Odors have been shown to impact human performance across a range of contexts. Indeed, the focus of the current special issue reflects incidental industrial/environmental odor exposure, e.g., [1,2]. This issue has, for example, shown that the inter-relation between malodors is not a simple additive/synergistic process [1,2]. These findings can, therefore, guide how malodors can be masked within urban areas. Since (indirect) instrumental measures of odor detection have been shown to correlate with direct (olfactometry) measures of odor threshold [3], it is important to consider how the instrumental measures of odor presence within populated areas may be affecting human cognitive performance. For example, Rotton [4] found that exposure to a malodor (ethanethiol) negatively affected performance on a complex task (proofreading) but not a simple mental arithmetic task. Rotton [4] argued that malodors exert analogous detrimental effects on attention to that of other sensory pollutants such as noise. The malodor used by Rotton [4] was shown to have negative effects on well-being, further illustrating the noxious characteristics of the stimuli. However, a number of other studies have shown that positively rated odors can exert negative (as well as positive) effects on cognition, e.g., [5–9]. These findings illustrate that it is not a simple relationship between odor valence and cognitive performance, but that positively rated odors can have both positive and negative effects on cognition. Such observations are of importance with the increased use of commercial odorants in both personal and professional environments. Indeed, one must not only consider odors epiphenomenal to industrial operations, but also the intentional utilization of ‘pleasant’ odors in commercial settings. In order to maximize the effectiveness of these odors one must isolate the effects of specific odors and understand the mechanism(s) underpinning such effects.

Essential oils and other commercially available scents have, for example, been shown to positively affect memory, e.g., [5,6], vigilance, e.g., [7,10], pain perception, e.g., [11,12], self-perception/confidence [13], consumer decision making, e.g., [14], and alertness, e.g., [8]. Jellinek [15] identified four potential mechanisms that might explain the odor-induced cognitive facilitation. The first concerns an odor-specific pharmacological mechanism, wherein volatile compounds enter the bloodstream following inhalation and impact neural activity. Such an account predicts odor-specific effects (potentially independent from odor valence). The second explanation is an epiphenomenal hedonically-driven mechanism, wherein effects on cognition are secondary to the increase in mood following odor exposure. Such an account predicts that similar valence odors would produce analogous effects on cognition. The third explanation is that odor effects are purely psychological, in that a prior belief/expectancy pertaining to the qualities of the odor underpins any benefits. Under such an account one might predict that an odor should only impact cognition when: (1) the participant is aware of the odor’s presence, and (2) the participant possesses a belief that the odor should induce specific effects. The final explanation is a contextual/associative account, such that odors have specific effects because their presence has been associated with a particular stimulus/mood/behavior. From a memory perspective, this account would predict that if a participant learnt material in the presence of a specific odor, re-presentation of that odor would operate as a subsequent cue to recall for the learnt material (see [16] for a review of context-dependent memory effects). The current paper will review the evidence that odors presented intentionally (rather than arising following industrial/urban activity) can facilitate/decrement cognitive performance. This concerns odors that are used deliberately because it is thought that particular odorants may benefit mood/cognition (i.e., aromatherapy) or, more generally, simply because it is thought that such odors might improve the ambience of an environment. This review will be structured in terms of Jellinek’s [15] four potential mechanisms.

Pharmacological properties of the odor are one mechanism by which odor exposure may facilitate cognition [15]. Odor-specific compounds may stimulate the olfactory nerve which has close association with the limbic system (an area of the brain associated with cognitive functioning). Furthermore, volatile compounds of the inhaled odor may enter the bloodstream via the nasal mucosa (e.g., animal studies have reported traces of the compound in the blood stream after odor exposure, [17]) and thus impact neural activity following neurological delivery. However, there are two clear limitations to this proposition: first, as yet, human studies have not reported traces in the bloodstream following odor inhalation [18]. Second, the amount of active compounds internalized via inhalation is greatly reduced compared to more direct methods of consumption such as ingestion [18]; as a consequence, any pharmacological effects should be greatly attenuated.

An alternative account to the pharmacological hypothesis is that improvements/decrements to cognition may be secondary to changes in mood [15]. Elevation in positive mood is linked with improvements in cognition (e.g., [27]); therefore, mood (induced via odor pleasantness) may underpin cognitive facilitation, rather than pharmacological properties of the odor per se. Under such an account one might predict that odors with similar hedonic properties would produce similar effects on cognition.

The link between odor presentation and emotion is well established. For example, neurologically, dysfunction in areas of the brain associated with emotion (i.e., the limbic system) is related to odor sensitivity [28]. Furthermore, odor-evoked memories have been shown to activate the amygdala, indicating an association with the emotional component of the memory [29]. Behaviorally, positively/negatively valence odors induce emotionally congruent facial expressions in children [30,31], whilst positive odors have been shown to elevate the pitch of spoken language [32] (wherein spoken pitch is linked to emotional state, [33]) and increase the emotional content of memory reports [34]. In addition, the startle response, an emotionally-driven alarm reflex, is influenced by odor exposure [35], further indicating inter-relation.

More explicitly, as aforementioned, a number of studies have shown that odor exposure can impact self-rating mood measures, e.g., [4–6,11,20,36–39]. Physiological responses have provided some support for this shift in mood. Brauchli et al. [39] reported that presentation of a negative odor (valeric acid) increased both heart rate and skin concordance; whilst presentation of a positive odor (phenyl ethyl alcohol) decreased both measures (effects with EEG were less clear). Nevertheless, evidence that odor exposure can influence mood does not, per se, show that the elevation in mood is driving cognitive improvement. Indeed, it is arguably very difficult to empirically test whether the shift in mood is the causal factor in cognitive facilitation. Knasko [36] has proposed that personality factors (specifically locus of control) may be fundamental in determining whether odors affect the mood of participants. Moreover, past experience of odors can change the neural networks within the olfactory bulb; this can produce profound differences in both the experience and perception of odors (see [41] for review). Consequently, a pilot study identifying positive and negative valence odors (e.g., [10]) may be insufficient in guaranteeing the desired affective responses. In order to fully examine the mood elevation hypothesis it is therefore necessary to obtain mood ratings by the participants to establish (1) if the odors have produced the desired affective response and (2) whether there is a mismatch between cognitive facilitation and mood elevation. Such an investigation was conducted by Heuberger et al. [42] and revealed that both self-rated and physiological effects were related to the subjective hedonic evaluations of the odors (see also [12,43]).

In contrast to pharmacological and mood explanations, one might argue that the effects of odors on cognition are uniquely psychological in origin; that is, the expectancy of specific effects following exposure to certain odors underpins any effect rather than the chemical properties of the odor per se [15]. The olfactory perceptual pathway is richly interconnected with other cortical areas (including the hippocampus and amygdala) (for a review see [71]). Consequently, this interconnectivity increases the propensity for contextual and expectancy effects in odor perception. However, the expectancy hypothesis may be difficult to empirically examine since the ability to manipulate expectancy is reliant upon the participant having no prior knowledge as to the potential effects of the odor. Indeed, Howard and Hughes [72] note that a number of odor studies utilize participants with positive attitudes towards aromatherapy (although it should be noted that expectancy effects in individuals may be attenuated by their general limited ability at identifying odors, e.g., [73]). However, by using odors that are less well-known, manipulation of expectancy can enable the examination of three important ideas: (1) does presentation of the odor without expectancy provide any benefit to performance, i.e., are there effects due to the properties of the odor? (2) Are initial effects magnified by expectancy of those benefits? (3) Can effects be reversed through inducing expectancy that the odor will produce the opposite effects, i.e., can properties of the odor be overridden by expectancy?

The role of expectancy in odor-induced cognition facilitation was examined by Moss et al. [9], who manipulated expectations of mood changes following exposure to Roman chamomile. In a between-participants design, participants were assigned to one of the following groups: (1) receive Roman chamomile and expect arousing effects, (2) receive Roman chamomile and expect sedating effects, (3) receive Roman chamomile and given no expectancies (i.e., participants were deceived about the true nature of the study), and (4) no odor control. As in previous work by Moss and colleagues [5,6] participants completed the CDR battery. Moss et al. [9] reported that Roman chamomile does exert some effect upon cognition but that these effects are accentuated by expectancy. Specifically accuracy of attention and self-rated alertness were significantly lower and calmness was significantly higher in the no expectancy condition compared to the non-odor (control) condition. These effects suggest that Roman chamomile has sedative/relaxing properties. However, both quality of memory and secondary memory were significantly impaired in the expect sedation condition relative to the non-odor condition, illustrating that sedating effects of the odor are magnified when these effects are expected. Furthermore, these effects are reversed if participants expect the odor to have arousing properties, i.e., quality of memory, secondary memory, and accuracy of attention were significantly greater in the expect arousal condition compared to the expect sedation condition. These findings illustrate that, in respect to Roman chamomile, the odor does posses properties that can influence cognition. However, expectancy does influence these effects; accentuating or reversing effects depending upon prior belief (for commentary regarding the additive effects of placebo, see [74]). These findings are mirrored by Knasko et al. [75], who observed that individuals who (erroneously) believed that they were being presented a pleasant odor reported more pleasant mood that those who believed that they were being presented an unpleasant odor (or no odor at all); this was found despite the absence of any odor.

The above expectancy effects may be compounded in the literature by the common failure to employ an adequate placebo [72]. Specifically, comparison of a specific odor to a no-odor control (1) prevents the researcher differentiating between odor-specific effects and more general effects of odor presentation, (2) encourages demand characteristics in participants, and (3) allows experimenter expectancy effects [72]. These issues were addressed by Howard and Hughes [72] who presented participants with lavender or a non-sedative placebo odor employing an experimenter who was unaware of which odor was presented to the groups. Participants were then told that the odor would assist relaxation, inhibit relaxation, or were given no information about the odor. The authors reported no effect of odor type on physiological measures of relaxation (i.e., galvanic skin concordance); however, this measure was mediated by the expectancy prime (see also [76]). That is, participants demonstrated higher relaxation states if told that the odor would inhibit relaxation, purportedly because individuals tried harder to relax following the prime. The authors therefore conclude that an evaluation of participant’s knowledge and expectancies following odor exposure are an important consideration in the domain.

The above studies indicate that expectancy may provide, at the very least, an addictive influence on mood and cognition. However, the full influence of expectancy on odor-effects is unknown due to the lack of appropriate placebos and control of experimenter expectancy [72].

The fourth proposed mechanism for odor-induced cognitive facilitation relates to odor context priming mood/behavior [15]. Rather than pharmacological properties of the odor affecting state, individuals associate a specific stimulus/behavior/mood with an odor. The close association between odors and memories has been shown both cognitively and neurologically. Engen [77] proposes that odors become intrinsically entwined with the memory, forming a holistic representation of the event. Indeed, it has been shown that odor-cued autobiographical memories contain greater detail and more emotional content than those cued by a verbal label [78], with such odor-evoked memories shown to activate the amygdala [29]. Such interconnectivity between odors, emotion, and memory reflects the close link with the limbic system during olfactory processing (for review see [71]). Furthermore, odor-specific neurons have been shown to also respond to the contextual cues associated with that odor [79–81]. Due to this interconnectivity, it is possible that representation of an odor may cue the memory, mood, or behavior present during initial odor exposure.

The pharmacological and mood accounts above are both premised on the notion that the chemical compounds of odors affect cognition either directly or indirectly following mood elevation. However, a number of studies have shown that the memorial associations made with odors are influenced by semantic knowledge rather than the pharmacological properties of the odor. For example, neural activation following odor exposure can reflect what an individual believes they are smelling rather than the actual chemical stimulus [82]. In this study participants were presented words (cheddar cheese or body odor) contemporaneously with an odor (isovaleric acid mixed with cheddar cheese flavor). When the cheese label was provided, the odor was rated as more pleasant. Furthermore, both isovaleric acid and clean air (control) showed significantly more activation in the rostral anterior cingulated cortex/medial orbitofrontal cortex when labeled as cheddar cheese. This shows that odor experience (both cognitively and neurologically) can be influenced by the top-down processing of semantic association. Consistent with these findings [82], hedonic responses to both culturally important [83–85] and novel odors have been shown to be culturally-specific [86]. Indeed, Cain [87] argued that odor discrimination is poor and that top-down processing is an essential component of the olfactory process. This further illustrates the role of experiential learning in odor perception and thus highlights how learnt associations may influence variability in odor-induced cognitive facilitation.

This review has shown that odors arising from urban pollution are not the only circumstances in which odors may impact cognition. The intentional utilization of odors in both commercial/industrial and private settings can have a direct impact upon cognition which is not exclusively beneficial. As a consequence, decisions to employ such odors needs to be empirically-driven or such interventions may have a detrimental effect on performance. Commercially available odors have been shown to facilitate and decrement both measures of memory and alertness. Effects are, therefore, not simply related to the hedonic quality of the odor.

Evidence for the four potential explanations of odor facilitation [15] has been reviewed. Essential oil compounds internalized without inhalation can have a direct influence on cognition; however, this does not necessarily show that the pharmacological properties of the odor induce post-inhalation effects. Indeed, more work is required to find evidence of olfactory compounds entering the human system following inhalation. It is though apparent that odor effects can be accentuated/attenuated via improvements to mood, expectancy of certain effects, and contextual congruity. More research is required in respect to each of these accounts. It needs to be established whether the chemical compounds within odors can facilitate/decrement specific cognitive operations; if this is shown, those compounds require isolation. Indeed, a more refined summary of the specific effects of a range of odors would enable the appropriate employment of certain odors to certain circumstances (i.e., matching odors to specific tasks/requirements). Furthermore, it is unclear how quickly individuals habituate to such benefits; data is required on the longevity of such effects in order to rationalize the financial outlay. In respect to the mood elevation hypothesis, it is important that mood measures are taken in conjunction with cognitive measures. Such an assessment would enable evaluation of the situations in which cognition improves independently of mood elevation. Limited research has been conducted on the role of expectancy, however initial data indicate that expectancy can accentuate/attenuate the effects of an odor [9,71]. Further research is required to ascertain the tasks that can be influenced via expectancy of odor-related effects and whether such expectancy effects are robust over prolonged odor exposure. Finally, odors can be used as contextual cues to cognitions, behavior, and mood. It is, however, unclear the extent to which odors can be used to cue behaviors/moods across a range of associations and whether the formation of such associations are odor-specific.