The health effects of hotter summers and heat waves in the population of the United Kingdom: a review of the evidence

It is widely acknowledged that the climate is warming globally and within the UK. In this paper, studies which assess the direct impact of current increased temperatures and heat-waves on health and those which project future health impacts of heat under different climate change scenarios in the UK are reviewed. This review finds that all UK studies demonstrate an increase in heat-related mortality occurring at temperatures above threshold values, with respiratory deaths being more sensitive to heat than deaths from cardiovascular disease (although the burden from cardiovascular deaths is greater in absolute terms). The relationship between heat and other health outcomes such as hospital admissions, myocardial infarctions and birth outcomes is less consistent. We highlight the main populations who are vulnerable to heat. Within the UK, these are older populations, those with certain co-morbidities and those living in Greater London, the South East and Eastern regions. In all assessments of heat-related impacts using different climate change scenarios, deaths are expected to increase due to hotter temperatures, with some studies demonstrating that an increase in the elderly population will also amplify burdens. However, key gaps in knowledge are found in relation to how urbanisation and population adaptation to heat will affect health impacts, and in relation to current and future strategies for effective, sustainable and equitable adaptation to heat. These and other key gaps in knowledge, both in terms of research needs and knowledge required to make sound public- health policy, are discussed. Electronic supplementary material The online version of this article (10.1186/s12940-017-0322-5) contains supplementary material, which is available to authorized users.

The current paper is a review of the evidence between high ambient temperatures / heat waves and different health outcomes in the UK. The topic of this review is timely and of public health interest. The results may be useful when discussing the impact of heat today, impacts of future climate change on human health and adaptive capacity of populations. However, there are limitations to the study.

Major comments:
My main concern is that no inclusion criteria are presented in the current review of the UK based studies. The review would benefit from letting the reader know how the literature search was conducted. For instance, what terms were used to identify relevant studies, when the studies should have been published and what kind of studies that were included.
In addition, the literature search is incomplete. I am missing a number of publications where the impact of heat on mortality in the UK were investigated, albeit in a multicity setting. For instance, de Donato et al (2014), Gou et al (2014) and Gasparrini et al (2015). The current review sometimes includes multicity studies, but not in a consistent way. Presenting all such studies would improve the quality of the review considerably.
It would be easier to get an overview of the studies included in the review were included in a table where the key information such as study setting and thresholds were presented.

Minor comments:
R 54-55: Mortality and morbidity from many causes is too unspecific, please specify. R 75: I believe the referred study used daily maximum temperature with a lag of 0-1 days. The section on work productivity and heat could be complemented with the review by Gubernot et al (2015) in American journal of industrial medicine. Socio-economic status: I think this section could be improved by a discussion regarding individual level SES versus neighbourhood level SES. See also Urban et al in IJERPH (2016) for another study in European setting. R 216: There is a nice study from the Netherlands regarding evidence of declining vulnerability to heat covering more than a century (Ekamper et al 2009, Demographic Research). This would provide additional information in a European setting as compared to the cited US studies. In addition, Gasparrini recently published a study on the declining effects of temperature on mortality within summers.
MMT is not found in the main document.
Competing interests declaration: I declare that I have no competing interests.

Reviewer 2: Simon Gosling
I have added line numbers to the manuscript (attached) and refer to them here throughout the review.
I like the idea of this review, which aims to summarise the impact of current increased patterns of high temperatures and heat waves on health as well as studies that have projected future health impacts of heat under different climate change scenarios, for the UK. The research could, potentially be a very useful resource.
However, despite my enthusiasm for a piece of work such as this, it is my opinion that the review requires major revision before it can be published and of use to researchers (and a policy audience in particular). The overarching reason for this recommendation is that I find the review to be significantly deficient in its depth of and breadth of literature coverage. Numerous articles are missing from the review and to this end the article does not achieve its aim. Sadly this means that the article does not provide an accurate summary of the current state of science on UK heat and mortality studies (present and projected). In order to address this, the authors need to conduct their review far more systematically and in turn refer to, and discuss, the literature that is missing.
I have provided some general and specific comments below which I hope that the authors will consider and find constructive. Addressing these comments will unfortunately require a significant amount of reading and reviewing, and new writing, to be undertaken by the authors, but I believe that if these comments are addressed then the overall quality of the manuscript will be improved greatly.

General comments
The title does not mention anything about health. The immediate impression to the reader is that this is going to be a review of climatological evidence. To enhance the chances of this article, if it is published, showing up in academic search engine results it might be worth changing the title to better reflect that contents of the review.
Numerous articles that I expected to see in the review are missing, so I am curious to know the inclusion/exclusion criteria used for selecting articles. There should really be a short Methods section that describes how the review was conducted. I acknowledge that the authors note in the Introduction that they only include epidemiological studies, but the problem with doing this is that it misses out a very large amount of evidence that is crucial, and needed, to actually address the authors' aim. This is one of the reasons that the review does not accurately represent the current state of knowledge on heat and health in the UK.
As a result of very little information on how the review was conducted, the reader is left asking several important questions, which really need to be answered so that the reader can understand the context and quality of the review: • What tools were used for conducting the literature review (e.g. PubMed, Google Scholar, Web of Science)?
• What key search terms were used?
• Were any journals included/excluded? • Did the authors only search for articles that focus on the whole of the UK, or did they include studies that investigated impacts for smaller spatial domains, e.g. regions, counties, cities? And were global-scale studies included where data for the UK could be extracted? • What years of publication did the authors limit their searches to?
A selection of the many articles that I expected to see citied which are missing are listed below. This is certainly not a comprehensive list of missing articles (I don't have time to list all of them unfortunately -a comprehensive and systematic literature search would yield them). I suggest that the authors consider including the articles listed below (and others) and moreover that they revise their review so that it is a more complete representation of the current state of science on this topic. Conducting the review with a more systematic and rigorous approach is really needed to address this. In its present form, the manuscript reads more like an informal ad-hoc review that may have been included as part of a report. I doubt that this was the authors' intention. Unfortunately, what is more concerning, as a result of the lack of literature coverage, is that there is a risk of some statements and conclusions in the article being factually inaccurate because the authors have overlooked key studies. A couple of examples of this occurring are: 1. Lines 213-214 where the authors state that "The lack of UK studies which have estimated the burden of health effects of heat considered attributable to climate change that has already occurred represents an area for further research". This is not the case because attribution studies on heat and health have been done for England and Wales (Christidis et al. 2010) but sadly the authors seem unaware of this area of research. 2. The situation described around lines 323-330 with regards to policy level interventions is far more complex and nuanced than the authors suggest. A highly relevant discussion of this is presented by Wolf et al. (2015;Int. J. Environ. Res. Public Health 12: 13321-13349) but no reference to this is included (see also Lindley et al. 2006;Journal of Risk Research).
Harvesting (i.e. mortality displacement) is mentioned in several different parts of the manuscript. Some readers might find it helpful if a specific section was included on harvesting that draws all of this information together.
Throughout the manuscript the text weaves between referring to UK and non-UK studies. This is somewhat unhelpful as the abstract suggests the aim is to focus on the UK. This is highlighted in lines 42-48, which is also a little confusing -are the authors saying here that they tried not to include studies that were: 1) written by authors who are based at non-UK institutions (e.g. a study on heat and health in London but written by authors from the Netherlands); or 2) written about health impacts in non-UK countries (e.g. a study about heat and health in the Netherlands)? If they mean the former, then why would they try to extrapolate to the UK? This would not be appropriate. And why review such studies anyway, if the aim is to focus on the UK? Lines 186-188, for example, also drift away from the UK. Overall, the jumping between UK and non-UK studies in the manuscript results in a feeling of inconsistency in approach to the review. I suggest that the authors re-evaluate their overall approach used to include/exclude studies.

Specific comments
Line 33: Need a brief description of why there is a range in projected temperature change (it is quite a large range). Also double-check whether the warming quoted really is relative to the last century -the values appear to be relative to pre-industrial (1800s), not last century (1900s).
Lines 22-23: This sentence is somewhat ambiguous. Need to better differentiate here whether the increases in mortality are due to population change only, climate change only, or climate change and population change with the majority of the change being attributable to population change.
Line 35: General comment: It already is acknowledged as being very important, e.g. see the health chapter in the forthcoming UK Climate Change Risk Assessment.
Lines 62-63: It would be helpful to indicate what some of the other confounders tend to be, beyond just air pollution.
Lines 64-66: Considering the content of this section of text it seems odd that the authors do not mention the term Relative Risk (RR).
Line 65: It is worth noting that threshold temperatures are also known as "optimum temperatures" and "minimum mortality temperatures".
Lines 69-70: Fair point here about the contentious issue of defining heatwaves but considering the aim of this paper, does the UK have a formal definition of a heat wave, perhaps referred to in the Met Office heat health warming system, that can be included here? It would be nice to provide a more UKcentric focus here. Line 129: This subsection is very short and its value as a standalone section is questionable. I suggest the authors consider merging this with one of the previous sections.
Lines 136-149: A number of global-scale assessments of the impact of climate change on labour productivity have been conducted recently (e.g. Dunne et al. (2013;Nature Climate Change) and see also a body of research by Tord Kjellstrom), so it seems a pity not to extract results from these global studies and report them here, especially when there is so little work done on the UK specifically.
Line 144: Why is a US study being discussed when the aim is to review work focussed on the UK?
Line 153: please provide more details here, at the very least some references to support this statement. Line 225: Suggest changing from "parameters and processes included in climate change models" to "parameters, processes and initial conditions included in climate change models".
Lines 247: Worth also noting as a caveat that the emissions scenarios used are somewhat "old" now as they used SRES scenarios. It would be nice to see some studies using the more recent RCPs and SSPs for the UK -it's good to see that this is mentioned in the Conclusion.