Hyperthermia impairs the executive function using the Attention Network Test

Abstract

Purpose: The primary aim of this study was to examine the effect of passive hyperthermia on the human attention system using the Attention Network Test (ANT), which has been used in both healthy controls and patients. Using target contrasts between conditions within a Flanker paradigm, the ANT can isolate three essential networks in the attention system: maintaining an alert state that is receptive to stimulus input and ready for responding; orienting, which involves selection of sensory input; and executive control, which monitors for and resolves conflict in responding or other aspects of cognitive processing.

Materials and methods: The ANT was administered to 16 right-handed subjects in a heat stress condition (50°C, 40% relative humidity) and a control condition (20°C, 40% relative humidity), for 1 hour each. Reaction time (RT) and accuracy rate as well as mean body core temperature (T-core) and body weight loss were recorded.

Results: Compared to the control condition, T-cores significantly increased and body weight was not significantly reduced in the heat stress condition. Overall, there were non-significant group differences for RT and accuracy rate. Although significant changes in neither alerting nor orienting effects were modulated by the simulated hyperthermia, the executive control effect on RTs was significantly larger in the heat compared to the control condition.

Conclusions: Passive hyperthermia impaired executive control function, whereas alerting and orienting effects were unaffected.

• Attention network
• executive control
• passive hyperthermia

Introduction

Effects of the thermal environment on cognitive performance have been extensively investigated based on the type of cognitive task [1], duration of exposure [2], psychological adaptability of the individual [3], and use of high incentives for testing [4]. Passive heat exposure for long periods can cause cognitive performance decrement by disturbing the central nervous system [5–7]. However, for short periods, effects of heat stress can be overcome in highly motivated and experienced subjects [8]. Hancock proposed that subjects with high skill level on a task are better able to withstand the subsequent effects of heat stress [3]. Indeed, participants assigned to complex tasks requiring a high state of vigilance, cooperation and coordination need to overcome possible negative effects of thermal stress on cognitive function that can result in a significant impairment of specific types of cognitive performance. Indeed, several studies have shown a significant impairment from heat stress in short-term memory, working memory, vigilance performance, and visual recognition [9–10]. To date, however, no studies have directly investigated the effect of heat exposure on attention function.

It has been well demonstrated that the attention system formed by specific anatomical areas can be further broken down into three networks: (1) alerting, defined as achieving and maintaining an alert state to incoming stimuli and being ready for responding, (2) orienting, which is the selection of information from sensory input that can be voluntary or involuntary, overt or covert, location-based or object-based, and (3) executive control, defined as resolving conflict among responses or other aspects of cognitive processing including planning or decision-making, error detection, novel or badly learned responses, and overcoming habitual actions [11–15]. The Attention Network Test (ANT) can measure the efficiency of these three networks by averaging reaction time and accuracy scores across several different cued reaction times in flanker tasks developed by Fan and Eriksen [12], [16]. To date, the ANT has been widely used to measure attention network function in many research areas, for example, in the assay performance of children [17–18] and in clinical patients with attention disorders (e.g. borderline personality disorder [11], [19], schizophrenia [20], and attention-deficit/hyperactivity disorder [21]). Functional magnetic resonance imaging (fMRI) results suggest that the ANT does measure the three mostly separate networks related to components of attention, and these networks are associated with separate anatomical brain areas [22], neurotransmitter systems [23], and genetic markers [24]. To this end, in the present study we used an ANT task to examine the effects of heat stress on the attention networks.

Materials and methods

Participants

The participants (N = 16; 5 women; mean age 24.2 ± 3.6 years; height 173 ± 5 cm and weight 70 ± 13 kg) had not previously taken part in heat stress experiments. All the participants reported normal or corrected-to-normal vision, had no history of current or past neurological or psychiatric illness, and took no medications known to affect the central nervous system. They signed an informed consent form approved by the Jinan Ethical Committee prior to the experiment and were paid for their participation. The investigation complied with the laws and regulations in China.

Experimental procedures

All participants were tested in two conditions: a control condition and a heat stress condition. The order of testing was counterbalanced. To avoid weight loss, each participant was given 100 mL of water before the start of the experiment. Body weight was measured after participants emptied their bladder. During the experiment, participants were asked not to drink or urinate. After 10 min of rest, body core temperature (T-core) was measured by a thermocouple (Honour, ZWJ-2, Shanghai, China) inserted beyond the anal sphincter. In the heat stress condition, participants entered the experimental chamber (50° ± 1°C, 40 ± 1% relative humidity) and remained in a sitting posture for 60 min. Subsequently, the ANT was carried out. T-core and body weight were measured again after the ANT. The procedure in the control condition was the same as in the heat condition, except for the conditions in the chamber (20 ± 0°C, 40 ± 1% relative humidity).

Attention network test

A modified version of the Attention Network Test (ANT) proposed by Fan et al. [22] was used in the experiment. As shown in Figure 1, there were three cue conditions (no cue, centre cue, and spatial cue) and two target types (congruent and incongruent). During the entire trial, a fixation cross appeared at the centre of the screen. In each ANT trial, the stimulus consisted of a central arrow (the target) with a leftward or rightward pointing arrowhead at the centre and four arrows surrounding the target. Two arrows were on each side, termed flanking stimuli, and were either pointed in the same (congruent) or the opposite (incongruent) direction. Each trial began with the presentation of a fixation cross followed by one of the three cue conditions with a presentation duration of 200 ms. After a variable duration of 300–11 800 ms, the target was presented 1.06° above or below the fixation point. The task was to identify the direction (left or right) that the target pointed to by pressing a button as quickly as possible. The target display disappeared until a response was made or after 2000 ms with no response. The fixation point remained on the screen for a variable duration of 3000–15 000 ms. One run contained 36 trials, and each participant was tested for 6 runs. This took about 35 min. To measure the efficiency of the three attention networks, behavioural analysis of RTs was conducted for different cue and target conditions. The efficiency of these three networks was measured by mean reaction time (RT):

Figure 1. Schematic of attention network test in the present study. In each trial a fixation cross appears in the centre of the screen all the time. Depending on the cue condition, a cue condition (no cue, centre, or spatial cue) appears for 200 ms. After a variable duration (300–11 800 ms), the target (congruent or incongruent flankers) are presented until the participant makes a response with a button press, but for no longer than 2000 ms. Once the subjects makes a response, the target disappears immediately and after a variable duration (3000–15 000 ms) the next trial start.

Data collection and statistical analysis

Statistical analyses were performed with SPSS version 17.0 (Chicago, IL, USA). Weight loss and T-core changes were analysed between control and heat conditions. For each subject, completion rate <90%, accuracy rate <90%, and the RT data after error trials and outliers were removed for each condition (three cue conditions: no cue, centre cue, and spatial cue; two target conditions: congruent and incongruent). ANT data, including the reaction time and accuracy rates for each condition, were analysed using a three-way ANOVA, with heat condition (control, heat), cue condition (no cue, centre cue, spatial cue) and target condition (congruent, incongruent) as within-subjects factors. In addition, for the analysis of the efficiency of the three attention networks, the difference between control and heat conditions was examined for alerting effect, orienting effect, and executive control.

Results

Body weight and temperature responses

Compared to control condition, T-core was significantly increased (37.3 ± 0.3°C and 38.4 ± 0.4°C for control and heat conditions, respectively; p < 0.001) and the body weight loss was not significantly reduced, i.e. less than 0.1% (69.6 ± 13.2 kg and 68.9 ± 12.8 kg for control and heat conditions, respectively; p = 0.89).

Reaction time and accuracy rates measures

Table 1 and Figure 2 show the mean RTs and accuracy rates for each condition of the ANT, respectively.

Figure 2. Three ANT effects of RTs, altering, orienting and executive control, in each condition. Heat stress resulted in larger executive control effect, while both the alerting and orienting effects did not show significant changes.

Table 1.  RT (ms) and accuracy (%) for each condition (Mean and SD).

		No cue		Centre cue		Spatial cue
		Mean	SD	Mean	SD	Mean	SD
Control condition
Congruent	RT	644	89	593	76	554	79
	Accuracy	99.7	0.9	99.7	1.1	99.9	0.6
Incongruent	RT	684	82	666	81	611	90
	Accuracy	98.8	2.1	97.0	2.8	99.3	1.3
Heat condition
Congruent	RT	624	75	591	65	543	69
	Accuracy	99.3	1.2	99.6	1.0	99.8	0.7
Incongruent	RT	697	89	676	80	611	92
	Accuracy	99.2	1.0	99.0	1.1	99.5	1.0

For RT, the ANOVA showed a significant main effect of cue condition, F(2, 30) = 234.5, p < 0.0001, revealing that RTs were faster for the spatial cue condition (581 ms) than for the centre cue (632 ms; p < 0.0001) and no cue (663 ms; p < 0.0001) conditions, and for the centre cue compared to the no cue condition (p < 0.0001). The main effect of target condition was also significant, F(1, 15) = 161.8, p < 0.0001, showing that response speed was slower for the incongruent (658 ms) than for the congruent (592 ms) condition. Although there was no significant main effect of heat condition (627 ms and 624 ms for control and heat conditions, respectively; F < 1), importantly, the interaction of heat condition × target condition was significant, F(1, 15) = 4.9, p < 0.045. Further analysis revealed that the target condition effect was more pronounced in the heat (RT incongruent − RT congruent = 75 ms) than in the control condition (57 ms; p < 0.045). No other effects were significant (p > 0.05).

Similar to analysis of RTs, the ANOVA analysis of accuracy rates did not show a main effect of heat condition (99.2% and 99.2% for control and heat conditions, respectively; F < 1). A significant main effect of cue condition was found, F(2, 30) = 3.9, p < 0.035, revealing that accuracy rates were lower for the centre cue condition (98.8%) than for spatial cue (99.6%; p < 0.025) and no cue (99.3%; p < 0.05) conditions. There was no significant difference between the latter two conditions (p = 0.742). The main effect of target condition was also significant, F(1, 15) = 9.8, p < 0.01, showing that the accuracy rates were lower for the incongruent (98.6%) compared to the congruent (99.8%) condition. No other effects were significant (p > 0.05).

Attention networks

To further clarify the above effects, we conducted paired sample t-tests for the three ANT RT effects: alerting, orienting, and executive control. We did not find significant changes for either the alerting (37 ms and 29 ms for control and heat conditions, respectively; t(15) = 1.11, p = 0.28) or the orienting (45 ms and 55 ms for control and heat conditions, respectively; t(15) = 1.36, p = 0.20) effects. However, the executive control effect was significantly distinctive between control and heat conditions, t(15) = 2.42, p < 0.03, indicating that heat stress resulted in a larger executive control effect (54 ms and 73 ms for control and heat conditions, respectively).

Discussion

The thermal environment can produce effects such as dehydration and physical fatigue [25]. There is evidence that a 2% or more decrease in body weight can result in dehydration, leading to deterioration in cognitive abilities [26]. However, possibly due to adequate fluid replacement, cognitive performance is unaffected when body weight changes do not exceed 0.5% [27–28]. In the present study all participants ingested water before the experiment to avoid dehydration effects, and body weight loss was controlled at less than 0.1%. Hence, the observed ANT changes were not caused by weight changes per se. Compared to the control condition, in the heat condition the increment in body core temperature was significant. Mean T-core temperature increased 1.1°C due to thermal stress in our experiment. In line with the current findings, previous research reported that when deep body temperature is perturbed by thermal stress exposure away from both normal and steady-state conditions, vigilance can be impaired [3].

This study aimed at examining the effects of passive hyperthermia on the attention system using the attention network test (ANT). We did not find significant heat stress effects on either the alerting or orienting effect. In contrast, a larger executive control effect of RTs was found for the heat compared to the control condition. It has been well documented that alerting is the ability to prepare for a sensory signal, allowing an individual to achieve and maintain a high sensitivity state of alertness to become ready for any incoming stimuli. Orienting is the ability to select information from many candidate inputs, turning attention toward a sensory signal and selecting one region of space by directing attention to cued areas. Neuroimaging evidence has revealed a functional dissociation between the alerting and orienting networks, with the former involving the thalamus, frontal, and parietal brain regions [13], [14], [22] and the latter involving parts of the superior and inferior parietal lobe, temporal parietal junction, frontal eye fields, and thalamus [15], [22]. The present data suggest that the aforementioned brain areas relevant to the alerting and orienting networks were not modulated by heat stress. However, previous studies have shown that heat stress exposure can impair vigilance, cause stress, and drain attention resources, leaving fewer of those resources to perform a task [29]. Therefore, the effects of heat stress on attention function of human beings should be investigated further.

Different from alerting and orienting effects of attention networks, executive control is the specific ability to resolve conflict among potential responses to a presented stimulus. This function is needed when an individual is requested to make a decision, for conflict resolution, error monitoring, planning an action, or inhibiting a response, associated with the activation of the anterior cingulate and lateral prefrontal cortices [16], [22], [30]. Using the ANT, previous studies have reported impaired executive function in pure developmental dyscalculia [31] as well as in childhood [18], with higher conflict scores compared to normal control participants. The present study revealed a deficit in the ability to inhibit responding and/or a deficit in conflict resolution after long-term heat stress. Indeed, our previous study indicated that 1 h of exposure at 50°C impaired task-irrelevant change detection processing under the non-attentional condition [32]. The present study expanded on previous research showing that mismatch or conflict processing was also impaired under the attentional condition.

Conclusion

Using the ANT, in conclusion, after long-term passive hyperthermia, the present study demonstrated deficits in executive control network of attention, while altering effect and orienting effect were uninfluenced.

Declaration of Interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.