cognitive performance, fatigue, emotional and

Cognitive performance, fatigue, emotional and physiological strains in

simulated long-duration flight missions

Eduardo Rosa1,, Eugene Lyskov

2, Mikael Grönkvist

5, Roger Kölegård

5, Nicklas Dahlström

4, 1

Igor Knez1, Robert Ljung

3, Johan Willander

1 2

1 Department of Psychology, University of Gävle, Gävle, Sweden 3

2 Centre for Musculoskeletal Research, Department of Occupational Health Sciences and 4

Psychology, University of Gävle, Sweden 5

3 Department of Environmental Psychology, University of Gävle, Gävle, Sweden 6

4 Lund University School of Aviation, Lund University, Lund, Sweden 7

5 Division of Environmental Physiology, Swedish Aerospace Physiology Center, KTH Royal Institute 8

of Technology, Stockholm, Sweden 9

10

* Correspondence: 11

Corresponding Author 12

Eduardo Rosa 13

[email protected] 14

Keywords: fatigue, heart rate variability, cognitive performance, emotions, long-duration 15

military missions 16

Abstract 17

Pilots in long-duration flight missions in single-seat aircrafts may be affected by fatigue. This study 18

determined associations between cognitive performance, emotions and physiological activation and 19

deactivation – measured by heart rate variability – in a simulated 11-hours flight mission in the 39 20

Gripen aircraft. Eleven participants volunteered for the study. Perceived fatigue was measured by the 21

Samn-Perelli Fatigue Index (SPFI). Cognitive performance and objective fatigue were measured by 22

non-executive and executive tasks. Emotions were assessed by the Circumplex Affect Space 23

instrument. Heart rate variability (HRV) was considered in relation to the cognitive battery test in 24

four time points – Hours 3, 5, 7, 9 – and their associations with emotional ratings. Results indicated 25

decrease in performance in the non-executive task after approximately seven hours. This result was 26

correlated with self-reported measures of fatigue. Heart rate variability, assessed by indices of 27

parasympathetic modulation – RMSSD, pNN50 and mean RR Interval – remained unchanged for 28

both non-executive and executive tasks over time. There were associations between increased 29

boredom as well as passiveness and decrease in stimulation as well as activeness, and increased 30

HRV. This suggests that a low self-regulatory effort for maintaining performance in these tasks in 31

these environmental conditions was required. Combined results indicate that pilots may be able to 32

adapt to environmental demands and fatigue in long-duration missions providing that low self-33

regulatory effort is prevalent. 34

35

Fatigue in long-duration flight missions

2

This is a provisional file, not the final typeset article

1 Introduction 36

Military pilots operating in single-seat multirole fighter aircrafts are exposed to a demanding 37

cognitive work environment, implying intense physical and psychological stress and fatigue (Driskell 38

and Salas, 1991). In such environments, vigilance decrements, complacency and loss of situation 39

awareness have long been identified as issues related to the increasing levels of complexity in 40

automation and other human-machine system errors (Parasuraman et al. 1993; Endsley and Kiris, 41

1995; Parasuraman and Riley 1997). The continuous increasing complexity of information 42

processing coupled with the changing operational characteristics of missions are factors that might 43

affect the performance of fighter pilots. One particular change in operational characteristics of multi-44

role fighter aircrafts relates to the duration of flight missions. Flight durations can now be prolonged 45

due to modern fighter aircrafts’ engineering features and purposes of employment, e.g., air policing 46

operations aided by air-refuelling or missions comprising multiple stops can now last more than 6 47

hours (personal communication, June 2018). Hence, the importance in assessing fighter pilots’ 48

cognitive performance, emotional strains and physiological variations in present long-duration 49

missions is paramount. Addressing pilots’ process capacity within these environmental 50

characteristics can support the design of missions with extended flight time, where fatigue may act as 51

a contributing factor to overall pilot performance. 52

The main purpose of this study is to investigate the effects of long-duration missions on cognitive 53

performance, fatigue, emotional and physiological strains. We examined mental fatigue influence on 54

cardiovascular (CV) responses throughout four time points during the 12 hours experimental session. 55

We have analysed heart-rate variability (HRV) in relation to cognitive task performance measures, 56

self-reported measures of fatigue and CV associations with emotional states. 57

In the military aviation context, fatigue-related errors during daytime operations may induce 58

performance degradation when pilots are exposed to extended periods of flight duty (Neville et al., 59

1994). Recently, Rosa et al. (2020) suggested that performance degradation in single-seat aircrafts 60

may occur in tasks targeting arousal and sustained attention after approximately seven hours into the 61

mission. Corresponding self-reported measures of fatigue with performance in response times were 62

also observed. Another possible indication of performance measurement relates to emotional ratings. 63

Rosa et al., (under review) have suggested that positive emotional ratings decrease and negative 64

emotional ratings increase also after about seven hours in long-duration missions. 65

The complex interplay between cognitive, emotion and also physiological regulation of goal-directed 66

behaviour is supported by a joint cortico-subcortical neural circuit (Grol & Raedt, 2020). These 67

processes are related to each other for adaptability to changing environmental demands, and there is 68

an overlap of the central autonomic network and neural circuits that support cognitive and emotion 69

regulation processes (Thayer, Hansen, Saus-Rose, & Johnsen, 2009; Thayer & Lane, 2000). The 70

balance between the excitatory sympathetic and inhibitory parasympathetic activity subsystems of the 71

autonomous nervous system (ANS) often interact in an opposing fashion, yielding varying degrees of 72

physiological arousal (Grol & Raedt, 2020). During physical and psychological stress, activity of the 73

SNS prevails, generating physiological arousal that will help us to adapt to the challenge up front. 74

Conversely, during conditions in which safety and stability is prevalent, the PNS is dominant, 75

maintaining a lower degree of physiological arousal (Appelhans & Luecken, 2006). Parasympathetic 76

inputs to the heart affect heart rate, such as more parasympathetic input reflects in more pronounced 77

acceleration and deceleration and more variable intervals between heart beats, that is, higher HRV 78

(Segerstrom & Nes, 2007). In this sense, an optimal HRV level is usually associated with self-79


3

regulatory capacity, and adaptability or resilience according to changing environmental demands 80

(Beauchaine, 2001; Berntson et al., 2008; Shaffer, & Ginsberg, 2017). 81

HRV addresses neurocardiac function and variations reflect heart-brain interactions and dynamic – 82

usually non-linear – ANS processes (Shaffer, & Ginsberg, 2017). HRV differences throughout time 83

reflect regulation of interdependent regulatory systems – autonomic balance, blood pressure, gas 84

exchange, gut, heart and vascular tone – in assisting in the adaptation during environmental and 85

psychological challenging conditions (Shaffer, & Ginsberg, 2017). The variability of non-linear 86

systems is one of the mechanisms of physiological adaptation that help us to cope with uncertain and 87

changing environments (Beckers et al., 2006). 88

HRV has been used to investigate fatigue influence on mental effort and associated cardiovascular 89

responses in challenging circumstances (Wright, & Stewart, 2012). Mental effort may be divided into 90

‘task effort’ and ‘state effort’ (Mulder, 1986). Task effort relates the properties of the task itself (task 91

difficulty and associated amount of controlled processing required). State effort relates to the amount 92

of effort required to protect task performance from negative influences of fatigue, environmental 93

pressures, etc. 94

As per the principle of HRV levels associations with adaptability, HRV may indicate self-regulatory 95

efforts in overcoming fatigue, particularly when an individual is under stressful tasks that involve 96

high level of mental load (Gonzalez et al., 2017), either by increases in task effort or by increases in 97

state effort. The principal premise is that fatigue may increase, decrease or have no effect on effort 98

and cardiovascular responses, and this depends on the difficulty of the impending challenge (LaGory 99

et al., 2011) 100

Hence, from the cognitive perspective, when the individual believes that success in achieving tasks’ 101

goals is possible, he or she will exert compensatory (self-regulatory) efforts to overcome fatigue 102

effects, experiencing increased arousal. On the other hand, when the individual does not believe that 103

success in achieving tasks’ goals is possible, people will suppress efforts to overcome fatigue effect, 104

experiencing minimal CV arousal. Lastly, when the individual beliefs on the impossibility in 105

achieving tasks’ goals is reinforced by fatigue, he or she will employ low effort, also displaying 106

minimal CV arousal (LaGory et al., 2011). This is an interesting mechanism as it forms the basis for 107

the variety in CV response patterns in relation to fatigue in different environments. 108

Certain studies on fatigue indicate that HRV is positively associated with mental fatigue state, i.e., 109

HRV increases as mental fatigue state increases, depending on the type of tasks (Ahsberg et al., 110

2000; Zhang & Yu, 2010; Huang et al., 2018). Also, HRV appears to index self-regulatory effort 111

when individuals are fatigued, showing that a parasympathetically mediated system is associated 112

with self-regulation (Segerstrom & Nes, 2007). This means that HRV is elevated during high self-113

regulatory effort compared with low self-regulatory effort. Another study pointed that HRV was 114

significantly increased due to time-on-task (Fairclough & Houston, 2004), whereas a more recent 115

finding indicated that HRV negatively associated with mental fatigue, i.e., decrease in 116

parasympathetic activity (RMSSD and pNN50), due to time-on-task effects (Melo et al., 2017). Tran 117

et al. (2009), however, observed that parasympathetic modulation such as RMSSD and pNN50 118

remained stable as participants experienced fatigue in a simulated driving task addressing 119

alertness/arousal, suggesting that fatigue may be associated with increased in sympathetic arousal. 120

Other studies pointed that CV responses in fatigued individuals can vary greatly according to 121

individual characteristics and environmental conditions (LaGory et al., 2011; Nolte et al., 2008). 122


4


Previous studies have also indicated changes in HRV during emotion regulation processes. Emotion 123

regulation is an attempt to change the present moment emotional experiences, behavioural 124

expressions and physiological responses (Gross, 1998). Thus, emotions experienced when humans 125

interact with their environment are associated with varying degrees of physiological arousal 126

(Levenson, 2003). As mentioned, HRV is usually associated with self-regulatory capacity and 127

depends on the environmental conditions. In this sense, emotion regulation depends on the 128

individual’s ability to continuously adjust physiological arousal (Gross, 1998) in accordance with 129

situational demands. HRV is, then, a measure of the interplay between sympathetic and 130

parasympathetic activity, providing information about the capacity for regulated emotional responses 131

(Grol & Raedt, 2020). 132

In general, there is a lack of consensus in the literature regarding HRV results concerning fatigue, 133

cognitive task performance and emotions. It seems that type of tasks (non-executive and executive 134

tasks), time-on-task, individual experiences and environmental circumstances play a role in HRV 135

results, with varying conclusions regarding the role of sympathetic and parasympathetic activity 136

related to mental fatigue and emotions. 137

Hence, the understanding of activation of ANS in long-duration flight missions – and its association 138

with cognitive performance and emotional experiences – may contribute to an integral assessment of 139

the psychology of pilots’ performance in such complex environment. 140

Regarding cognitive performance, considering that our experiment design did not involve time-on-141

task but challenging environmental conditions (physical limitations, discomfort and performance of 142

cognitive battery test in different points in time), we expected that participants would become 143

fatigued while trying to sustain cognitive performance (increase in self-regulatory effort), reflecting 144

an increase in HRV over time. Hence, the predictions of increase in HRV over time refers to the 145

assumption that participants would view success when performing the tasks as possible, exerting self-146

regulatory efforts to overcome fatigue effects. 147

Regarding emotions, we assumed that participants’ HRV would be associated with emotional ratings, 148

reflecting the attempt to self-regulate emotions according to environmental pressure. 149

Exact predictions regarding CV measures were problematic to estipulate, given the difficulty in 150

stipulating how fatigued participants would be and how much effort they would be willing to employ 151

when performing the tasks. In spite of this limitation, we formulated the following hypotheses: 152

H1. Effort-related CV responses would reflect fatigue development, implying that higher HRV is 153

positively correlated with fatigue (increased parasympathetic activity), in both non-executive and 154

executive components of task performance. 155

H2. Effort-related CV responses is associated with emotional states, such that higher HRV is 156

negatively correlated with positive emotions, and positively correlated with negative emotions. 157

2 Methods 158

This study is part of a larger research project. The methods section in this study is in accordance with 159

at least one previous publication (Rosa et al., 2020) within the project concerning the effects of long-160

duration flight missions. This study complies with the Declaration of Helsinki and was approved by 161

the Medical Research Ethics Review Board, in Stockholm (approval no:2018/806-3). 162


5

2.1 Participants 163

Participants comprised of twelve healthy individuals, and consisted of 10 men and 2 women. Mean 164

(SD) age, height and body mass were 28.2 (± 6.0) years, 179 (± 6.8) cm, 79.5 (± 16.9) kg, 165

respectively. Informed consent was obtained prior to participation. Participants could stop the 166

experiment at any time if they wished to do so. 167

2.2 Study Design 168

Experiments were conducted at the Flight Physiological Centre in Linköping, Sweden. A high-169

fidelity Dynamic Flight Simulator – DFS (Wyle Laboratories Inc., El Segundo, CA) was used. It 170

comprises of a 9.1m radius centrifuge equipped with a 39Gripen flight simulator mimicking the 171

aircraft controls and models. The hardware comprises of, e.g., a Martin-Baker seat (Martin Baker 172

Aircraft Co. Ltd., Middlesex, UK), stick and throttle (Page Aerospace Ltd., Middlesex, UK), and G-173

valve/breathing regulator (PSU BRAG valve, Honeywell Aerospace, Yeovil, UK). The total duration 174

of the experiment was 12 consecutive hours – calculated from the first cognitive test battery until the 175

last cognitive task battery. Both these tests occurred in a computer outside the DFS. The total mission 176

time was, then, 11 hours, considered by the time participants were strapped inside the DFS. 177

Participants wore a full-coverage anti-G suit, helmet with a mask covering the mouth and nose 178

(116H; Saab Ltd.), life jacket fitted with a chest bladder for assisted positive pressure breathing and 179

gloves (for details see Eiken et al., 2007) throughout the total duration of the experiment. Participants 180

were strapped with a 6-point seat belt while in the DFS. Participants could urinate by using an in-181

flight aircrew urine bladder relief system (AMXDmax®, Omni Measurement Systems, Inc, Milton, 182

USA) (Figure 1). Food was limited to three energy bars, three protein bars. Water was limited to 183

1.5L for the duration of the experiment. Intake were ad libitum. Front- and back-facing cameras were 184

used to monitor participants actions. 185

186


6


187

Figure 1. The gondola depicting a participant strapped in the 39Gripen seat simulator. Mask and 188

helmet were used uninterruptedly for 11 consecutive hours. Cognitive tests were presented on the 189

center screen head-down display. Emotional rating scale and self-reported measure scale of fatigue 190

was affixed in the upper right corner from the central display (not visible). 191

192

A within-subjects, repeated measures design was used. Dependent variables were cognitive test 193

performance, self-reported measures of fatigue and heart rate variability. Independent measure was 194

physiological load (fatigue development) up to the 12th hour of the experiment. Level of significance 195

was p < .05 for statistical differences for all analyses. Bonferroni multiple-comparison correction was 196

used as post-hoc analysis. 197

198

2.3 Measures 199

2.3.1 Fatigue 200

Self-reported data were collected by the 7-point Samn-Perelli Fatigue Index (Samn & Perelli, 1982). 201

2.3.2 Cognitive tasks 202

Five cognitive tasks were part of the cognitive test battery. These tasks are representative of pilots’ 203

cognitive processes in the flight operational environment. 204

Psychomotor Vigilance Task (PVT). This task is a response time (RT) task to stimuli presented at 205

random intervals. It addresses changes in arousal and sustained attention deterioration associated 206


7

with fatigue-related variations and time on task (Dinges & Powell, 1985). A blank screen is initially 207

presented, followed by fixation cross (200ms) in the center of the screen. Next, the fixation cross 208

disappears. Then, at random intervals, a red circle appears in the center of the screen. Participants’ 209

task was to respond to the appearance of the red circle as fast as possible by pressing the space bar on 210

a keyboard. Inter-stimulus interval varied between 2 to 12s. Response latency was recorded. The 211

duration of the task was set for 5 minutes. 212

Visual Search Task. This task assesses targeting saliency and contextual cueing (see Treisman, 213

1985; Eckstein, 2011; Emrich et al., 2009). A dispersed matrix of a set of letters (‘U’, ‘D’, ‘G’, ‘C’, 214

‘Q’ – distractors) and one of the letters (‘X’ or ‘O’ – target) is initially presented (randomly 10, 30 or 215

50 programmed screen load weight). Participants’ task was to quickly identify one of the two 216

randomly presented targets (‘X’ or ‘O’, between trials) among the distractors. Reaction time and 217

accuracy were recorded. The task comprised 36 trials and lasted approximately 3.5 minutes. 218

Match-to-sample Task. This task assesses pattern recognition and short-term memory (Lieberman et 219

al., 2002; Shurtleff et al., 1994). A 6 x 6 matrix of red and yellow checkerboard pattern appears on 220

the screen for 1000ms. It then disappears for 3500ms. Then, two matrices appear simultaneously next 221

to each other; the original and a second (similar) matrix. Participants’ task was to select the matrix 222

that matched the original sample. Response latency and accuracy of target detection were recorded. 223

The task comprised 30 trials and lasted approximately 3 minutes. 224

N-back Task. This task assesses working memory (Kane et al., 2007). Participants’ task was to 225

acknowledge whether the stimulus (a number) in the sequence matched the one that appeared n 226

(number of) items before. We used only numbers as a stimulus as this involves the communication 227

characteristics of flight environment. Three blocks were performed for each session, varying in 228

memory load; 1-back, 2-back and 3-back (number of items), in this order. Each stimulus (number) 229

appeared for 500ms. The inter-stimulus interval was 2000ms. Accuracy for each task was recorded. 230

The task contained 50 trials per block, which was equivalent to approximately 1.5min per n-back task 231

(all three tasks lasted, then, 5min30s). 232

Mental Rotation Task. This task assesses internal representation of an object in space as it involves 233

imagining analogue space representations (Shepard & Metzler, 1971; for a review, see Zacks, 2008). 234

First, two equal figures appeared on the screen at the same time. Participants’ task was to identify 235

whether the images were rotated or whether they were mirrored in relation to each other. Response 236

was done by pressing the left shift button to identify ‘rotated’, and right shift button to identify 237

‘mirrored’. Decision time was set to a maximum of 5s. Response latency and accuracy for choice 238

were recorded. The duration of the task was 3 minutes. 239

240

2.3.3 Emotions 241

Emotional experience was acquired and assessed by the Circumplex Affect Space (CAS) instrument 242

(Knez & Hygge, 2001; Larsen & Diener, 1992). Eight affect states are represented in a circle 243

comprised of eight octants (Figure 2). The horizontal axis depicts Pleasant (P) and Unpleasant (UP) 244

valences. The vertical axis depicts High Activation (HA) and Low Activation levels (LA). 245

The diagonal axes define another four valences, in which pleasant valences are on the right – 246

Activated Pleasant (AP), Pleasant (P) and Unactivated Pleasant (UAP), and unpleasant valences are 247

placed on the left – Unactivated Unpleasant (UAUP), Unactivated Pleasant (UP) and Activated 248


8


Unpleasant (AUP). HA and LA levels have no valence; they vary in magnitude only. Note that affect 249

octants that oppose each other (180°) have an antagonist association. This defines the circumplex 250

structure of space. 251

252

253

Figure 2. The circumplex affect space depicting the two-dimensional structure of pleasure-254

displeasure (horizontal axis) and high-low levels of perceived activity (vertical axis). It has eight 255

affective states with two adjectives each. Plus and minus signs indicate positive or negative affect, 256

respectively. Adapted from Knez (2014) 257

258

Two adjectives were chosen to represent each of the eight valences. This is a short version of this 259

instrument (see Knez, 2014). Participants rated their emotional experiences using a 1-5 scale (from 260

‘little to not at all’ to ‘very much’) when answering the question ‘How do you feel right now?’. 261

262

2.3.4 Heart rate variability 263

Electrocardiogram measurements were recorded with ECG-monitor/cardiometer (Gould 6600 264

ECG/Biotech module, Valley View, OH, USA). Ag/AgCl Electrodes were attached in a precordial 5-265

lead arrangement (two electrodes placed under the left and right collarbones and three electrodes 266

placed in line on the left lower rib cage). Recordings were acquired every 0.005s (sample rate: 267

200Hz). HRV variables were extracted from the original ECG data – measured continuously – for 268

analysis. For the PVT task, the initial and last 120s were considered. For the Visual Search, Match-269

to-Sample and Rotation tasks, the last 120 seconds in each task were considered. For the N-back 270

tasks (1, 2, and 3-back), the whole 90s of the duration of each task was considered plus 15s before 271

the first (1-back) task, the overlapping 15s between 1-back and 2-back and between the 2-back and 3-272

back, and the 15s after the 3-back task (totalling 120s for each N-back task). 273


9

The time-domain HRV indicators considered were root mean square of successive differences 274

(RMSSD), percentage of adjacent intervals that differ from each other by more than 50ms (pNN50), 275

and mean Inter-Beat-Intervals (RR intervals). Inter-beat intervals (IBIs) from ECG data were plotted 276

versus time for visual inspection and for editing, by means of linear interpolation, of abnormal and 277

ectopic beats. 278

Repeated measures analysis of variance (ANOVA) was performed on all dependent measures for the 279

for each cognitive test battery and for self-reported measures of fatigue, with test session as the 280

within-subject factor. Heart rate variability data were quantified as RMSSD, pNN50 and mean RR 281

Intervals. Then, HRV data was analysed via repeated measures ANOVA (cognitive task condition x 282

time of performance – Hour 3, 5, ,7, 9 only, when participants were inside the gondola). 283

Post hoc differences were established using Bonferroni multiple-comparison correction. The level of 284

significance was set at p < .05 for all analyses. A Pearson correlation (r) was performed to measure 285

the strength and direction of association between HRV variables and emotions. 286

Results of self-reported measures of fatigue and cognitive performance are similarly reported 287

elsewhere (Rosa et al., 2020). Similarly, results on emotions over time are reported elsewhere 288

(manuscript under review). These three results are summarized in the Results section. 289

290

2.4 Procedure 291

The experiment was performed during daytime for 12 consecutive hours. Of those 12 hours, the 292

experiment part that was conducted inside the DFS lasted 11 hours. 293

Participants arrived at 7:00am at the Flight Physiological Centre. The electrocardiograph electrodes 294

were fitted and electrocardiogram monitor was started. 295

The cognitive task battery and fatigue ratings were administered in six points in time; before entering 296

the DFS (hour 0), while inside the DFS (hour 3, hour 5, hour 7 and hour 9 – centrifuge stationary), 297

and after leaving the DFS (hour 12). Randomization between tasks was not programmed as to obtain 298

the same cognitive load for each task in each point in time and for every participant. 299

Ratings of self-perception of fatigue and emotional ratings were obtained after each cognitive test 300

battery. Ratings were obtained by reading the SPFI scale and the CAS – attached inside the DFS 301

gondola – and verbally stating the chosen index through the inter-communication system. 302

ECG data was recorded throughout the 11 hours while participants were inside the gondola. HRV 303

variables were considered in relation to the beginning and end of cognitive tasks, i.e., obtained HRV 304

results were synced with cognitive tasks’ timings for assessment of cognitive performance over time. 305

The four measurement points were Hours 3, 5, 7 and 9. Averaged HRV data was considered for 306

correlations with emotional ratings. 307

Participants could engage in ‘free flying’ in the simulator between the cognitive test batteries. 308

309

310


10


Table I. The experimental protocol 311 312

Time Procedure Time Procedure

0700 Arrive 1440 Air refueling task

0715 Start/Briefing 1450 Free flying

0730 Undress 1500 Cognitive tests

0735 Height/Weight 1530 Self-reported measures

0740 Urine sampling 1530 Free flying

0750 Blood sampling 1540 Air refueling task

0800 Instrumentation (ECG) 1550 Free flying

0810 Dressing 1700 Cognitive tests

0830 Cognitive tests 1730 Self-reported measures

0845 Self-reported measures 1730 Free flying

0850 Sits in the DFS 1740 Air refueling task

0900 Flight simulation phase starts 1750 Free flying

0915 G-tolerance test without anti-G suit (spinning

centrifuge)

1800 Spatial Disorientation task

(spinning centrifuge)

0945 G-tolerance test with anti-G suit


1900 G-tolerance test without anti-G suit


1025 Spatial Disorientation task


1930 G-tolerance test with anti-G suit


1100 Cognitive tests

(start stationary centrifuge time)

2000 Flight simulation phase ends

1130 Self-reported measures 2000 Moving out from the DFS

1135 Free flying 2005 Undress

1240 Air refueling task 2010 Urine sampling

1250 Free flying 2020 Blood sampling

1300 Cognitive tests 2030 Cognitive tests

1330 Self-reported measures 2055 Self-reported measures/Debrief

1335 Free flying 2100 Weight/Release

313

3 Results 314

3.1 Fatigue 315

Self-reported fatigue increased over time, F(5, 50) = 11.66, p < .001, p2 = .53. Bonferroni post hoc 316

tests revealed a difference between Hour 0 and Hour 7 (p = .04, Mean difference 95% CI [-3.2, -317

.59]), between Hour 0 and Hour 9 (p = .02, Mean difference 95% CI [-3.78, -.21]) and between Hour 318

0 and Hour 12 (p = .002, Mean difference 95% CI [-3.74, -.80]). 319

320

3.2 Cognitive performance 321

Significant results on differences in accuracy or reaction time were not observed for high-order 322

cognitive tasks throughout the 12 hours period. 323


11

PVT task results revealed increase response time over time, F(5, 50) = 7.87, p = .001, p2 = .44. 324

Bonferroni post hoc tests revealed that a difference occurred between Hour 0 and Hour 7 (p = .02, 325

Mean difference 95% CI [-75.6, -5.3]), between Hour 0 and Hour 9 (p = .01, Mean difference 95% 326

CI [-76.9, -7.8]) and between Hour 0 and Hour 12 (p = .01, Mean difference 95% CI [-38.04, -3.4]). 327

3.3 Emotions 328

Regarding positive valences, participants reported being significantly less peppy/enthusiastic and less 329

glad/cheerful over time, (F(5, 50) = 14.1, p < .001, p2 = .58; F(5, 50) = 8.4, p < .001, p

2 = .45, 330

respectively). Bonferroni post hoc tests for peppy/enthusiastic revealed a significant difference 331

between Hour 0 and Hour 7, Hour 0 and Hour 9 and Hour 0 and Hour 12, (p = .009, Mean difference 332

95% CI [0.35, 2.74]; p = .002, Mean difference 95% CI [0.57, 2.70]; p = .00, Mean difference 95% 333

CI [0.56, 2.89], respectively). Bonferroni post hoc tests for glad/cheerful revealed a difference 334

between Hour 0 and Hour 7 and between Hour 0 and Hour 9 (p = .02, Mean difference 95% CI [0.10, 335

2.43]); p = .05, Mean difference 95% CI 0.002, 2.18], respectively). 336

Regarding negative valences, participants reported being significantly more drowsy/bored over time 337

(F(5, 50) = 8.5, p < .001, p2 = .46). Bonferroni post hoc tests revealed a significant difference 338

between Hour 0 and Hour 7, Hour 0 and Hour 9 and Hour 0 and Hour 12, p = .001, Mean difference 339

95% CI [-2.77, -0.68]; p = .006, Mean difference 95% CI [-2.53, -0.37]; p = .003, Mean difference 340

95% CI [-2.31, -0.04], respectively). 341

High activation level results indicated that participants were significantly less active/stimulated over 342

time, F(5, 50) = 11.7, p < .001, p2 = .54. Significant differences disclosed by Bonferroni post hoc 343

tests were observed between Hour 0 and Hour 7, Hour 0 and Hour 9 and Hour 0 and Hour 12, p = 344

.005, Mean difference 95% CI [0.45, 2.82]), p = .001, Mean difference 95% CI [0.24, 2.84]); p = 345

.002, Mean difference 95% CI [0.57, 2.70], respectively). 346

Low activation levels results showed that participants were overall significantly more 347

passive/inactive over time, F(5, 50) = 4.17, p < .05 , p2 = .29. However, Bonferroni post hoc tests 348

did not show significant differences. 349

3.4 Heart rate variability and fatigue 350

A Pearson’s correlation analysis showed no significant associations between RMSSD, pNN50 and 351

mean RR Interval and self-reported fatigue, r’ = 0.34, n = 10, p > .05; r’ = 0.29, n = 10, p > .05; r’ = 352

0.20, n = 10, p > .05, respectively. 353

3.5 Heart rate variability and cognitive performance 354

RMSSD results for all tasks were not significant as a function of time, p > .05. pNN50 results were 355

non-significant for all tasks as a function of time, p > .05, except for the 2-back task, F(3, 27) = 3.25, 356

p = .03, p2 = .26. Mean RR Intervals were not significant for all tasks as a function of time, p > .05. 357

Multiple comparisons with Bonferroni adjustments revealed non-significant differences in RMSSD, 358

pNN50 and mean RR Intervals over time for all tasks (Table 2). 359

360

361


12


362

Table 2. Descriptive values (Mean (SD)) for HRV measures for the cognitive tasks in Hour 3, 5, 7 and 9 while 363 participants were inside the gondola in the DFS. 364

365 DFS phase

Hour 3 Hour 5 Hour 7 Hour 9

Mean (SD) Mean (SD) Mean (SD) Mean (SD) p

PVT (first 120s)

RMSSD 48.5 (25.7) 52.0 (22.0) 50.0 (29.5) 52.2 (30.1) 0.92

pNN50 13.2 (11.2) 11.0 (7.9) 11.8 (10.2) 12.5 (9.5) 0.48

RR Interval 0.84 (.13) .81 (.12) .83 (.17) .84 (.14) 0.64

PVT (last 120s)

RMSSD 46.1 (24.1) 50.5 (23.5) 48.2 (27.5) 47.2 (28.0) 0.65

pNN50 11.3 (9.9) 10.2 (9.0) 12.4 (11.2) 10.2 (9.8) 0.42

RR Interval .85 (.13 .83 (.14) .83 (.17) .84 (.17) 0.66

Visual Search

RMSSD 44.6 (24.4) 48.4 (26.3) 44.5 (25.2) 47.3 29.5) 0.63

pNN50 10.2 (10.5) 10.6 (10.0) 9.9 (9.4) 11.0 (10.4) 0.58

RR Interval .83 (.13) .81 (.13) .82 (.16) .83 (.15) 0.62

Match-to-Sample

RMSSD 44.7 (24.2) 48.8 (27.4) 51.9 (29.9) 46.6 (26.1) 0.32

pNN50 10.1 (9.2) 10.8 (9.7) 11.1 (9.5) 10.7 (9.4) 0.92

RR Interval .82 (.12) .81 (.14) .81 (.16) .81 (.13) 0.96

1-back

RMSSD 54.1 (34.6) 45.7 (21.7) 44.6 (23.7) 45.9 26.6) 0.49

pNN50 9.5 (7.8) 8.6 (7.1) 10.4 (9.1) 10.0 (9.0) 0.43

RR Interval .80 (.10) .79 (.12) .80 (.17) .80 (.15) 0.94

2-back

RMSSD 40.5 (27.1) 46.3 (26.4) 47.3 25.9) 46.2 (28.8) 0.20

pNN50 7.8 (8.4) 10.8 (9.5) 9.5 (8.7) 10.6 (10.2) 0.03*

RR Interval .80 (.13) .80 (.12) .80 (.16) .82 (.15) 0.80

3-back

RMSSD 39.7 (22.2) 41.3 (18.7) 44.3 (25.0) 43.4 (31.1) 0.73

pNN50 8.4 (9.1) 7.8 (6.4) 10.0 (9.2) 9.9 (10.2) 0.34

RR Interval .81 (.14) .79 (.12) .80 (.16) .83 (.18) 0.51

Mental rotation

RMSSD 47 (28.7) 39.7 (19.7) 42 (26.2) 45.5 (31.9) 0.21

pNN50 9.7 (9.9) 9.1 (8.6) 9.0 (8.9) 10.6 (10.6) 0.59

RR Interval .83 (.14) .80 (.15) .81 (.19) .82 (.17) 0.37

Note: ª N=10 due to data loss

* Bonferroni post hoc test revealed NS results

366

3.6 Heart rate variability and emotions 367

Overall results of the Pearson correlation are depicted in Table 3. HRV variables significantly 368

correlated with each other; RMSSD significantly correlated with pNN50, r’ = 0.98, n = 10, p < .01; 369


13

RMSSD significantly correlated with mean RR Interval, r’ = 0.90, n = 10, p < .01; and pNN50 370

significantly correlated with mean RR Interval, r’ = 0.88, n = 10, p < .01). 371

3.6.1 RMSSD 372

Results of the Pearson correlation indicated that there was a significant positive association between 373

the negative valence drowsy/bored, as well as the LA level passive/inactive and RMSSD (r’ = 0.69, n 374

= 10, p < .05; r’ = 0.79, n = 10, p < .01, respectively). There was a significant negative association 375

between the positive valence glad/cheerful as well as the HA level active/stimulated and RMSSD (r’ 376

= -0.64, n = 10, p < .05; r’ = 0.78, n = 10, p < .01, respectively). 377

This indicates that as participants become more drowsy/bored, passive/inactive and less 378

active/stimulated, cheerful, RMSSD increases. 379

3.6.2 pNN50 380

Correspondingly to RMSSD results, there was a significant positive association between the LA level 381

passive/inactive and pNN50 (r’ = 0.76, n = 10, p < .01). The LA level passive/inactive revealed a 382

near significant association with pNN50 (r’ = 0.61, n = 10, p = .059). There was a significant 383

negative association between HA level active/stimulated and pNN50 (r’ = -0.74, n = 10, p < .05). 384

Similarly, we observed a near significant negative association between the positive valence 385

glad/cheerful and pNN50 (r’ = -0.61, n = 10, p = .057). 386

Interestedly, there was a negative association between the negative valence sad/gloomy and pNN50 387

(r’ = -0.65, n = 10, p < .05). 388

These results indicate that as participants became more passive/inactive, less active/stimulated, but 389

also less sad/gloomy, pNN50 increases. 390

3.6.3 RR Interval 391

Results considering the mean RR Interval revealed a significant positive association between the 392

negative valence drowsy/bored (r’ = 0.64, n = 10, p < .05). There was a significant negative 393

association between the positive valence calm/serene, the HA level active/stimulated and mean RR 394

Interval (r’ = -0.66, n = 10, p < .05; r’ = -0.71, n = 10, p < .05, respectively). 395

These results suggest that as participants became more drowsy/bored, less calm/serene and less 396

active/stimulated, mean RR interval increases, indicating increase in HRV. 397

Combined results of HRV measures and positive/negative valences of emotions suggest that there is 398

a correlation of being more drowsy/bored, and increased heart rate variability. Considering high/low 399

activation levels, results exhibited a correlation of being more passive/inactive, as well as less 400

active/stimulated and increased heart rate variability. 401

402

Table 3. Correlations between HRV variables considered (RMSSD, pNN50 and mean RR Interval) and the eight 403 emotional states. 404

405


14


Table 3. Correlations between HRV variables considered (RMSSD, pNN50 and mean RR Interval) and the eight 406 emotional states. 407 408

Measures 1 2 3 4 5 6 7 8 9 10 11

1. RMSSD ⎻

2. pNN50 .987** ⎻

3. RR Interval .909** .889** ⎻

4. Peppy/enthusiastic -0.567 -0.469 -0.534 ⎻

5. Glad/cheerful -.648* -0.618 -0.52 .749* ⎻

6. Calm/serene -0.574 -0.528 -.667* 0.436 0.625 ⎻

7. Nervous/anxious -0.286 -0.393 -0.262 -0.576 -0.201 0 ⎻

8. Sad/gloomy -0.57 -.659* -0.546 0.004 0.289 0.329 0.56 ⎻

9. Drowsy/bored .694* 0.613 .643* -.867** -.696* -0.44 0.367 -0.207 ⎻

10.Active/stimulated -.782** -.745* -.712* .771** .811** 0.549 -0.21 0.329 -.774** ⎻

11. Passive/inactive .799** .765** 0.613 -.742* -.833** -0.412 0.045 -0.463 .832** -.784** ⎻

Note: N = 10 due to HRV data acquisition loss.

* Correlation is significant at the 0.05 level.

** Correlation is significant at the 0.01 level.

409 Table 3. Continued 410 411

Variables 1 2 3 4 5 6 7 8 9 10 11

M 46.461 10.325 0.8205 3.025 3.275 3.575 1.25 1.225 2.275 2.7 2.125

SD 24.30 8.97 0.14 0.58 0.65 0.87 0.37 0.30 0.66 0.62 0.83

412

4 Discussion 413

The purpose of this study was to examine physiological response to cognitive and emotional loads in 414

simulated long-duration flight missions. Hypothesis 1, stating that participants would exhibit increase 415

in variability while trying to sustain task performance as they become fatigued, was not confirmed. 416

Results from participants’ absolute RMSSD, pNN50 and mean RR Intervals referring to increase in 417

parasympathetic indices were shown to be stable while performing both non-executive and executive 418

tasks over time. Hypothesis 2, regarding emotions – stating that higher HRV is negatively correlated 419

with positive emotions, and positively correlated with negative emotions – was partially confirmed. 420

Considering the positive valences, i.e., peppy/enthusiastic, glad/cheerful and calm/serene, 421

glad/cheerful negatively correlated with RMSSD (near significance was observed with pNN50) and 422

calm/serene negatively associated with mean RR Interval. Concerning the negative valences, i.e., 423

nervous/anxiousness, sad/gloomy, drowsy/bored, drowsy/bored positively correlated with RMSSD 424

and mean RR Interval (near significance was observed for pNN50). Interestedly, in an opposing 425

direction, sad/gloomy was negatively associated with pNN50. As for high/low activation levels, i.e., 426

active/stimulated, passive/inactive, for which there was no hypothesis, active/stimulated negatively 427

correlated with RMSSD, pNN50 and mean RR Interval, whereas passive/inactive was positively 428

associated with RMSSD and pNN50. 429


15

Additionally, participants reported significantly increased levels of fatigue after approximately seven 430

hours into the mission. This was shown by results from the self-reported measures of fatigue and the 431

significant performance decrease in sustained attention – as shown by decrease in performance by the 432

non-executive PVT task – around Hour 7. There was no support for the effect of fatigue on executive 433

tasks targeting saliency and contextual cueing, short-term memory, pattern recognition, working 434

memory and analog spatial representations. Participants were significantly less peppy/enthusiastic, 435

less glad/cheerful, less active/stimulated and more drowsy/bored and more passive/inactive also after 436

approximately 7 hours into the mission. 437

Regarding cognitive performance, previous associations between HRV and mental effort generally 438

revealed increased in CV arousal as mental effort increases. Self-regulatory effort may covary with 439

HRV, such that higher HRV may reflect greater self-regulatory effort (Segerstrom & Nes, 2007). In 440

this sense, it is expected higher HRV during high self-regulatory effort compared with low self-441

regulatory effort. Self-regulatory tasks are associated with activation of prefrontal cortex (PFC), and 442

self-regulatory fatigue affects cognitive task performance that are frontal or executive (Schmeichel et 443

al., 2003; Small et al., 2001). The central autonomic network (CAN) is in part responsible for 444

parasympathetic input to the heart (Ahern et al., 2001). Since the CAN and structures that are central 445

for self-regulation colocalize in the brain, cortical activities that follows self-regulation may also play 446

a role in increased vagal input to the heart. 447

Mulder (1986) divided mental effort in ‘task effort’ and ‘state effort’. In the present study, it might be 448

the case that task effort was low. The associated amount of controlled processing for performing the 449

executive tasks in our experiment might not have been demanding; tasks were not particularly hard to 450

perform. It might also be the case that ‘state effort’ was low. Similarly, the amount of effort required 451

to protect task performance from detrimental effects of fatigue or environmental pressures over time 452

also might have been low. This reasoning finds support in the observed non-significant performance 453

differences in accuracy and response times for the executive tasks over time. Also, no penalty was 454

implied if not performing well, which might have been a reason for a possible lower effort in 455

performing the tasks. 456

Considering this, it is possible to argue that individuals believed that success in achieving tasks’ 457

goals was possible, implying that recruitment of self-regulatory strength over time to maintain 458

performance was not required. This follows the arguments proposed by LaGory et al. (2011), in 459

which CV responses depends on the difficulty of the impeding challenge; stable indices of HRV and 460

observed sustained accuracy in performance suggest that mental effort, i.e., ‘task effort’ and ‘state 461

effort’, was low, indicating that participants were not vulnerable to the negative consequences of self-462

regulatory fatigue. As self-regulatory efforts are presumably associated with prefrontal cortex 463

activities, as well as influences of the CAN in controlling vagal inputs to the heart when self-464

regulation is present, as mentioned, unchanging HRV responses corroborates to the assumption that 465

maintaining performance as fatigue developed in these executive tasks did not infer self-regulation. 466

As per the non-executive task, i.e., PVT task, addressing sustained attention, our results are in line 467

with results from Tran et al. (2009). Parasympathetic modulation shown by RMSSD, pNN50 and 468

mean RR Interval indices remained stable from pre- to post task as individuals performed a similar 469

attention task. In the present experimental conditions, this occurred even after degradation of 470

performance in this task after approximately seven hours. The processing of the PVT task – a low-471

order processing task sensible to fatigue effects – is considered to occur in brain regions responsible 472

for attention and motor function, involving the frontoparietal sustained-attention network 473

(Drummond et al., 2005) – therefore different from executive tasks processed in the PFC. The 474


16


recruitment of a different information processing network as compared to executive tasks indicates 475

that performance in the PVT task is not dependent on effort, hence the lack of observation in 476

differences in parasympathetic activity. This offers another description for the stable indices of 477

parasympathetic activity relative to this task. 478

Regarding emotions, nevertheless, these results are in line with previous findings over time in these 479

environmental conditions (manuscript under review), where emotionality ratings addressing 480

glad/cheerful significantly decreased after approximately 7 hours into the mission. Similarly, 481

emotions addressing drowsy/bored significantly increased after 7 hours and, following the same 482

trend, participants reported being less active/stimulated and more passive/inactive, also after around 483

hour 7. Here we find associations concerning these variables and increase in HRV. Increases in HRV 484

usually reflect self-regulatory capacity when individuals are fatigued, implying that a 485

parasympathetically mediated system is associated with self-regulation (Segerstrom & Nes, 2007). In 486

this study, emotional variables associated with increase in HRV were related to increased levels of 487

boredom and passiveness, and decreased levels of cheerfulness and stimulation. 488

A possible interpretation relates to a reduction of available energy resources after approximately 7 489

hours, reflected by a decrease of PVT task performance (decreased arousal). Prior consumption of 490

cognitive resources impaired emotion regulation. Grillon et al. (2015) suggested that depletion of 491

cognitive resources – or the extent to which resources is drained and recover over time – can impair 492

emotion regulation. Still, described associations between emotional ratings and increase in 493

parasympathetic activity when individuals were fatigued may be related to the environmental 494

conditions in the simulated mission, as the situation was safe and stable. This is in line with the 495

established consensus indicating that PNS is dominant when the individual is not subjected to high 496

levels of psychological stress (Appelhans & Luecken, 2006). Conversely, SNS exhibits long-term 497

dominance over the PNS when individuals are vulnerable to high stress levels, having detrimental 498

effects at psychological and physical dimensions (Pinna & Edwards, 2020). 499

In summary, effort-related CV responses were stable irrespective of fatigue development in executive 500

and non-executive tasks in simulated long-duration missions. The lack of significant differences in 501

HRV over time may be due to insufficient task demand to induce performance decrease in executive 502

tasks due to fatigue. This may also have occurred due to the repetitive nature of the battery test, 503

facilitating a practice effect. We find support for these assumptions since performance decrements in 504

these tasks over time were not observed. This is a limitation in this study. 505

These results demand careful interpretation as the three HRV variables significantly correlated with 506

each other. The premise that there might be low statistical power cannot be discarded, as the sample 507

size in this study is small and some results yielded moderate effect sizes concerning correlations of 508

emotional valences. Correlations concerning emotions varying in magnitude, i.e., the high/low 509

activation levels, on the other hand, yielded high correlation coefficients, and significant results were 510

observed between both ratings and all three HRV variables considered, except for passive/inactive 511

and mean RR Interval. 512

Considering this, it is important to notice that the results of executive tasks and emotional ratings 513

should not be taken detached from the present conditions, i.e., the simulated environment. It is 514

problematic to generalize that executive tasks are not affected by fatigue in other circumstances. 515

Rather, differences in performance in these tasks are not easy to detect, and depend on type of tasks, 516

environmental conditions and individual characteristics. Regarding the latter, another limitation in 517

this study relates to the observation of individuals who have a strong need to achieve versus 518


17

individuals who do not. Probably, the first will have high levels of success importance in challenging 519

conditions. As such, they should be more likely to display CV responses differences in response to 520

fatigue compared to individuals who do not have a strong need to achieve. Since our sample 521

comprised of pilots and nonpilots – and pilots supposably would have higher will to perform in these 522

conditions compared to nonpilots – this difference may also have played a role in the non-significant 523

amplification of CV responses (and our sample was too small for a between-group comparisons). 524

Fatigue effects on HRV responses have been broadly different, with some studies reporting positive 525

effects, others reporting negative effects and others reporting null effects, as cited. Associations 526

between HRV and fatigue need careful consideration of type of tasks and environmental conditions. 527

Nevertheless, HR/HRV may be an instrument for assessing pilot’s performance in more realistic and 528

complex flight missions (Mansikka et al., 2016). 529

530

5 Conclusions 531

In this simulated long-duration mission environment, we did not observe differences in HRV in non-532

executive and executive task performance over time. This denotes that a low self-regulatory effort 533

was required for maintaining performance in executive tasks in these environmental conditions. 534

Relatedly, the non-executive task – being not dependent on effort – did not show associations with 535

increased HRV. 536

We observed associations between increased boredom as well as passiveness and decrease in 537

stimulation as well as activeness, and increased HRV. These results indicate a prevalence of 538

parasympathetic activity, reflecting a low degree of physiological arousal. Associations with these 539

emotional variables may relate to the fairly safe and stable conditions in this simulation. 540

Future research in this context might consider the investigation of CV responses during the 541

performance of simulated flight tasks. A collective investigation comprising HRV, cognitive 542

performance and emotions may enhance the understanding of the psychology of pilots in similar 543

conditions. Combined results indicate that pilots may be able to adapt to environmental demands and 544

fatigue in long-duration missions providing that low self-regulatory effort is prevalent. 545

546

6 Funding 547

The study was supported by the Swedish Armed Forces grant no 922:0918. 548

7 Conflicts of interest 549

The authors declare that the research was conducted in the absence of any commercial or financial 550

relationships that could be construed as a potential conflict of interest. 551

8 Ethics Approval 552

The Ethics Application was reviewed by the Medical Research Ethics Review Board – n° 2018/806–553

31, named “Fysiologiska och kognitiva effekter av långvariga flygpass i JAS39Gripen”. 554


18


9 Consent to participate 555

Informed consent was obtained from all individual participants included in the study. 556

10 Acknowledgements 557

We are grateful to the staff and engineers of the Dynamic Flight Simulator at the Flight Physiological 558

Center for their kind assistance in the study. 559

11 References 560

Ahsberg, E., Gamberale, F., & Gustafsson, K. (2000). Perceived fatigue after mental work: An 561

experimental evaluation of a fatigue inventory. Ergonomics, 43, 252–268. 562

Appelhans, B. M., & Luecken, L. J. (2006). Heart rate variability as an index of regulated emotional 563

responding. Review of general psychology, 10(3), 229-240. 564

Beauchaine, T. (2001). Vagal tone, development, and Gray's motivational theory: Toward an integrated 565

model of autonomic nervous system functioning in psychopathology. Development and 566

psychopathology, 13(2), 183-214. 567

Beckers, F., Verheyden, B., & Aubert, A. E. (2006). Aging and nonlinear heart rate control in a healthy 568

population. American Journal of Physiology-Heart and Circulatory Physiology, 290(6), H2560-569

H2570. 570

Berntson, G. G., Norman, G. J., Hawkley, L. C., & Cacioppo, J. T. (2008). Cardiac autonomic balance 571

versus cardiac regulatory capacity. Psychophysiology, 45(4), 643-652. 572

Davidson, R. J., Pizzagalli, D., Nitschke, J. B., & Putnam, K. (2002). Depression: Perspectives from 573

affective neuroscience. Annual Review of Psychology, 53, 545–574. 574

https://doi.org/10.1146/annurev.psych.53.100901.135148 575

Driskell, J. E., and E. Salas. (1991). “Group Decision Making under Stress.” Journal of Applied 576

Psychology, 76 (3): 473. 577

Drummond, S. P., Bischoff-Grethe, A., Dinges, D. F., Ayalon, L., Mednick, S. C., & Meloy, M. J. 578

(2005). The neural basis of the psychomotor vigilance task. Sleep, 28(9), 1059-1068. 579

Endsley, M. R. and Kiris, E. O. 1995, The out-of-the-loop performance problem and level of control 580

in automation, Human Factors, 37, 381–394. 581

Fairclough, S. H., & Houston, K. (2004). A metabolic measure of mental effort. Biological 582

psychology, 66(2), 177-190. 583

Gonzalez, K., Sasangohar, F., Mehta, R. K., Lawley, M., & Erraguntla, M. (2017, September). 584

Measuring fatigue through Heart Rate Variability and activity recognition: A scoping literature 585

review of machine learning techniques. In Proceedings of the Human Factors and Ergonomics 586

Society Annual Meeting (Vol. 61, No. 1, pp. 1748-1752). Sage CA: Los Angeles, CA: SAGE 587

Publications. 588

Grillon, C., Quispe-Escudero, D., Mathur, A., & Ernst, M. (2015). Mental fatigue impairs emotion 589

regulation. Emotion, 15(3), 383. 590

Grol, M., & De Raedt, R. (2020). The link between resting heart rate variability and affective 591

flexibility. Cognitive, Affective, & Behavioral Neuroscience. 592

https://doi.org/10.1146/annurev.psych.53.100901.135148


19

Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of General 593

Psychology, 2(3), 271–299 594

Huang, S., Li, J., Zhang, P., & Zhang, W. (2018). Detection of mental fatigue state with wearable ECG 595

devices. International Journal of Medical Informatics, 119, 39-46. 596

LaGory, J., Dearen, B. B., Tebo, K., & Wright, R. A. (2011). Reported fatigue, difficulty, and 597

cardiovascular response to an auditory mental arithmetic challenge. International Journal of 598

Psychophysiology, 81(2), 91-98. 599

Levenson, R. W. (2003). Blood, sweat, and fears: The autonomic architecture of emotion. Annals of 600

the New York Academy of Sciences, 1000(1), 348–366. 601

Mansikka, H., Simola, P., Virtanen, K., Harris, D., & Oksama, L. (2016). Fighter pilots’ heart rate, 602

heart rate variation and performance during instrument approaches. Ergonomics, 59(10), 1344-603

1352. 604

McCraty, R., Atkinson, M., Tomasino, D., & Bradley, R. T. (2009). The coherent heart heart-brain 605

interactions, psychophysiological coherence, and the emergence of system-wide order. Integral 606

Review: A Transdisciplinary & Transcultural Journal for New Thought, Research, & Praxis, 5(2). 607

Melo, H. M., Nascimento, L. M., & Takase, E. (2017). Mental fatigue and heart rate variability (HRV): 608

The time-on-task effect. Psychology & Neuroscience, 10(4), 428 609

436. https://doi.org/10.1037/pne0000110 610

Mulder, G. (1986). The concept and measurement of mental effort. In Energetics and human 611

information processing (pp. 175-198). Springer, Dordrecht. 612

Neville, K. J., Bisson, R. U., French, J., Boll, P. A., & Storm, W. F. (1994). Subjective fatigue of C-613

141 aircrews during Operation Desert Storm. Human Factors, 36(2), 339-349. 614

https://doi.org/10.1177/001872089403600213. 615

Nolte, R. N., Wright, R. A., Turner, C., & Contrada, R. J. (2008). Reported fatigue, difficulty, and 616

cardiovascular response to a memory challenge. International Journal of Psychophysiology, 69(1), 617

1-8. 618

Parasuraman, R. (1993) Effects of adaptive function allocation on human performance, in D. J. Garland 619

and J. A. Wise (eds), Human factors and advanced aviation technologies (Daytona, Beach, FL: 620

Embry-Riddle Aeronautical University Press), 147–158. Approaches to human-centred automation 621

151 622

Parasuraman, R. and Riley, V. (1997). Humans and automation: use, misuse, disuse and abuse, Human 623

Factors, 39, 230–253. 624

Pinna, T., & Edwards, D. J. (2020). A systematic review of associations between interoception, vagal 625

tone, and emotional regulation: Potential applications for mental health, wellbeing, psychological 626

flexibility, and chronic conditions. Frontiers in Psychology, 11. doi: 10.3389/fpsyg.2020.01792 627

Rosa, E., Eiken, O., Grönkvist, M., Kölegård, R., Dahlström, N., Knez, I., Ljung, R., Jönsson, F.U., 628

Willander, J. (2020). Effects of Fatigue on Cognitive Performance in Long-Duration Simulated 629

Flight Missions. Aviation Psychology and Applied Human Factors, 10(2), 82-93. 630

https://doi.org/10.1027/2192-0923/a000191. 631

Schmeichel, B.J., Vohs, K.D., & Baumeister, R.F. (2003). Intellectual performance and ego depletion: 632

Role of the self in logical reasoning and other information processing. Journal of Personality and 633

Social Psychology, 85, 33–46. 634

https://psycnet.apa.org/doi/10.1037/pne0000110

https://doi.org/10.1177%2F001872089403600213

https://doi.org/10.1027/2192-0923/a000191


20


Small, D.M., Zatorre, R.J., Dagher, A., Evans, A.C., & Jones-Gotman, M. (2001). Changes in brain 635

activity related to eating chocolate: From pleasure to aversion. Brain, 124, 1720–1733. 636

Segerstrom, S. C., & Nes, L. S. (2007). Heart rate variability reflects self-regulatory strength, effort, 637

and fatigue. Psychological science, 18(3), 275-281. 638

Shaffer, F., & Ginsberg, J. P. (2017). An overview of heart rate variability metrics and norms. Frontiers 639

in public health, 5, 258. 640

Thayer, J. F., & Lane, R. D. (2000). A model of neurovisceral integration in emotion regulation and 641

dysregulation. Journal of Affective Disorders, 61(3), 201–216. 642

Thayer, J. F., Hansen, A. L., Saus-Rose, E., & Johnsen, B. H. (2009). Heart rate variability, prefrontal 643

neural function, and cognitive performance: The neurovisceral integration perspective on self-644

regulation, adaptation, and health. Annals of Behavioral Medicine: A Publication of the Society of 645

Behavioral Medicine, 37(2), 141–153. https://doi.org/10.1007/s12160-009-9101-z 646

Tran, Y., Wijesuriya, N., Tarvainen, M., Karjalainen, P., & Craig, A. (2009). The relationship between 647

spectral changes in heart rate variability and fatigue. Journal of Psychophysiology, 23(3), 143-151. 648

Wright, R. A., & Stewart, C. C. (2012). Multifaceted effects of fatigue on effort and associated 649

cardiovascular responses. 650

Zhang, C., & Yu, X. (2010). Estimating mental fatigue based on electroencephalogram and heart rate 651

variability. Polish Journal of Medical Physics and Engineering, 16(2), 67-84. 652

653

https://doi.org/10.1007/s12160-009-9101-z

cognitive performance, fatigue, emotional and

Documents