University of Southern Denmark Campus Esbjerg University of Southern Denmark Campus Esbjerg

A systematic Review on nature´s impact on the Autonomic Nervous System measured by Heart Rate Variability

Supervisor: Professor of Public Health Arja R. Aro Co-supervisor: Postdoc, ph.d. Maja Larsen Master Thesis Master of Science in Public Health

Lea Neuberg January 28 th 1975 Lea Neuberg Total number of characters: 264.768 January 28 th 1975 Total number of characters:

Title sheet Title: A systematic Review on nature´s impact on the Autonomic Nervous System measured by Heart Rate Variability Author: Lea Neuberg University: University of Southern Denmark Faculty: Faculty of Health Science Education: Master of Science in Public Health Supervisor: Professor of Public Health Arja R. Aro Co-supervisor: Postdoc, ph.d. Maja Larsen Research Unit for Health Promotion Institute of Public Health University of Southern Denmark Pages: 133 Number of figures: 12 Number of tables: 10 Number of appendices: 12

Til Johnny Kuhr Dine ambitioner bragte mig på universitetet - Så satte jeg mig et mål: Jeg ønskede ikke bare at lære noget nyt, men at lære noget andre kunne bruge til noget nyt, jeg håber det er lykkedes.

Abstract Background: Urbanization accelerates, meantime little is known about the importance for humans to live in natural environments. The relation between nature and mental health has been studied and documented. Theories derive from environmental psychology, now researchers seek a physiological explanation; evidence documenting why and how nature impact human health, prerequisite nature-interventions beeing implemented efficiently in public health work and policies. The autonomic nervous system (ANS) has been suggested as the physiological explanation, specifically the Parasympathetic nervous system (PNS). Objective: To explore if ANS can explain the relationship between nature and human health by mapping the knowledge obtained in studies using Heart Rate Variability (HRV) to measure human reaction to nature. Methods : A systematic review of articles published in the last 5 years, using PsycINFO, Global health, Embase, PubMed Medline, Amed, CINAHL and Web of science. Additionally, a search though most cited key articles, followed by snowballing and contact to experts and lastly alerts on databases. The PICO method was used to select studies; Healthy adults, nature, control and HRV-parameters. Eight RCT studies (n=3125) were eligible for inclusion, a Narrative Synthesis was conducted. Results: All studies found at least one significant outcome, documenting the association between nature and ANS by reactions from PNS or the Sympathetic Nervous System (SNS) or both. Real nature could be superseded by virtual nature environments. Several moderators of different categories were indentified: Duration of exposure, exercise, and ANS activity prior to exposure, stress induction, type of nature, senses exposed, personality, preferences and believes. Impact from respiration was unclear. Discussion: The results were based on RCT´s, as cross-over trails. Limitations were; incorrect analyses, lack of blinding of outcome assessments, heterogeneity in the use of HRV-parameters, uncertainties in the interpretation of ANS activity based on HRV. Furthermore lack of possibilities to evaluate practical significance and determine cross-over effects. However enough strong evidence to state; ANS could explain nature’s impact on human health, without mutually excluding other explanations. Substantiating further research to reveal the precise influence from the moderators, develop ideal analyze methods and valid HRV-parameters. Meantime the researcher must be aware if they steer to develop an artificial substitute for nature or creation of the most efficiently real nature to initiate health benefits. Conclusion: A convincing conclusion was the increase in the PNS initiated by viewing nature compared to build environment, documenting the suggested physiological explanation. ANS and nature challenge science to cooperate across disciplines, as more knowledge on ANS is essential, sophisticated analyze methods required and insights to environmental possibilities. This thesis gave documentation to use nature to improve health, prevent disease and indentifies the complexity of investigating nature and ANS, meantime the potential to develop nature as evidence based medicine.

Resume Baggrund: Urbaniseringen accelererer; betydningen af at leve i naturlige miljøer for mennesker er ikke afklaret. Sammenhængen mellem natur og mental sundhed er undersøgt og dokumenteret. De forklarende teorierne stammer fra miljø psykologi, nu søger forskerne en fysiologisk forklaring som kan dokumenterer, hvorfor og hvordan naturen påvirker menneskers sundhed. Denne viden kan bidrage til at gennemføre effektive natur-interventioner i folkesundhedsarbejde og politik. Det autonome nervesystem (ANS) er foreslået som den fysiologiske forklaring, mere præcist det ParaSympatiske nervesystem(PNS). Formål: Undersøge om ANS kan forklare sammenhængen mellem natur og sundhed ved at revidere studier, der har anvendt Hjerte Rytme Variabilitet (HRV) som biomakør for naturens påvirkning af mennesket. Metoder: Et systematisk review af artikler publiceret de sidste 5 år, via PsycINFO, global sundhed, Embase, PubMed Medline, Amed, CINAHL og Web of Science. Efterfulgt af søgning på mest citerede artikler, efterfulgt af snowballing og kontakt til eksperter og endelig nyhedsalarm fra databaser. PICO metoden blev anvendt til at vælge studier; raske voksne, natur, kontrol og HRV -parametre. Otte RCT studier(n= 3125) blev inkluderet og en narrativ syntese gennemført. Resultater: Alle undersøgelser fandt mindst ét signifikant resultat, der dokumenterede sammenhængen mellem natur og ANS; reaktioner fra PNS eller det Sympatiske Nervesystem(SNS) eller begge. Reel natur kunne erstattes af virtuelle natur miljøer. Effektmoderatere af forskellige kategori blev identificeret: Varighed af eksponering, bevægelse, ANS aktivitet forud for eksponeringen, stress induktion, type af natur, sanser, personlighed, samt præferencer og forventninger til natur. Betydning af respiration var uvis. Diskussion: Resultaterne var baseret på RCTs; cross-over designs. Begrænset af; forkerte analyser, manglende blinding, heterogenitet i brugen af HRV-parametre, usikkerhed i fortolkningen af ANS aktivitet baseret på HRV. Desuden manglende mulighed for evaluering af praktisk signifikans, samt cross-over effekt. Dog var der overbevisende evidens for at ANS kunne forklare naturens indvirkning på menneskers sundhed, uden at udelukke andre forklaringer. Samt belæg for yderligere forskning; afklaring af effektmoderaternes præcise betydning for resultatet, udvikle ideelle analyse metoder og valide HRV -parametre. Forskeren skal hvad resultatet kan bruges til; udvikle af en kunstig erstatning for naturen eller oprettelse af den mest effektivt virkelige natur, der kan generere sundhedsfremme og sygdomsforebyggelse. Konklusion: En overbevisende dokumentation for den foreslåede fysiologiske forklaring; en stigningen i PNS initieret af natur, sammenlignet med by miljø eksponering. Forskning i ANS og natur udfordre videnskab til at samarbejde på tværs af discipliner, mere viden om ANS er essentiel, sofistikeret analysere metoder nødvendige og kendskab til miljømæssige muligheder afgørende. Denne afhandling gav dokumentation for at bruge naturen til at forbedre folkesundheden, identificerede kompleksiteten af natur og ANS som forskningfelt, samt skabte et potentiale til at udvikle naturen som evidensbaseret medicin.

Indhold Indhold................................................................................................................................................................. List of abbreviations ............................................................................................................................................ List of abbreviations related to HRV.................................................................................................................... List of Figures ....................................................................................................................................................... List of Tables ........................................................................................................................................................ Chapter 1 ........................................................................................................................................................... 1 Introduction ....................................................................................................................................................... 1 1.1 Reading guide for the thesis .................................................................................................................... 3 Background ........................................................................................................................................................ 4 1.2. Existing theories on the relationship between nature and health......................................................... 5 1.2.1. Biophilia hypothesis......................................................................................................................... 5 1.2.2. Attention Restoration Theory.......................................................................................................... 6 1.2.3. Affective Aesthetic Theory. ............................................................................................................. 7 1.3 Autonomic Nervous System .................................................................................................................... 8 1.3.1. How natures affect the Autonomic Nervous System. ..................................................................... 9 1.3.2. The Autonomic Nervous System (ANS) and Heart Rate Variability (HRV)..................................... 10 1.4 Heart Rate Variability ............................................................................................................................ 11 1.4.1. Measurements of HRV – knowledge from Kubios......................................................................... 13 1.4.3 Interpretation of HRV-parameters ................................................................................................. 16 1.5 Theoretical model: Natures impact on humans explained by ANS and measured by HRV. ................. 18 1.6. The research question .......................................................................................................................... 20 1.7 Delineations ........................................................................................................................................... 20 Chapter 2 ......................................................................................................................................................... 21 Method ............................................................................................................................................................ 21 2.1 The systematic review ........................................................................................................................... 21 2.1.1 Objective ......................................................................................................................................... 22 2.1.2 Procedure ....................................................................................................................................... 22 2.2 Reference management ........................................................................................................................ 23 2.3 Identification of studies ......................................................................................................................... 23 2.3.1. Types of studies ............................................................................................................................. 24 2.3.2 Search strategy ............................................................................................................................... 24 2.4 Eligibility criteria - Selection of studies................................................................................................. 27

2.4.1 People - Study participants............................................................................................................. 28 2.4.2 Outcome ......................................................................................................................................... 28 2.4.2 Exposure ......................................................................................................................................... 28 2.4.3 Comparison..................................................................................................................................... 29 2.5 Exclusion criteria.................................................................................................................................... 29 2.5.1 Exclusion based on language .......................................................................................................... 29 2.5.2 Exclusion based on duplicity ........................................................................................................... 30 2.6 Data extraction ...................................................................................................................................... 31 2.7 Synthesis and analyzes of the included studies and their results ......................................................... 35 Chapter 3 ......................................................................................................................................................... 39 Results ............................................................................................................................................................. 39 3.1 Developing a preliminary synthesis ....................................................................................................... 39 3.1.1. Characteristics of the studies by tabulation: ................................................................................. 40 3.1.2. Summary of the findings by tabulation ......................................................................................... 42 3.1.3 Grouping and clustering of the authors. ........................................................................................ 46 3.1.4 Transforming data into a common rubric ...................................................................................... 48 3.1.5 What is the size of effect ................................................................................................................ 51 Summary of the preliminary synthesis ........................................................................................................ 54 3.2. Exploring relationships within and between studies ........................................................................... 55 3.2.1 Graphic of relationships ................................................................................................................. 56 3.2.2 Moderator variables ....................................................................................................................... 57 3.2.3 Qualitative case description ........................................................................................................... 58 3.3 Summary: Study characteristics and relationships within and between studies. ................................. 67 3.4. Assessing the robustness of the synthesis ........................................................................................... 68 3.4.1 Risk of bias assessment .................................................................................................................. 68 3.4.2 EPPI ................................................................................................................................................. 72 Chapter 4 ......................................................................................................................................................... 74 Discussion ........................................................................................................................................................ 74 4.1 Short summary of the main findings ..................................................................................................... 74 4.2 Reflecting critically on the synthesis process ........................................................................................ 74 4.2.1 The methodology............................................................................................................................ 74 4.2.2 Evidence used. ................................................................................................................................ 77 4.2.3 Assumptions made. ........................................................................................................................ 80

4.2.5 Expected changes in technology or evidence ................................................................................ 83 4.2.6 Aspects that may have an influence on the implementation of the technology and its effectiveness in real settings. .................................................................................................................. 84 Chapter 5 ......................................................................................................................................................... 85 Conclusion ....................................................................................................................................................... 85 Appendix .......................................................................................................................................................... 86 Appendix I

Organization of the nervous system.................................................................................... 86

Appendix II

Anatomy of the Autonomic Nervous System .................................................................. 87

Appendix III

The sinoatrial (SA) node................................................................................................... 88

Appendix IV

PubMed screening of 7 articles ....................................................................................... 89

Appendix V

Documentation of the literature search.......................................................................... 90

Appendix VI

Search through most cited key articles ........................................................................... 91

Appendix VII

Pilottest............................................................................................................................ 92

a) Appendix VII 3 articles .......................................................................................................................... 92 (1)................................................................................................................................................................. 92 (2)................................................................................................................................................................. 93 (3)................................................................................................................................................................. 94 b) Appendix VII Test data sheets .............................................................................................................. 95 Appendix VIII List over excluded rapports............................................................................................. 96 Appendix IX

Risk of bias ....................................................................................................................... 98

Appendix X

CONSORT Cheklist NPT .................................................................................................. 102

Appendix XI

Datasheets ..................................................................................................................... 104

a) Appendix XI

Analysis .......................................................................................................................... 104

b) Appendix XI

Autors ............................................................................................................................ 106

c) Appendix XI

Conclusions .................................................................................................................... 108

d) Appendix XI

Meassurements ............................................................................................................. 110

e) Appendix XI

Method .......................................................................................................................... 112

f) Appendix XI

Overview ........................................................................................................................ 114

g) Appendix XI

Participants .................................................................................................................... 115

h) Appendix XI

Setting ............................................................................................................................ 116

Appendix XII

Flowchart of the synthesis process ............................................................................... 118

List of abbreviations AAT

Affective Aesthetic Theory

ANS

Autonomic nervous System

ART

Attention restoration theory

BP

Blood pressure

CNS

Central nervous system

ECG

Electrocardiography

EEA

Environment of Evolutionary Adaptation

HR

Heart rate

HRV

Heart rate variability

NCD

Non Communicable Diseases

NPT

Checklist of Items for Reporting Trials of Non- Pharmacologic Treatment from the CONSORT Statement

NS

Narrative Synthesis

PNS

Parasympathetic nerve system

RCT

Randomized controlled trials (RCT);

RSA

Respiratory sinus arrhythmia

SA

Sinoatrial

SCL

Skin conductance level

SNS

Sympathetic nervous system

SR

Systematic review

Cochrane Handbook

Cochrane Handbook for Systematic Reviews of Interventions

The Task Force

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology

List of abbreviations related to HRV

HF

High Frequency

LF

Low Frequency

n.u

Normalized units

RMSSD

Square root of the mean squared differences of successive NN intervals

SD1

Short term variability

SD2

Long term variability

SDNN

Standard deviation of NN intervals

List of Figures

Figure 1: Visualized theory; ANS reaction to nature exposure. ...................................................................... 10 Figure 2 ECG show how heart rate varies with every heartbeat= Heart Rate Variability. .............................. 12 Figure 3 A screen shot from the Kubious program with analyzed HRV-parameters. ..................................... 15 Figure 4: Natures impact on humans explained by the ANS and measured by HRV. ..................................... 19 Figur 5 Flowchart illustrating the identification and selection of studies. ..................................................... 31 Figure 6 Procedure for data-extraction ........................................................................................................... 32 Figure 7 Grouping by authors .......................................................................................................................... 46 Figure 8 Visualization of the procedure .......................................................................................................... 49 Figure 9 Grafic on relationships ....................................................................................................................... 57 Figure 11 Risk of bias ....................................................................................................................................... 70 Figure 12 Risk of bias by Study ....................................................................................................................... 71

List of Tables

Table 1 Strategy for the block search .............................................................................................................. 25 Table 2 The search strategy to identify eligible studies for inclusion in the review ....................................... 27 Table 3 Included studies. ................................................................................................................................. 39 Table 4 Study charateristics............................................................................................................................. 41 Table 5. Summary of conclusions .................................................................................................................... 43 Table 6 Counting significant results................................................................................................................. 44 Table 7. Counts for the specific effect triggering ANS activity. ....................................................................... 45 Table 8. The result sheet. ................................................................................................................................ 51 Tabel 9 Moderators ......................................................................................................................................... 57 Table 10 Overall weight of studies by EPPI ..................................................................................................... 73

Chapter 1 Introduction There is limited knowledge on the link between nature and health. Consequently, knowledge of this potential link is not implemented in health politics and public health thinking in Denmark. More research on the causal relationship between nature and health is needed, (“Folkesundhedsrapporten” - report on public health) (Kjøller M, Juel K, & F, 2007). In 2009-2011 the Nordic Council of Ministers financed the project ”Frisk i Naturen” (fresh in nature), carried out by Nordic outdoor organizations, with the purpose to gather arguments for nature’s health promoting effect. The final report from 2012 concluded that there is enough documentation to utilize nature for health promoting purposes in the Nordic countries. However, there is still a need to document which and how physiological mechanisms are affected. The health promotion focus is based on the industrialized countries, dominated by non-communicable diseases (NCD). The project pointed out the advantage of nature as complementary and also an important strategy to reduce health inequalities in the populations. Cross- disciplinary, research is crucial to increase the use of nature to improve public health. (Norden, 2012). NCDs are complex and multifactorial and call for a number of assumptions to relate the health effects to risk factors, which also are inter-dependent. The behavioral and biological factors are affected by the living conditions, determined by education and work, which is a function of the way society, is organized. (Juel, J., & H., 2006) The organization of the society determines the way humans interact with nature; industrialization and urbanization result in people moving further away from nature, elaborating accelerating changes in living conditions, with unknown consequences on behavioral and biological factors affecting human health. According to World Health Organization (WHO) more than half of the earth's population lives in cities, in Europe the percentage is 70 and this number is expected to rise. It affects the way society is organized and results in rapid changes in our lifestyles when our behavior is influenced by living conditions in cities. (WHO, 2010; 2012, p. 7).

Introduction

1

The consequences of industrialization and technology in transport have resulted in a physically inactive behavior. Concurrently, there was an explosion in the incidence of type 2 diabetes in the younger population and a new syndrome occurred - the metabolic syndrome. Only recently researchers identified the changing biological factors; the discovery of the muscles as endocrine organ that explain causality and document why physical activity is essential for human and why lack of activity rises the risk of developing metabolic syndrome and type 2 diabetes (Pedersen & Febbraio, 2012; Sundhedsstyrelsen, 2011). This thesis seeks to relate nature to human health; the focus area is the underlying mechanisms, defining how and why natural environments can have positive effects on human biological factors and thus affect health. A Doctoral Thesis from the Swedish University of Agricultural Sciences in Alnarp by Matilda Annerstedt, published in 2011, suggested that the relaxing effect from nature may partly be mediated by an active relaxation mechanism within the parasympathetic nerve system(PNS), therefore an attempt to find the causal relationship between nature and public health should explore the autonomic nervous system(ANS). Thus underlying mechanisms, explaining the relationship between nature and health, calls for studies using physiological outcomes such as biomarkers. Looking for reviews to find out the state of the art; subjective self-reported emotions were the most common outcome used. There was limited empirical knowledge based on biomarkers, especially there are very few studies using similar biomarkers as outcome. Biomarkers used are; heart rate variability (HRV), heart rate (HR), blood pressure(BP), salivary amylase activity, cerebral activity in the frontal lobe, anticancer proteins, natural killer (NK) cell activity and number of NK cells, endocrine markers; adrenaline/noradrenaline in the urine and cortisol concentration in saliva (Bowler, Buyung-Ali, Knight, & Pullin, 2010; Gladwell, Brown, Wood, Sandercock, & Barton, 2013; Lee et al., 2011). A review concluded that a standard measurement of relevant biomarkers as outcome would facilitate further meta-analyses on the health benefits of the natural environment, when more datasets become available (Bowler et al., 2010) The need for more knowledge on nature’s impact on humans, combined with Annerstedt´s suggestion about the underlying mechanisms within the ANS, which can be measured by HRV, the focus of this thesis is to map the knowledge obtained in studies using HRV as the outcome to document nature’s impact on humans. The aim is to explore if the causal relationship between nature and human health can be explained by the ANS.

Introduction

2

1.1 Reading guide for the thesis This thesis is divided into seven main chapters. This reading guide is placed in the middle of chapter one; the introduction has already been read. The rest of chapter one contains background information on the Autonomic Nervous System and Heart Rate Variability in an effort to provide the reader with an understanding of basic terms, physiological mechanisms and technical aspects. The background information was expanded with more detailed theory to meet the challenge of developing a model of theory, as the first of four elements in a narrative synthesis (NS). This knowledge made basis to choose data extraction, understanding key aspects of the studies included in this review and to judge the objectives, assumptions, methods, outcomes and results of each study. In Chapter 2 the research question are presented as well as the methodologies applied throughout the thesis. This concerns in particular how the systematic review (SR) was conducted, how studies were identified through the literature search, and the selection process of the studies included in the review. The last part of this chapter concerns the methods used for extracting data from the selected studies and the methods from NS applied to draw conclusions from the SR. The purpose of this chapter is therefore to illuminate the choices that were made throughout conducting the review in order to create transparency of how results and conclusions are reached in the remaining parts of the thesis. Chapter 3 provides the results and chapter 4 discuses the results, first by finishing the NS with a systematic approach to the discussion of the results, followed by a discussion of potentials for health promotion and disease prevention. The chapter also presents needs for further research. Lastly, conclusions are drawn in chapter 5 and implications for public health practice (and research) are presented. The remaining chapters contain the bibliography as well as the appendices, which provide additional information on specific issues in the SR.

3

Background The reason this thesis makes an effort to map the knowledge obtained in studies using HRV, is not to investigate HRV, but to investigate if the causal relationship between nature and human health, can be explained by ANS. Therefore the research must first explore the physiological mechanisms of ANS, in order to develop a theoretical model of how nature could have impact on health. Before developing a theoretical model, the justification to assume there is a relation between nature and health and that the relation is causal must be presented, this will also add to the identification of the gab this thesis aims to fill in. Assuming that ANS explains the relation between nature and humans is related to the biological plausibility. Failure to make biological sense does not necessarily negate the possibility of a causal association. In some cases, our understanding of the biological mechanisms may be incomplete, and what does not make sense today may make sense sometime in the future. This is an attempt to make sense in the present and fill in the evidence gap to increase knowledge on relation between nature and health. Impact on health is the core element. Health effects believed to be determined by nature are summarized below and indication of causality identified. Nature can potentially increase resistance against diseases, prevent osteoporosis, adjust the human biological clock, improve sleep quality, reduce stress, prevent depression, stabilize blood pressure, increase concentration and motor skills, reduce headaches, fatigue, medication use, digestion problems and sickness absence, and patients with windows to green areas have been shown to be discharged from hospital nearly a day earlier compared to patients with windows to build environment (Holm & Tvedt, 1998; Norden, 2012). Nature increases physical activity and quality of life and reduces the risk of obesity and stress (Stigsdotter et al., 2011). These findings come from different populations, different methodologies and study designs. The replications of findings points in the same direction: Nature is associated with health benefits; this increases the probability that the association is causal. There are also indications that there is a doseresponse relationship, as frequencies of park visits are related to level of stress, obesity and physical activity (Stigsdotter et al., 2011). The relation between nature and mental health, in terms of stress, depression and mood, is the most frequently studied and documented (Hansen & Nielsen, 2005; Koksvik, 2009; Norden, 2012; Wamalas, 2008). Going back in history to Hippocrates; in the ancient Egypt garden walks were prescribed for mentally ill royalty. As the medical science developed the interest for nature and gardening as a medical therapy

Background

4

declined. Not until the beginning of the 1980s did the theories about nature’s impact on health see a new light; these theories are rooted in the tradition of environmental psychological values, closely linked to theories of stress, mental fatigue, and restoration (Annerstedt, 2011). These theories go in line with the well documented effect on mental health. Focus on the biological plausibility requires a theoretical explanation of the physiological mechanisms, which is also related to the medical science field. The theories rooted in the psychological values are well described and accepted knowledge in the environmental science disciplines. These theories form background in the development of a physiological theoretical model explaining mechanisms of ANS as the biological plausibility. The theoretical model must include HRV as a measure for ANS. This will be done in a visual model represented in the end of the background, just before presenting the research question of this thesis.

1.2. Existing theories on the relationship between nature and health Tree theories were selected based on; a) the theories used by Annerstedt; b) the reference list from the reviews found searching the state of the art in the explanation on nature´s impact on humans and c) the theories used by authors in the literature found through the homepage from “Frisk I Naturen”, which purpose was to gather arguments for nature’s health promoting effect on a scientific basis. The three theories are: Biophilia hypothesis, Attention Restoration Theory (ART) and the Affective Aesthetic Theory (AAT).

1.2.1. Biophilia hypothesis The question why nature promotes health is by many authors answered within the Biophilia hypothesis; the theory claims humans to have an inherent inclination to affiliate with nature and therefore imply affection for plants and other living things. The theory also answers the question what happens when humans move away from nature; built environments lack these natural elements, risking a consequentially negative impact on human mind. The term Environment of Evolutionary Adaptation, or EEA, is used to denote the qualities of the environment Background

5

humans are adapted to live in. Obviously this environment comprised a closer presence of nature compared to what most people experience today. Plants were of crucial importance for survival during most of our evolutionary history as a food resource, for shelter, and as an indicator of water. On a purely theoretical ground, one would expect the presence of plants, as an integral part of the human EEA, to have had an impact on the evolution of the brain. We are presumably adapted to live in a green environment. Deviations from the way of life for which we are genetically designed have been referred to as mismatches. Some mismatches are beneficial, such as sleeping on a mattress instead of on the ground, while others may contribute to disease or reduce quality of life. The word discord is used for mismatches that have a negative impact; i.e., they cause some form of stress, at least in susceptible individuals (Grinde & Patil, 2009).

1.2.2. Attention Restoration Theory How nature affects humans, many authors answered within the Attention restoration theory (ART), developed by an American environmental psychologist team headed by Kaplan & Kaplan. ART consider nature as a "restorative environments" - facilitating relaxation, based on the stress theoretical aspect, where nature creates new mental resources, because nature stimulates the spontaneous attention. Kaplan & Kaplan developed the ART, explaining how nature promotes health. According to this theory two diverse types of attention are functioning within the human mind – directed and spontaneous. Nature, which is filled with intriguing stimuli, grabs attention spontaneous, in a bottom-up fashion, allowing topdown directed-attention abilities a chance to replenish. Unlike natural environments, urban environments are filled with stimulation that capture attention dramatically and additionally requires directed attention (e.g., to avoid being hit by a car), making them less restorative. What is popularly called "being stressed out" may more accurately be a reflection of a worn-out attention capacity. Recovering from this all too common state is aided by settings, e.g. nature, which minimizes demands on our attention by grabbing attention spontaneous. Such restorative experiences can be of very short duration; perhaps helping explain why having nature in the view from the window has pervasive psychological impacts with improved directed-attention abilities. (Berman, Jonides, & Kaplan, 2008). The essential mechanism is the central inhibition capacity, Annerstedt explains; to keep directed-attention requires suppressed and inhibited concurring stimuli. This mechanism is largely under intentional control

Background

6

and is effortful. Hence directed attention is a resource that may be depleted; the inhibitory mechanism will become fatigued with prolonged or intensive use, causing fatigued, reduced self-control which eventually cause stress and irritability. If the resource becomes completely emptied we experience exhaustion. The spontaneous attention is the key to recovery. The spontaneous attention is an involuntary and noneffortful kind of action. It is an automatic mechanism, triggered by spontaneous fascination for stimuli. When the involuntary, spontaneous attention is engaged, the demands on directed attention are diminished, thus rendering a chance for restoration of attention, and bringing mental competence back to normal. Involuntary attention occurs in settings where the amount of stimuli is adequate and manageable, often referred to as restorative environments e.g. nature. The theory has gained some empirical support, mainly in quasi-experimental studies, where individuals’ cognitive competence has significantly increased by regular visits to nature(Annerstedt, 2011, p. 16).

1.2.3. Affective Aesthetic Theory. How nature affects humans, can also be explained by Affective Aesthetic Theory (AAT). This theory differs from ART by focus on psychological and affective reactions, rather than cognitive ones. As many other authors Annerstedt refers to Ulrich, who made empirical support for this theory by studies mainly focusing on short, restorative experiences, showing for example immediately induced blood pressure reduction from contact with nature(Annerstedt, 2011, p. 18). The psychological and affective reactions are assumed to evolve from visual stimuli. AAT considers the stress reduction to be achieved in nature as the main factor for nature’s health effects and is sometimes called Stress Reduction Theory. By a certain visual stimulus we innately and immediately react with wellbeing, calmness, and relief from stress – all restorative responses. In this perspective visual stimuli from nature, such as water or greenery, would induce this instinctive, positive, affective relaxation-response. AAT proposes that restoration can occur when a scene elicits feelings of mild to moderate interest, pleasantness, and peacefulness. It evokes interest and positive effect, holds attention, and thereby displaces or restricts negative thoughts and allows arousal heightened by stress to drop to a more moderate level; these have been demonstrated by for example lower blood pressure, heart rate, and muscle tensions. (Annerstedt, 2011, pp. 17-18).

Background

7

A very simple interpretation: The Biophilia hypothesis explains why nature effects humans, ART and ATT attempt to explain how nature affects health. The ART suggests that relaxation reduces stress due to a cognitive automatic mechanism triggered by nature, and AAT suggests it is the decreased arousal level, which reduces stress due to psychological and affective reactions when viewing nature. ART and AAT are not mutually exclusive.

1.3 Autonomic Nervous System In the nervous system, afferent pathways carry nerve impulses from the body e.g. skin, ears, eyes, tongue and many other organs, muscles and joints, towards the central nervous system (CNS). The opposite activity of direction or flow is efferent pathways carrying nerve impulses away from CNS to effectors e.g. muscles or organs (Blessing & A Gibbins, 2008). ANS is part of the human nervous system; it is the branch of the nervous system that controls the visceral functions in order to maintain the “homeostasis/homodynamic” state of the body, being an interface between body, CNS, and the external stimuli (Montano, Tobaldini, & Porta, 2012). Appendix I (Page86) The anatomy of the ANS can be seen in Appendix II (page 87), ANS innervates, and thereby controls, organs and tissues throughout the body (eyes, salivary glands, heart, stomach, urinary bladder, blood vessels, etc), independently of voluntary control (Binder, Hirokawa, & Windhorst, 2009). ANS consist of the sympathetic nervous system (SNS) and the parasympathetic system (PNS). SNS primarily controls activation and mobilization and is often called the fight or flight response. PNS controls restoration and relaxation, also called the healing, rest and digest system, because it conserves energy as it slows the heart rate, increases intestinal and gland activity, and relaxes sphincter muscles in the gastrointestinal tract (Blessing & A Gibbins, 2008).

In the last few decades, intensive research in this area allowed to better understand the physiological interaction between SNS and PNS efferent pathway (Montano et al., 2012). It is a misleading simplification to interpret that the two systems oppose each other; neither is ever activated in its entirety. Each system consists of a series of discrete functional pathways that may be activated from CNS either independently or in patterns contributing to bodily homeostasis, according to the particular requirement of the particular daily activity (Binder et al., 2009; Blessing & A Gibbins, 2008).

Background

8

The function and mechanisms of the ANS, as part of the nervous system and part of different body functions is not totally explored. ANS is a complex system based on the integration of several reflexes having both negative and positive feedback characteristics, continuously balancing each other, playing a crucial role as interface between visceral function and physical/emotional stress challenges. This reflex activity is directed both upstream, impinging upon CNS structures related to sleep–wake cycle and activating the arousal system, as well as downstream, regulating visceral (cardiovascular, gastrointestinal and genitourinary) functions. It has recently been observed that ANS also is able, not only to be modulated, but also to modulate immune response, thus strengthening the importance of ANS as a possible target system to elaborate countermeasures to better cope with stressful conditions.(Montano et al., 2012) This relationship between the brain and the heart has been described in the model of neurovisceral integration (Thayer & Lane, 2009).

The complexity of the reflex activity covering knowledge on the CNS, limbic system, hypothalamus, efferent and afferent pathways, baroreceptors, pre- and postganglionic neurons, neurotransmitter substances, synaptic transmission, re- and polarization and detailed S-A node function, are left out in this presentation.

ANS must be tested by measuring the end-organ responses to the exposure of interest, because the anatomical location makes it inaccessible to simple, direct physiological testing (Freeman & Chapleau, 2013). The exposure of interest in this thesis is nature and the end organ is the heart. The simple version of the physiological mechanisms of ANS; all external stimuli summed, e.g. smell of flowers, viewing nature, traffic noise or dogs barking, determine the nerve impulses in the afferent pathways from the body to CNS, the efferent pathways from CNS to the body determines the reaction; SNS makes the body ready to fight or flight and PNS controls restoration and relaxation. SNS and PNS activity can be measured through the heart. ANS plays an important role in various body functions related to stress coping, sleep, immune response, emotions and cognitive performance. The interplay between PNS and SNS is essential for the body’s ability to respond appropriately to the requirements.

1.3.1. How natures affect the Autonomic Nervous System.

Background

9

Figure 1 (page 10): Developed by the reviewer from the literature about ANS combined with the AAT and ART as the first step in the development of a theoretical model on natures impact on humans explained by the ANS and measured through HRV. Figure 1: Visualized theory; ANS reaction to nature exposure.

Viewing, smelling, hearing, sensing = nature exposure

Afferent pathways carry nerve impulses to CNS AAT: Views of nature trigger innate reaction ART: The intriguing stimuli, grab attention spontaneous

SNS decrease AAT: Arousal heightened by stress drops Measurable changes in the heart

Affecting the heart

PNS increase ART: Attention recovery =cognitive resources increase

CNS processes all impulses and the effect of the nature exposure dominates the efferent pathways in the ANS

ANS positively regulates the bodily homeostasis through visceral impact AAT: “feel good” emotions↑ ART: Cognitive performance ↑

1.3.2. The Autonomic Nervous System (ANS) and Heart Rate Variability (HRV) The heart as end-organ can be measured by the heart rate (HR).The heart rate is the rate of heart beats pr. minute, it is regulated through the sinoatrial (SA) node of the heart Appendix III (page 88), the SA –node is innervated from the cardiac plexuses, combining branches from SNS by thoracic cardiac nerves and PNS though the vagus nerve, the 10th cranial nerve (Appendix II). Activity in the vagus nerve is referred to as vagal tone; the specific activity in branches to the cardiac plexuses from the vagal nerve is referred to as cardiovagal activity. The combined innervation of the heart is referred to as the sympathovagal balance (Binder et al., 2009; Blessing & A Gibbins, 2008). CNS controls body functions both by increasing and decreasing PNS and SNS. ANS determines the variability between consecutive heart beats –HRV (Marek, 1996). HRV is a result of ANS regulation of the S-A node. The heart rate is kept low though vagal tone which is controlled by PNS. SNS is activated to increase heart rate during stress, anxiety or exercise, to help the body to go into action (fight or fight). SNS activity tends to increase HR and decrease HRV, whereas PNS tends to decrease HR and increase HRV (Tarvainen, Niskanen, Lipponen, Ranta-aho, & Karjalainen, 2013). Background

10

The degree of HRV provides information about the functioning of the nervous control on the HR and the heart’s ability to respond.

Other physiological factors determine the HR, but ANS is the most prominent; the reason for the HR to go up during inspiration and down during expiration is also because SNS is more active during inspiration than during expiration, where PNS is more active (Binder et al., 2009) . This naturally occurring variation in heart rate that occurs during a breathing cycle; increasing HR during inspiration and decreasing HR during expiration, is termed Respiratory sinus arrhythmia (RSA) and it generally correlates with PNS, although nonlinearity in the relationship has been noted, particularly at high levels of PNS (Freeman & Chapleau, 2013). Maintaining the “homeostasis/homodynamic” state of the body is essential; as the PNS is the restorative system and greater HRV comes from PNS and lower HRV comes from SNS dominance, it is preferred to have a greater HRV.(Rajendra, Paul, Kannathal, Lim, & Suri, 2006; Tarvainen et al., 2013) Using the heart as the end-organ measuring HRV to test ANS, is described both by Freeman & Chapleau and Montano et al. Whereas Freeman & Chapleau present HRV as a parameter mainly for the PNS, Montano presents HRV as a parameter for both PNS and SNS. Further statements about the interpretation of HRV will be presented in the next section describing HRV, but it is important to be aware of the complexity of ANS.

1.4 Heart Rate Variability HRV is the physiological phenomenon of variation in the time interval between heartbeats; the beat-tobeat interval, also known as R-R intervals or the NN intervals. HRV is the interval between consecutive beats rather than the heart rate per se. An average heart rate of 60 beats per minute (bpm) does not mean that the interval between successive heartbeats would be exactly 1.0 sec.; instead it may fluctuate/vary from 0.5 sec up to 2.0 sec. (Marek, 1996) Explanation to Figure 2: The numbers in the figure are milliseconds (ms) and NN interval 1-4 vary from 732845 ms. Source: www.polar.com

Background

11

Figure 2 ECG show how heart rate varies with every heartbeat= Heart Rate Variability.

R-R (NN)interval (1)

(2)

(3)

(4)

The Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology described HRV as one of the most promising biomarker. The specific goals of this Task Force were to standardize nomenclature, develop definitions of terms and specify standard methods of measurement to avoid incorrect conclusions and for excessive or unfounded extrapolations, as the significance and meaning of the many different measures of HRV are more complex than generally appreciated (Marek, 1996). HRV has been developed to detect disease in a medical world; however, HRV is nowadays also used by others than cardiologists; it has been clearly shown that HRV provides reliable information on the ANS function also in healthy people, believed to be a reliable tool capable of investigating SNS and PNS modulation (Montano et al., 2012). The Task Force has not published any updates since 1996. HRV measurements are easy to perform, being noninvasive and having a good reproducibility if used under standardized conditions. Electrocardiography (ECG) or R-R monitors are used to analyze HRV. There are three ways of analyzing the data from the ECG; Time domain, Frequency domain and the Non-linear method. Those will be presented in the next section, followed by a presentation of the kubios computer program to analyze HRV-parameters translating the data to interpret the ANS activity. Time domain method is the easiest to understand, it reflects mean NN interval, mean heart rate, the difference between the longest and shortest NN interval. The time domain method can register variations of instantaneous heart rate due to e.g. respiration, postural changes, medicine, stress or environment

Background

12

changes. The standard deviation of NN intervals (SDNN) is used, as variability is the target, however it only shows the width of the deviation, not the variability between continuous intervals, so called short-term variation estimating high frequency variations in heart rate. To find the variability between continuous intervals RMSSD is preferred. RMSSD is the square root of the mean squared differences of successive NN intervals and has better statistical properties than NN50, which is the number of interval differences of successive NN greater than 50 ms, and pNN50, which is the proportion derived by dividing NN50 by the total number of NN intervals. It is inappropriate to compare time–domain measures obtained from recordings of different durations.(Marek, 1996).

Frequency domain methods are power spectral analyses only looking at the variability. The total variance (the “power”) of a continuous series of beats is set into its frequency components. Low Frequency (LF) and High Frequency (HF) are mainly used, Very Low Frequency (VLF) is much less defined and the existence of a specific physiological process attributable to these heart period changes might even be questioned. LF range from 0.04–0.15 Hz, HF ranges from 0.15–0.4 Hz. LF and HF power components are usually made in absolute values of power (ms2), but LF and HF may also be measured in normalized units (n.u.). The representation of LFnu and HFnu emphasizes the controlled and balanced behavior of PNS and SNS. N.u. should always be quoted with absolute values of LF and HF power in order to describe in total the distribution of power in spectral components (Marek, 1996).

Non-linear methods received a growing interest in the past few years due to the hypothesis that ANS control is based on nonlinear dynamics (Montano et al., 2012). In 1996 the Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology stated a need of mathematical development to assess the full scope of these methods (Marek, 1996).

1.4.1. Measurements of HRV – knowledge from Kubios To present some of the newest publications on HRV, the Finnish article on the Kubios HRV – Heart rate variability analysis software is summarized. Kubios is developed at the University of Eastern Finland. The software is available free of charge at http://kubios.uef.fi.

Background

13

Kubios HRV is an advanced tool for studying the variability of heart beat intervals. Due to its wide variety of different analysis options and the easy-to-use interface, the software is suitable for researchers and clinicians with varying premises. The software computes all the commonly used time-domain and frequency-domain HRV parameters and several non-linear parameters. There are several adjustable analysis settings through which the analysis methods can be optimized for different data. The ECG derived respiratory frequency is also computed, which is important for reliable interpretation of the analysis result (Tarvainen et al., 2013). Looking through a Kubios tutorials(Medleg, 2011), the differences in HRV values between the supine position and the standing position is used to show that PNS is dominating the supine position and SNS gets active when standing up and SNS is more active in standing than in supine. It is stated that measurements are individual and measurements should always be compared to one self. However a table is presented with values from three groups; controls, trained and over trained persons. It indicates that some scale of indexing values for the ANS activity could be obtained, as over trained persons have much higher values in the supine position compared to the control.

Background

14

Figure 3 A screen shot from the Kubious program with analyzed HRV-parameters.

Background

15

Figure 3: Source: http://kubios.uef.fi/. The figure is meant to help get an overview of one possible way to analyze HRV from ECG. Starting from the top of the screen shot; The first “RR interval time series” shows the RR interval, the yellow is measured in a supine position and the white in a standing position, demonstrating the change in HRV follows change in postural position. The selected “detrended RR series” is a 5 minute RR series used to analyze the ECG into time-domain, frequency-domain and non-linear parameters: The time-domain results are listed with variables, units and values. Showing that the mean RR is 1221,7 ms, looking at the diagram of kolonnes showing the distribution it is clear that the variability is much larger than the width of the deviation, SDNN value is 47,1 ms, indicates. This is way RMSDD is used, the value in this RR-series is 60,5 ms. The number of NN-interval greater than 50 ms, NN50, is 108 counts and equals to 44,1 %, which is the pNN50 value. The frequency-domain results are taken from the AR spectrum. The advantage of FFT based methods is the simplicity of implementation, while the AR spectrum yields improved resolution especially for short samples, as this series is a short term sample of 5 minutes, the choice was to use the values from the AR spectrum to present the results. The power or the weight of a frequency; how much does the LF or HF contribute to the wave. The LF is 766 ms2 and LFnu is 45,8, whereas the HF is 930 ms2 and the HFnu is 54,8. The LF/HF is 0,824 ms2 saying that for every ms2 of HF there is 0,824 ms2 of LF.

The Non-linear results are showing the SD1 and SD2 in a Poincare plot. Every dot in the plot is a combination of two on each other following RR intervals. Therefore SD1 represent short term variability and SD2 represent long term variability. The value for SD1 is 42,9 ms and the value for SD2 is 51,1 ms.

1.4.3 Interpretation of HRV-parameters The taskforce describe HF to reflect PNS, while the LF, although more complex, is often assumed to have a dominant SNS component(Marek, 1996). Montanto et al. stated that HF is synchronous with respiration and agreed HF is a marker of PNS and LF is a marker of SNS(Montano et al., 2012). Rajendra et al. also agreed and explained the separate rhythmic contributions from SNS and PNS to modulate the R-R intervals

Background

16

at distinct frequencies. This difference in frequency ranges allows HRV analysis to separate SNS and PNS contributions evident. (Rajendra et al., 2006). Freeman states that LF reflects combined modulation of efferent PNS and SNS. Both the vagus and the cardiac sympathetic nerves mediate oscillations in HR at frequencies less than 0.15 Hz., therefore the respiratory frequency should be carefully controlled (kept at a rate at or above 0.15 Hz) to minimize the effects of respiratory modulation of HRV within this frequency band. Interpretation of LF power should be undertaken with caution, as also the interpretation that LF/HF represents the sympatogovagal balance(Freeman & Chapleau, 2013). According to Kubios the HF component is mediated almost solely by ANS activity, LF both SNS and PNS. Using the LF/HF then means PNS+SNS/PNS, which reasons that the value expresses the ratio of SNS to PNS. Also the LFnu could be used to asses SNS. (Medleg, 2011) Thayer agreed with Freeman that LF reflects both PNS and SNS and states that RSA, RMSSD and HF are closely related, and all reflect PNS(Thayer & Lane, 2009). The Task force also stated that RMSSD is related to HF(Marek, 1996). More experience and theoretical knowledge exists on the physiological interpretation of the frequency– domain measures compared to the time–domain measures(Marek, 1996). Freeman only describe frequency-domain measures, not time domain and not non-linear methods(Freeman & Chapleau, 2013). Several aspects of power spectral analysis of HR remain controversial or misunderstood. It is often inferred that spectral indices are a direct measure of PNS or SNS activity, whereas these indices in fact are a measure of sinus node activity(Freeman & Chapleau, 2013). The method selected should correspond to the aim of each study. When investigating short-term recordings Frequency–domain methods should be preferred to the time–domain methods. Recording of approximately 1 min is needed to assess the HF components of HRV while approximately 2 min are needed to address the LF component (Marek, 1996). The SNS effects are slow, on the time scale of seconds, whereas the PNS effects are fast, on the time scale of milliseconds. Therefore PNS can produce rapid changes. The differential effects of the ANS on the S-A node, and thus the timing of the heart beats, are due to the differential effects of the neurotransmitters (Thayer, Ahs, Fredrikson, Sollers, & Wager, 2012). HF unambiguously reflects PNS. PNS is stated certain to interpret, SNS is more difficult and uncertain.

Background

17

1.5 Theoretical model: Natures impact on humans explained by ANS and measured by HRV. On the basis of the information presented above a theoretical model was developed, with the purpose to inform decisions about the review question and what types of studies to review, to contribute to the interpretation of the review’s findings and to assess the generalizability, meanwhile providing a model of theory to test throughout the process (Popay et al., 2006) It is theoretically reasonable to believe there is an association between nature and health and the relationship could be causal as several studies have suggested through replications and consistency of findings and dose-response relationships. However, the research is still on a pioneer phase where it is not possible neither to close nor identify the width of the gap in research within a master thesis. But an attempt to test the theoretical model of causality in figure 4 is a step on the way to answer the question: How does nature affect human physiology? This ultimate goal is to be able to identify the causality on the molecule level, just as researchers documented that muscles are endocrine organs important for our immune system(Pedersen & Febbraio, 2012). Figure 4 (page 19): A diagram developed by the reviewer to visualize the theoretical model on natures impact on humans explained by the Autonomic Nervous System and measured by Heart Rate Variability. The diagram attempts to explain the reaction within ANS after nature exposure; follow the thick arrow.

Background

18

Figure 4: Natures impact on humans explained by the ANS and measured by HRV.

Non-deceased human adults

Viewing, smelling, hearing, sensing (e.g.wind) =nature exposure

Arousal heightened by stress drops

Follow the thick arrow Afferent pathways carry nerve impulses to CNS

Other physiological and pathophysiological factors Respiratory sinus arrhythmia (RSA )

Arousal heightened by stress drops

Psychological affect

CNS processes all impulses and the effect of the nature exposure dominates the efferent pathways in the ANS

Modulate immunesystem

ANS positively regulates the bodily homeostasis though visceral impact

PNS increases

SNS decreases

Other issues effecting ECG, e.g. individual training level

The sympathovagal balance increase HRV

Non-linear todmethods

ECG change

HF, RMSDD = PNS

Poincare plot: SD1 short term variability SD2 long term variability

Linear methods LF = SNS+PNS or SNS

LFnu may, represent SNS LF/HF ratio of SNS to PNS

Health improves

Background

19

1.6. The research question The research questions for this thesis are: What can be learned and concluded about ANS from studies using HRV to measure nature’s impact on humans? How can these conclusions explain a relationship between nature and health?

The questions will be instrumentalized through a systematic review (SR)

1.7 Delineations To focus on HRV measurement method to investigate the function of ANS, the reliability of the equipment is not included in reflection of the results, despite the fact it could change the results. It is assumed that the equipment used has provided reliable measurements across brands and that the persons handling the equipment are well educated. This thesis only deals with aspects relevant for healthy population making at scientific basis to make statements on concerns about public health promotion and disease prevention.

Background

20

Chapter 2 Method This chapter will outline the methodology of the systematic review (SR) explaining its relevance for answering the research question. The findings, discussion and conclusions presented in chapters 3-5 were determined by the methodological choices, therefore the reflections and decisions behind the review, are made explicit in this chapter, to give the reader a better assessment of the quality of the review results and conclusions. The first choice of guidance was the Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook); it provided an overall guidance to authors for the preparation of Cochrane Intervention reviews and included a stepwise description on how to conduct a full review with high standards. In this paper nature is an exposure and not an intervention, therefore the word exposure replaces the word intervention. In addition to the Cochrane Handbook, other methodological guidance resources were used as described in the reading guide for this thesis.

2.1 The systematic review Investigating physiological responses in the human body, when presented to nature environments, is a new area of health research. HRV can now be measured in field studies as well as in laboratory studies; therefore a SR was relevant to summarize and evaluate the experiences of different methods for data collection to optimize future studies. The conclusion from an earlier SR demanding more datasets to make a meta-analyses documenting the health benefits of the natural environment, measured by a relevant biomarkers gave merit to explore if enough datasets were available though a new SR.

Method

21

2.1.1 Objective Supported by the general necessity and purpose (Higgins & Green, 2011, pp. sektion: 1.2.1, 2.3.1) of SR described in Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook), the objectives for this SR was to: 

Identify, appraise and synthesize the research-based evidence on studies using Heart Rate Variability to measure the reactions within the Autonomic Nervous System, to investigate nature’s impact on humans and present it in an accessible format;



Explore variations in practice and map current practice of using HRV to measure ANS when humans are exposed to nature;



Highlight the needs for future research.

The above objectives should make this SR useful to develop new knowledge and be a brick in the puzzle of understanding the mechanisms of the human body´s interaction with the environment. This knowledge may contribute to targeted integration of nature interventions in public health in relation to the growing urbanization. No previous review´s on HRV-parameters used to document the human reactions to nature exposure was found. Some recommendations for reviews could not be met in this thesis; being two review authors, and using a Trials Search Coordinator. This was substituted by the possibility to consult a professional university librarian and the thesis supervisor and co-supervisor. Neither was there an expert on ANS to consult; nor practical experiences with similar studies or specialist knowledge on HRV. The advantage was avoidance of pre-formed opinions that could bias experts’ assessments of both the relevance and validity of articles(Booth, Papaioannou, & Sutton, 2012).

2.1.2 Procedure An effort was made to follow the steps suggested in the Cochrane handbook when doing a review (Higgins & Green, 2011, p. section 1.2.2): 

A systematic search that attempts to identify all studies that would meet the eligibility criteria;

Method

22



An assessment of the validity of the findings of the included studies, for example through the assessment of risk of bias;



A systematic presentation, and synthesis, of the characteristics and findings of the included studies.

To finish the thesis within the 6 month timeframe, some choices had to be taken considering the efficient use of time. The choices and how the above mentioned steps were handled in this thesis are described in sections 2.2 – 2.7.

2.2 Reference management Endnote was used from the early beginning to save all relevant information and take advantage of the technology to export citations from the databases and find full texts automatically. The literature was categorized into folders; HRV, ANS, nature interventions, outdoor life, physiological biomarkers, public health surveys, and reviews on nature interventions, studies relevant for the systematic review, and studies relevant for the discussion.

2.3 Identification of studies Identifying reports of studies is currently the most convenient approach to identifying the majority of studies and obtaining information about them and their results; therefore that was the aim of the search, well knowing that the ideal way was to use data material from studies, not reports of studies, to include in SR (Higgins & Green, 2011, p. section 6). In the selection of studies caution was taken to identify if two or more reports represented the same study. Some studies and reviews of different characters had already been found, exploring the state of the art on nature and health through the homepages of the Danish national health service and the project “Fresh in Nature” financed by the Nordic Council of Ministers. These starting point studies were used in the first step of the search strategy.

Method

23

Studies eligible for the review should somehow investigate how HRV-parameters could be used to measure human reactions to natural environments. The issues can be outlined by following PICO-method, used as criteria for considering studies for this review: People: Healthy adult people Intervention/Exposure: Nature Comparison: Urban, city, or man-made environments can be used as a contrast for the intervention Outcome: HRV-parameters

2.3.1. Types of studies When conducting a systematic review, one should always strive to include only the best available evidence. The best or highest level of evidence in epidemiological research is generally accepted to originate from randomized controlled trials (RCT); because they are more likely to provide unbiased information. The issues to be investigated set the stage for experimental studies and open the possibility to only use RCT´s; however, an earlier review on curative and health enhancing effects of natural environment only found two randomized controlled trials, but excluded seven crossover designs; the review was published in 2012 and included all health effects. (Kamioka et al., 2012). Other reviews on nature’s impact on humans also showed an overall low number of studies using HRV as measurement (Bowler et al., 2010; Keniger, Gaston, Irvine, & Fuller, 2013). Therefore it was expected to mainly find studies recently published. To be sure to find all possible studies no limits were set for study design or languish.

2.3.2 Search strategy Step 1 To investigate as many varieties of search terms as possible, the starting point studies were screened for related and citing articles in PubMed; snowballing, Appendix IV(page89). Relevant search terms from the titles, abstracts, keywords and MeSH terms (HRV, had no MeSH terms) relevant to the research question were extracted and synonyms and related terms were found.

Method

24

According to the Cochrane Handbook, too many different search concepts should be avoided, but a wide variety of search terms should be combined with OR within each concept. The Cochrane Handbook does not advice to use the PICO method rigorously, because the concepts of PICO may not be well described in the title or abstract of an article and are often not well indexed with controlled vocabulary terms and thus do not lend themselves well to searching. Therefore, depending on the review´s research question it is better to have a search strategy with typically three sets of terms, in this case: (1) terms to search for the effect, (2) terms to search for the exposure evaluated; and (3) terms to search for the health condition of interest. (Higgins & Green, 2011, p. section 6.4). The health condition was translated to include the outcome under investigation. This search strategy is called a block search by the university librarians, illustrated in Table 1. After gaining experience through experimentation and advice from the Librarians, to strike a balance between striving for comprehensiveness and maintaining relevance when developing the search strategy, a block search was chosen for the systematic search strategy applying Boolean search phrase as follows: Table 1: Strategy for the block search

Block 1

Block 2

Block 3

effect OR effects OR affect

Park OR parks OR Green* OR Natural

HRV OR stress recovery OR

OR Affects OR response* OR

environment OR natural

autonomic* OR Physiolo* OR stress

Recovery OR Physiological

environments OR Open space OR

OR heart rate OR Heart Rates OR

validity OR Convalescence

open spaces OR Garden OR gardens

Pulse Rate OR Pulse Rates OR

OR influence OR outcome

OR gardening OR Horticulture OR

cortison OR sympathetic nervous

OR outcomes

Wilderness OR ecotherapy OR

activity OR parasympathetic

Countryside OR Outdoor* OR Nature

nervous activity OR Cardiac

OR nature sounds OR Biological

Chronotropic OR Cardiac

Diversity OR Biodiversity OR Wood

Chronotropism OR Heart Rate

OR woods OR tree OR trees OR

Control OR heart rate variability OR

Allotment OR allotments OR Forest

Higher Nervous Activity OR

OR forests OR virtual environment

nervous activity OR Higher Nervous

OR rual

Activities OR mental health

Using the filters the search was limited to last 5 years, peer reviewed, adults, and humans.

Adults were chosen because HRV is different in children. Method

25

The systematic searches were made in; (OvidSP) = PsycINFO, Global health, Embase and PubMed Medline and (Ebsco) = Amed, CINAHL and Web of science. The databases were chosen on the basis of previous reviews on nature and health and under supervision of the librarians. Slightly different approaches were used in the different databases; however, most often records were searched in all text fields or in the title, abstracts and keywords, depending on the database. Table 2: Show the search strategy. More details on the exact search in each database are found in Appendix V (Page 90). Step 2 Then a search through most cited key articles from the first authors who explored a relationship between nature and health; Stephan and Rachel Kaplan, Roger Ulrich R. and Terry Harting, was conducted. Appendix VI (Page 91). Step 3 Snowballing from identified studies was conducted. Step 4 Experts were contacted: The Japanese professor Yoshifumi Miyazaki replied with information on the book: Forest Medicine. Public Health in the 21st century/Environmental Health - Physical, Chemical and Biological Factors. Edited by Li, Qing published in 2012 by Nova Biomedical, Nova Science Publishers, Inc. in New York. Step 5 Alerts from the databases were activated for weekly updates in case new studies matched the conducted search.

Method

26

Table 2: The search strategy to identify eligible studies for inclusion in the review

Name “Presearch” Step 1: Searching databases PubMed Medline PsycINFO (OvidSP) Global health Embase AMED (Ebsco) CINAHL Web of science

Number of hits 3125 (all together) 643

Potentially relevant articles based on title 12

952 1092

20 16

146

2

292

Total (when removing report duplicates) Name Number of hits

13 59 23 Number of additional articles identified

Step 2: Searching authors using Web of Science Hartig, T (2003) 157 1 Ulrich, RS (1991) 210 0 Ulrich, RS (1984) 259 1 Kaplan, S (2008) 99 0 Kaplan, S (1995) 233 0 Step 3: Searching references Each of the 25 studies identified were reviewed in their reference list in order to locate additional relevant studies. Additionally, 18 studies were screened. No additional studies were however found during this step.

2.4 Eligibility criteria - Selection of studies One of the features that distinguish a systematic review from a narrative review is the pre-specification of criteria for including and excluding studies in the review – i.e. the use of eligibility criteria. Eligibility criteria are a combination of aspects of the research question plus specification of the types of studies that have addressed these questions. This review was an exception from the normal procedure by translating all the PICO- issues in the research question directly into eligibility criteria, because it legitimately restricts eligibility to a specific outcome – HRV-parameters, in order to meet the objective of the review. Eligibility criteria set out boundaries for which types of studies potentially can be included in the review. Any eligibility criteria should balance between being sufficiently broad to encompass the diversity of studies, but also sufficiently narrow to ensure that a meaningful answer can be obtained when studies are considered in aggregate.(Higgins & Green, 2011, p. section 5.1.2)

Method

27

The eligibility criteria for this SR were narrow for the outcome measure; HRV-parameters, compared to the exposure; nature, which can be represented both by virtual environments, green parks, forests, nature sounds and the comparison which can be urban, city, or man-made environment. A pilot test of the eligibility criteria on a sample of three reports was done. Taken the novelty of the research issues into account, it was estimated time-efficient to extend the pilot test with extraction of data from eleven studies, and to use the pilot test both to refine and clarify the eligibility criteria and start the assessment of the content for data extraction. Appendix VII (page 92). In the following sections the eligibility criteria translating all the PICO- issues in the research question for this SR are specified in further detail: 2.4.1 People - Study participants The Cochrane Handbook, section 5.2, recommends two steps to follow when considering the types of people that are of interest. First, the diseases or conditions of interest, in this case healthy people. It is important that no diseases have influence on the heart rate. Second, the broad population and setting of interest should be defined. This involves deciding whether a special population group is of interest, determined by factors such as age, sex, race, educational status is important. The study population for this SR is all adult people.

2.4.2 Outcome The outcome is the most narrow eligibility criteria and the studies included must have HRV-parameter reported in a sufficiently detailed way, so that it was clear, how it was handled and what the results were.

2.4.2 Exposure An important step in setting out eligibility criteria for the studies to be included in the SR was defining the exposure of interest. Nature is difficult to define; this review sets a broad definition. Therefore anything representing nature was accepted as eligible: wild nature e.g. forest, virtual environments, urban parks, green space, flowers, and also sounds and odor were acceptable. The criteria were nature in all forms.

Method

28

2.4.3 Comparison The comparison works as control opposing the exposure to make it possible to measure a difference and document the effect of nature. Built, virtual or artificial environments were accepted as eligible.

2.5 Exclusion criteria The list of excluded studies, detailing any studies that a reader might plausibly expect to see among the included studies, covers all studies that may on the surface appear to meet the eligibility criteria but on further inspection do not, and also those that do not meet all of the criteria but are well known and likely to be thought relevant by some readers. The purpose of listing such studies as excluded and giving the primary reason for exclusion, is to show that consideration has been given to these studies. The list of excluded studies is made as brief as possible. Fejl! Henvisningskilde ikke fundet. (page 97)(Higgins & Green, 2011, p. section 7.2.5) The exclusion criteria selected are listed in the following. For the brief list mentioned above the exclusion criteria or the inclusion criteria not met are described:

2.5.1 Exclusion based on language The systematic review was furthermore limited to studies described in publications available in English, as no resources were available to translate important information on methods and results of studies that was reported in languages other than English. It was important to know if there were many studies made in countries that not traditionally publish their work in English. But if it was not possible to get the full text in English the study had to be excluded. Also excluding studies described only in non-English publications increased the risk of introducing bias into the review. Language bias occurs because statistically significant results are more likely than insignificant results to be published in English literature. (Higgins & Green, 2011, p. section 10.12.12.14)

Method

29

2.5.2 Exclusion based on duplicity A systematic review is a review of studies that meet pre-specified criteria for inclusion of studies in the review. Since each study may have been reported in several articles, abstracts or other reports, a comprehensive search for studies for the review may identify many reports from potentially relevant studies. Two distinct processes were used to determine which studies to include in the review; linking together multiple reports of the same study and using the information available in the various reports to determine which studies were eligible for inclusion (Higgins & Green, 2011, p. Section 7.2.1). Duplicate publications can introduce substantial biases if studies are inadvertently included more than once in a meta-analysis. It can be difficult to detect duplicate publications; the following checklist from the Cochrane Handbook was used: 

author names



location and setting



specific details of the intervention



numbers of participants and baseline data



date and duration of the study.

By applying the eligibility criteria described in section 2.4 and 2.5 a total of eight studies remained for inclusion in the review.

Method

30

Figur 5 Flowchart illustrating the identification and selection of studies. 23 potentially eligible studies identified in Pubmed, OvidSP, Ebsco and web of science

2 potentially eligible studies identified Searching Authors using web of science

No additional studies identified searching thought citations, related articles and citing articles

25 potentially eligible studies identified 1 Excluded based on language 8 excluded based on abstract: - Outcome was not HRV 16 potentially eligible studies to be reviewed in full text

6 excluded because of duplicate data 1 excluded because exposure not nature 1 excluded because no full text available 8 potentially eligible studies identified

Figur 5: Flowchart illustrating the overall identification and exclusion of studies for the SR.

2.6 Data extraction The eight studies included in the SR were RCT’s from which data had to be appraised, synthesized and analyzed. The definition ‘data’ is any information about (or deriving from) a study, including details of methods, participants, setting, context, exposure, outcomes, results, publications and investigators. The data extraction must enable appraisal of the included studies, to make a quality assessment in terms of validity and reliability (both internal validity) and applicability/generalizability(external validity). Synthesizing and analyzing of the content in the included studies must explore similarities and variations in methods and results, in order to estimate the total effect and to highlight the needs for future research(Higgins & Green, 2011, p. section 1.2.1). The first step to assess the evidence base by extracting data from each included study must be done in a systematic way. Every study is reported differently, according to e.g. requirements of the publishing journal (Booth et al., 2012, p. 101).

Method

31

Data extraction is one of the key elements in the SR to detect and understand relationships between the variables (Booth et al., 2012). Nature was the independent and HRV the dependent variable in this SR, even though not used to make any analyzes, as none of the results made it possible. It was chosen to make the quality assessment simultaneously, because it took advantages of the fact that both data extraction and quality assessment require in-depth reading of a study. The systematic procedure of the data extraction is presented in Figure 6 (page 32).

Crochrane Handbook, to estimate internal validity in terms of risk of bias

NPT – a CONSORT Statement, to be systematic 

RevMan

and estimate the internal validity in terms of quality assessment, with emphasis not to produce critical appraisal "nihilism"



Data sheets

Also shoosing data to extract based on the pilot test + model of theory



Result sheet

Data extraction of the eight included studies

Simply extracting the results from the included studies, to start the estimate of Heterogeneity

Figure 6 Procedure for data-extraction

Internal validity in terms of reliability of significant results

Method

32

Figure 6: RevMan is developed by Cochrane Collaboration and it links directly to the Cochrane handbook. For the appraisal the RevMan includes a table for the risk of bias assessment; selection bias, performance bias, detection bias and reporting bias; all systematic errors, but also internal validity depends on random error, imprecision and confounding(Higgins & Green, 2011, p. section 8.2.1). As all studies were RCT´s; confounding due to study participants were taken care of. Confounding due to the setting was implemented in the data extraction sheets. Quality of reporting (imprecision) only focused on issues important for evaluating the results. Reliability (random error) was evaluated by the significance of the results in the results sheet. Furthermore, issues within cross-over trails turned out to be relevant, as seven out of eight studies were crossover trails. Therefore the described risks in part three “special topics” within chapter 16 section four of the Cochrane Handbook was added to the table of risk of bias assessments in RevMan. The purpose of the risk of bias assessment is to identify the limitations of the primary research and appraise the research-based evidence from the included studies, because the key source of potential bias in the SR stems from limitations in the original studies included in the analyses in a SR (Booth et al., 2012). The choice to assess the risk of bias and not do a full assessment of methodological quality, as the CONSORT recommends, were based on the Cochrane Handbook, which discusses the CONSORT statement and distinguishes between assessment of methodological quality and assessment of risk of bias; and recommends a focus on the latter. A study may be performed to the highest possible standards, yet it still can have an important risk of bias. The Cochrane Collaboration recommends a tool for assessing risk of bias that is neither a scale nor a checklist, perfering a domain-based evaluation, in which critical assessments are made separately for different domains. Because it is impossible to know the extent of bias (or even the true risk of bias) in a given study, the possibility of validating any proposed tool is limited. The most realistic assessment of the validity of a study may involve subjectivity (Higgins & Green, 2011, p. section 8). The systematic work assessing risk of bias in the included studies of this SR can be read in Appendix IX (page 98). RevMan included features for a graphic presentation and meta-analyses, meanwhile making it possible to write the SR in a framework ensuring to comply the guideline of the Cochrane Handbook which was the reason to download and make use of the program. Later it turned out that it was not compatible to print

Method

33

and hand in a master thesis, therefore the program was only used to make the risk of bias assessment, however the graphic presentation of the risk of bias assessment was not possible to transport into a word document and printing and scanning made it unreadable, therefore a regular figure was made in words and the results typed in. This added a risk to make a typing mistake, but the figure was carefully checked by three people at separate times. Data sheets of data extraction made for this SR, have an intent to be interpretative - based on the objectives of the SR (Booth et al., 2012, pp. 128-131). The systematic basis was formed by using the Checklist of Items for Reporting Trials of Non- Pharmacologic Treatment from the CONSORT Statement (NPT) Appendix X (page 103). Even though The Consort Statement was opted out as an independent tool to assess the quality of the included studies, it was a relevant tool for systematic data extraction, and included important and useful information to estimate the variations in the findings and access the generalizability (external validity) of the SR. Seeking to answer the research question, an attempt to predict, what would be useful findings in the included studies, besides the inclusion criteria was based on the pilot test (Appendix VII) and the model of theory; Figure 4: Natures impact on humans explained by the ANS and measured by HRV. (page 19). Some information from assessing risk of bias was also included in the sheets; the focus was to fill in all relevant data before starting synthesizing and analyzing. Throughout the data extraction, issues were added, when important topics were found in a study. A table with rows for each study and column for each topic, giving a field to enter the quotation describing the extracted data, was made. Nine sheets were made altogether, each with a theme; analyses, authors, background, conclusions, measurements, method, overview, participants and setting. A “?” was added to the sheet in a cell with missing data, due to lack of reporting in the included study. Appendix XI (page 105) The data from the eight studies included in the review were extracted by thorough assessment of each study. In some instances secondary publications of the same study were necessary in order to gather all relevant information on the study. Result sheet was extracted to assess the reliability and begin estimating heterogeneity. Causal relationships are also determined by the strength of association, which is another reason to investigate the effect size. As many details on the results as possible were extracted with considerations on the practical reliability as well. Where it seemed reasonable, a calculation of the effect size in percentage was conducted from the presented result, to produce similar and comparable effect sizes. Results from conducted calculations were marked in red. P-value