. Klapp; Clínica Dávila: S. Illanes, C. Aguilera, A. Castro; Complejo Asistencial Dr. Víctor Ríos Ruiz: C. Figueroa, J. Benavides, P. Salamanca, M.C. Concha, J. Pajarito; Hospital Naval Almirante Nef: P. Araya, F. Guerra; China—Baotou Central Hospital: Y. Li, G. Liu, B. Wang, J. Zhang, Y. Chong; Beijing Shijitan Hospital: M. He, L. Wang, J. Liu; Beijing Tongren Hospital: X. Zhang, C. Lai, H. Jiang, Q. Yang, S. Cui; Chang Ning District Central Hospital: Q. Tao, Y. Zhang, S. Yao, M. Xu, Y. Zhang; Changsha Central Hospital: Z. Liu, H. Xiao, J. Hu, J. Tang; Gongli Hospital, Pudong New Area, Shanghai: J. Sun, H. Ji, M. Jiang; Haidian Hospital, Beijing: F. Yu, Y. Zhang, X. Yang, X. Guo; Hejian City People's Hospital: Y. Wang, L. Wu, Z. Liu, Y. Gao, D. Sun; Hunan Province Brain Hospital: X. Huang, Y. Wang, L. Liu, Y. Li, P. Li; Jiangsu Province Hospital of Traditional Chinese Medicine: Y. Jiang, H. Li, H. Lu; Nanjing First Hospital: J. Zhou, C. Yuan; Navy General Hospital: X. Qi, F. Qiu, H. Qian, W. Wang, J. Liu; Peking University First Hospital: Y. Huang, W. Sun, F. Li, R. Liu, Q. Peng; Peking University Shougang Hospital: Z. Ren, C. Fan, Y. Zhang, H. Wang, T. Wang; People's Hospital of Beijing Daxing District: F. Shi, C. Duan, S. Chen, J. Wang, Z. Chen; Pinggu County Hospital, Beijing: X. Tan, Z. Zhao, Y. Gao, J. Chen, T. Han; Qinghai Province People's Hospital: S. Wu, L. Zhang, L. Wang, Q. Hu, Q. Hou; Qinghai University Affiliated Hospital: X. Zhao, L. Wang, G. Zeng, L. Ma, F. Wang; Ruijin Hospital Affiliated to Shanghai Jiaotong University School of Medicine: S. Chen, L. Zeng, Z. Guo, Y. Fu, Y. Song; Second Hospital of Hebei Medical University: L. Tai, X. Liu, X. Su, Y. Yang, R. Dong; Shijiazhuang 260 Hospital: Y. Xu, S. Tian, S. Cheng, L. Su, X. Xie; The Affiliated Hospital of Xuzhou Medical College: T. Xu, D. Geng, X. Yan, H. Fan, N. Zhao; The Branch Hospital of the First People's Hospital: S. Wang, J. Yang; The Chinese PLA No. 263 Hospital: J. Zhang, M. Yan, L. Li; The Fifth Affiliated Hospital Sun Yat-Sen University: Z. Li, X. Xu, F. Wang; The First Affiliated Hospital of Baotou Medical College: L. Wu, X. Guo, Y. Lian, H. Sun, D. Liu; The First Affiliated Hospital of Fujian Medical University: N. Wang, Q. Tang; The First Affiliated Hospital of Wenzhou Medical College: Z. Han, L. Feng; The Fourth Hospital of Jilin University: Y. Cui, J. Tian, H. Chang, X. Sun, J. Wang; The Second Affiliated Hospital Suzhou University: C. Liu, Z. Wen; The Second Affiliated Hospital of Guangzhou Medical College: E. Xu, Q. Lin; The Second Affiliated Hospital of Wenzhou Medical College: X. Zhang, L. Sun, B. Hu, M. Zou, Q. Bao; The Second Hospital of Qinghuangdao: X. Lin, L. Zhao, X. Tian, H. Wang, X. Wang; The Second Hospital of Tianjin Medical University: X. Li, L. Hao, Y. Duan, R. Wang, Z. Wei; Third Hospital of Hebei Medical University: J. Liu, S. Ren, H. Ren, Y. Wang, Y. Dong; Tianjin Medical University General Hospital: Y. Cheng, M. Zou, W. Liu, J. Han, C. Zhang; Tianjin Third Central Hospital: Z. Zhang, J. Zhu, Y. Wang, Q. Li; Traditional Chinese Medicine Hospital, Zhangjiagang: J. Qian, Y. Sun, K. Liu, F. Long; Wangcheng County People's Hospital of Hunan Province: X. Peng, Q. Zhang, Z. Yuan, C. Wang, M. Huang; Wuxi People's Hospital: J. Zhang, F. Wang, P. He, Y. You, X. Wang; Xiangya Hospital Central-South University: Q. Yang, H. Wang, J. Xia, L. Zhou, Y. Hou; Xining First People's Hospital: Y. Wang, L. Liu, Y. Qi, L. Mei, R. Lu; Xuzhou Central Hospital: G. Chen, L. Liu, L. Ping, W. Liu, S. Zhou; Yutian County Hospital, Hebei Province: J. Wang, L. Wang, H. Li, S. Zhang, L. Wang; Zengcheng People's Hospital: R. Zou, J. Guo, M. Li, W. Wei; Finland—Helsinki University Central Hospital: L. Soinne, S. Curtze, M. Saarela, D. Strbian, F. Scheperjans; France—Centre Hospitalier de Saint Denis–Hôpital Delafontaine: T. De Broucker, C. Henry, R. Cumurciuc, N. Ibos-Augé; Centre Hospitalier de Versailles André-Mignot: A.-C. Zéghoudi, F. Pico; CH Calais: O. Dereeper, M.-C. Simian, C. Boisselier, A. Mahfoud; CHRU de Brest: S. Timsit, F.M. Merrien; CHU de Nantes—Hôpital G&R Laënnec: B. Guillon, M. Sevin, F. Herisson, C. Magne; Hôpital de Meaux: A. Ameri, C. Cret, S. Stefanizzi, F. Klapzcynski; Hôpital Kremlin Bicêtre: C. Denier, M. Sarov-Riviere; Hôpital Lariboisière: C. Stapf, P. Reiner, J. Mawet, D. Hervé, F. Buffon; Hôpital Ste-Anne: E. Touzé, V. Domigo, C. Lamy, D. Calvet, M. Pasquini; Hôpital Tenon: S. Alamowitch, P. Favrole, I.-P. Muresan; Pitié Salpêtrière: S. Crozier, C. Rosso, C. Pires, A. Leger, S. Deltour; Roger Salengro Lille: C Cordonnier, H. Henon, C. Rossi; Service de Neurologie et Neurovasculaire, Groupe Hospitalier Paris Saint Joseph: M. Zuber, M. Bruandet, R. Tamazyan, C. Join-Lambert; Germany—Charité-University Medicine Berlin—Center for Stroke Research Berlin (CSB): E. Juettler, T. Krause, S. Maul, M. Endres, G.J. Jungehulsing; Department of Neurology University of Heidelberg UMM Mannheim: M. Hennerici, M. Griebe, T. Sauer, K. Knoll; Department of Neurology, University of Ulm: R. Huber, K. Knauer, C. Knauer, S. Raubold; Dresden University of Technology, University Hospital, Department of Neurology: H. Schneider, H. Hentschel, C. Lautenschläger, E. Schimmel, I. Dzialowski; Goethe University Hospital Frankfurt: C. Foerch, M. Lorenz, O. Singer, I.M. R. Meyer dos Santos; Klinikum Frankfurt (Oder): A. Hartmann, A. Hamann, A. Schacht, B. Schrader, A. Teíchmann; Martin Luther University: K.E. Wartenberg, T.J. Mueller; University Hospital Düsseldorf: S. Jander, M. Gliem, C. Boettcher; University Medical Center Hamburg–Eppendorf: M. Rosenkranz, C. Beck, D. Otto, G. Thomalla, B. Cheng; Hong Kong—Prince of Wales Hospital, Chinese University of Hong Kong: K.S. Wong, T.W. Leung, Y.O.Y. Soo; India—Apollo Hospitals: S. Prabhakar, S.R. Kesavarapu, P.K. Gajjela, R.R. Chenna; Baby Memorial Hospital: K. Ummer, M. Basheer, A. Andipet; CARE Hospital, Nampally: M.K.M. Jagarlapudi, A.U.R. Mohammed, V.G. Pawar, S.S.K. Eranki; Christian Medical College & Hospital: J. Pandian, Y. Singh, N. Akhtar; GNRC Hospitals: N.C. Borah, M. Ghose, N. Choudhury; Jehangir Clinical Development Centre Pvt Ltd: N.R. Ichaporia, J. Shendge, S. Khese; Lalitha Super Specialities Hospital: V. Pamidimukkala, P. Inbamuthaiah, S.R. Nuthakki, N.M.R. Tagallamudi, A.K. Gutti; Postgraduate Institute of Medical Education & Research: D. Khurana, P. Kesavarapu, V. Jogi, A. Garg, D. Samanta; St. John's Medical College & Hospital (1): G.R.K. Sarma, R. Nadig, T. Mathew, M.A. Anandan; Italy—Central follow up for Italy: E. Caterbi; Nuovo Ospedale Civile, AUSL Modena: A. Zini, M. Cavazzuti, F. Casoni, R. Pentore, F. Falzone; Ospedale di Branca: S. Ricci, T. Mazzoli, L.M. Greco, C. Menichetti, F. Coppola; Ospedale di Città di Castello: S. Cenciarelli, E. Gallinella, A. Mattioni, R. Condurso, I. Sicilia; San Giovanni Battista: M. Zampolini, F. Corea, M. Barbi, C. Proietti; Sapienza University Unità di Trattamento Neurovascolare: D. Toni, A. Pieroni, A. Anzini, A. Falcou, M. Demichele; The Netherlands—University Medical Center Utrecht (2): C.J.M. Klijn; Norway—Sørlandet Sykehus HF Kristiansand: A. Tveiten, E.T. Thortveit, S. Pettersen; Sykehuset Innlandet HF Lillehammer: N. Holand, B. Hitland; University Hospital North Norway: S.H. Johnsen, A. Eltoft; Pakistan–Aga Khan University: M. Wasay, A. Kamal, A. Iqrar, L. Ali, D. Begum; Portugal—Centro Hospitalar Sao Joao: G. Gama, E. Azevedo, L. Fonseca, G. Moreira; Centro Hospitalar Vila Nova de Gaia: L.M. Veloso, D. Pinheiro, L. Paredes, C. Rozeira, T. Gregorio; Spain—Complejo Hospitalario Universitario de Albacete: T. Segura Martin, O. Ayo, J. Garcia-Garcia, I. Feria Vilar, I. Gómez Fernández; Hospital Clinico de Barcelona: A. Chamorro, S. Amaro, X. Urra, V. Obach, A. Cervera; Hospital Universitari de Girona, Dr Josep Trueta: Y. Silva, J. Serena, M. Castellanos, M. Terceno, C. Van Eendenburg; Switzerland—University of Bern, Inselspital: U. Fischer, M. Arnold, A. Weck, O. Findling, R. Lüdi; United Kingdom—Addenbrookes Hospital: E.A. Warburton, D. Day, N. Butler, E. Bumanlag; Bristol Royal Infirmary: S. Caine, A. Steele, M. Osborn, E. Dodd, P. Murphy; County Durham & Darlington NHS Foundation Trust: B. Esisi, E. Brown, R. Hayman, V. K.V. Baliga, M. Minphone; John Radcliffe Hospital: J. Kennedy, I. Reckless, G. Pope, R. Teal, K. Michael; King's College Hospital: D. Manawadu, L. Kalra, R. Lewis, B. Mistry, E. Cattermole; Leeds General Infirmary: A. Hassan, L. Mandizvidza, J. Bamford, H. Brooks, C. Bedford; Musgrove Park Hospital: R. Whiting, P. Baines, M. Hussain, M. Harvey; New Cross Hospital: K. Fotherby, S. McBride, P. Bourke, D. Morgan, K. Jennings-Preece; Northumbria Healthcare–Wansbeck and North Tyneside General Hospitals: C. Price, S. Huntley, V.E. Riddell, G. Storey, R.L. Lakey; Nottingham University Hospital: G. Subramanian; Royal Bournemouth Hospital: D. Jenkinson, J. Kwan, O. David, D. Tiwari; Royal Devon and Exeter Hospital: M. James, S. Keenan, H. Eastwood; Royal United Hospital Bath NHS Trust: L. Shaw, P. Kaye, D. Button, B. Madigan, D. Williamson; Royal Victoria Infirmary Hospital NHS Foundation Trust: A. Dixit, J. Davis, M.O. Hossain, G.A. Ford; Salford Royal NHS Foundation Trust: A. Parry-Jones, V. O'Loughlin, R. Jarapa, Z. Naing; St George's Healthcare NHS Trust: C. Lovelock, J. O'Reilly, U. Khan; St. Thomas Hospital: A. Bhalla, A. Rudd, J. Birns; University College London Hospitals NHS Foundation Trust: D.J. Werring, R. Law, R. Perry, I. Jones, R. Erande; University Hospital of North Staffordshire (2): C. Roffe, I. Natarajan, N. Ahmad, K. Finney, J. Lucas; University Hospitals of Leicester NHS Trust: A. Mistri, D. Eveson, R. Marsh, V. Haunton, T. Robinson; USA—Mayo Clinic: A.A. Rabinstein, J.E. Fugate, S.W. Lepore.
Conceived and designed the experiments: DZ HA EH CSA. Analyzed the data: DZ HA AG SL. Contributed reagents/materials/analysis tools: HA EH CD YH JW CS TR PL JC CSA. Wrote the paper: DZ HA SS. Critical appraisal of the study: HA SS AG EH CD SL YH JW CS TR PL JC CSA. Supervision of the study: CSA.
- 1. Krishnamurthi RV, Feigin VL, Forouzanfar MH, Mensah GA, Connor M, Bennett DA, et al. Global and regional burden of first-ever ischaemic and haemorrhagic stroke during 1990–2010: findings from the Global Burden of Disease Study 2010. Lancet Glob Health. 2013;1(5): e259–81. pmid:25104492
- 2. Rothwell PM, Wroe SJ, Slattery J, Warlow CP. Is stroke incidence related to season or temperature? The Oxfordshire Community Stroke Project. Lancet. 1996;347(9006): 934–6. pmid:8598757
- 3. Gomes J, Damasceno A, Carrilho C, Lobo V, Lopes H, Madede T, et al. Triggering of stroke by ambient temperature variation: a case-crossover study in Maputo, Mozambique. Clin Neurol Neurosur. 2015;129: 72–7.
- 4. Ohwaki K, Yano E, Murakami H, Nagashima H, Nakagomi T. Meteorological factors and the onset of hypertensive intracerebral hemorrhage. Int J Biometeorol. 2004;49(2): 86–90. pmid:15257452
- 5. Passero S, Reale F, Ciacci G, Zei E. Differing temporal patterns of onset in subgroups of patients with intracerebral hemorrhage. Stroke. 2000;31(7): 1538–44. pmid:10884450
- 6. Bhaskaran K, Hajat S, Haines A, Herrett E, Wilkinson P, Smeeth L. Short term effects of temperature on risk of myocardial infarction in England and Wales: time series regression analysis of the Myocardial Ischaemia National Audit Project (MINAP) registry. BMJ. 2010;341: c3823. pmid:20699305
- 7. Schwartz J, Samet JM, Patz JA. Hospital admissions for heart disease: the effects of temperature and humidity. Epidemiology. 2004;15(6): 755–61. pmid:15475726
- 8. Fischer U, Cooney MT, Bull LM, Silver LE, Chalmers J, Anderson CS, et al. Acute post-stroke blood pressure relative to premorbid levels in intracerebral haemorrhage versus major ischaemic stroke: a population-based study. Lancet Neurol. 2014;13(4): 374–84. pmid:24582530
- 9. Wei JW, Arima H, Huang Y, Wang JG, Yang Q, Liu Z, et al. Variation in the frequency of intracerebral haemorrhage and ischaemic stroke in China: a national, multicentre, hospital register study. Cerebrovasc Dis. 2010;29(4): 321–7. pmid:20130397
- 10. Anderson CS, Heeley E, Huang Y, Wang J, Stapf C, Delcourt C, et al. Rapid blood-pressure lowering in patients with acute intracerebral hemorrhage. N Engl J Med. 2013;368(25): 2355–65. pmid:23713578
- 11. Delcourt C, Huang Y, Wang J, Heeley E, Lindley R, Stapf C, et al. The second (main) phase of an open, randomised, multicentre study to investigate the effectiveness of an intensive blood pressure reduction in acute cerebral haemorrhage trial (INTERACT2). Int J Stroke. 2010;5(2): 110–6. pmid:20446945
- 12. Hansson L, Hedner T, Dahlof B. Prospective randomized open blinded end-point (PROBE) study. A novel design for intervention trials. Prospective randomized open blinded end-point. Blood Press. 1992;1(2):113–9. pmid:1366259
- 13. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet. 1974;2(7872): 81–4. pmid:4136544
- 14. Maclure M. The case-crossover design: a method for studying transient effects on the risk of acute events. Am J Epidemiol. 1991;133(2): 144–53. pmid:1985444
- 15. Whitaker HJ, Hocine MN, Farrington CP. On case-crossover methods for environmental time series data. Environmetrics. 2007;18(2): 157–171.
- 16. Janes H, Sheppard L, Lumley T. Case-crossover analyses of air pollution exposure data: referent selection strategies and their implications for bias. Epidemiology. 2005;16(6): 717–26. pmid:16222160
- 17. Mittleman MA. Optimal referent selection strategies in case-crossover studies: a settled issue. Epidemiology. 2005;16(6): 715–6. pmid:16222159
- 18. Armstrong B. Models for the relationship between ambient temperature and daily mortality. Epidemiology. 2006;17(6): 624–31. pmid:17028505
- 19. Gasparrini A, Armstrong B, Kenward MG. Distributed lag non-linear models. Stat Med. 2010;29(21): 2224–34. pmid:20812303
- 20. Gasparrini A. Modeling exposure-lag-response associations with distributed lag non-linear models. Stat Med. 2014;33(5): 881–99. pmid:24027094
- 21. Gasparrini A, Guo Y, Hashizume M, Lavigne E, Zanobetti A, Schwartz J, et al. Mortality risk attributable to high and low ambient temperature: a multicountry observational study. Lancet. 2015;386(9991): 369–75. pmid:26003380
- 22. Sobel E, Zhang ZX, Alter M, Lai SM, Davanipour Z, Friday G, et al. Stroke in the Lehigh Valley: seasonal variation in incidence rates. Stroke. 1987;18(1): 38–42. pmid:3810768
- 23. Dawson J, Weir C, Wright F, Bryden C, Aslanyan S, Lees K, et al. Associations between meteorological variables and acute stroke hospital admissions in the West of Scotland. Acta Neurol Scand. 2008;117(2): 85–89. pmid:18184342
- 24. Jeong TS, Park CW, Yoo CJ, Kim EY, Kim YB, Kim WK. Association between the daily temperature range and occurrence of spontaneous intracerebral hemorrhage. J Cerebrovasc Endovasc Neurosurg. 2013;15(3): 152–157. pmid:24167793
- 25. Fang CW, Ma MC, Lin HJ, Chen CH. Ambient temperature and spontaneous intracerebral haemorrhage: a cross-sectional analysis in Tainan, Taiwan. BMJ Open. 2012;2(3).
- 26. Aubiniere-Robb L, Jeemon P, Hastie CE, Patel RK, McCallum L, Morrison D, et al. Blood pressure response to patterns of weather fluctuations and effect on mortality. Hypertension. 2013;62(1): 190–6. pmid:23648702
- 27. Alperovitch A, Lacombe JM, Hanon O, Dartigues JF, Ritchie K, Ducimetiere P, et al. Relationship between blood pressure and outdoor temperature in a large sample of elderly individuals: the Three-City study. Arch Intern Med. 2009;169(1): 75–80. pmid:19139327
- 28. Keatinge WR, Donaldson GC, Bucher K, Cordioli E, Dardanoni L, Jendritzky G, et al. Cold exposure and winter mortality from ischaemic heart disease, cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. Lancet. 1997;349(9062): 1341–1346. pmid:9149695