Studies meeting inclusion criteria are summarized in Table 1. All studies included adults only and sample sizes ranged from 4 to 124 participants. Apart from the studies that evaluated ion effects on patients with some form of depression, six studies [11, 14, 19, 25–27] also evaluated the influence of ions on mood states of persons with varying health conditions. Collectively, the findings from these six studies did not provide contrasting results from those studies that included only healthy subjects. Most studies examined negative air ionization only (n=24); one examined positive air ionization only; and eight studied the effects of both. Blinding of study subjects was not reported in three experiments, nor was it obvious upon review of the study methodology. Among the 30 studies that conducted blind experiments, 18 were double-blind. All but one study  was published in a peer-reviewed journal.
Air ion intensities and duration are summarized in Table 2. Air ion intensities were reported in 29 studies (range: 1000 ions/cm3 (ambient levels) to 27,500,000 ions/cm3). Air ionization duration ranged from 10 minutes at a single time point, to daily treatment periods administered for multiple days, to successive weeks at a time where air ion generators were switched on continuously. Collectively, many studies reported a mood-related response after exposure to ionized air; however, considerable variation by outcome, statistical significance testing, and degree of precision across the reported data was noted.
For reporting purposes, we have organized our review of studies by outcome, ascending year of publication, and the first author’s last name.
Activation, anxiety, and mood outcomes
Four studies examined the effects of negative and positive air ions on activation, anxiety, and mood [15, 16, 25, 27]. Silverman and Kornbleuh  conducted an experiment to examine the effect of negative and positive air ionization on the human electroencephalogram (blinding not reported). Ten healthy adults and two subjects with chronic stationary neurologic conditions participated in the study. Findings indicated a consistent decrease in alpha activity, a non-specific response, ranging from 0.5 to 1.5 cycle decrements during negative or positive air ionization, or both, in 10 subjects (9 healthy; 1 neurologically impaired).
Charry and Hawkinshire  examined the effect of positive air ions on mood in 85 subjects (age range: 18–60; mean age: 30) in contrast to ambient conditions in a single-blind experiment and found significantly greater tension and irritability in subjects’ mood states. In particular, ‘ion-sensitive’ subjects showed that activation decreased and reaction times increased during exposure to positive air ions while non-sensitive subjects showed increased activation and no effects on reaction time.
Dantzler et al.  reported that ratings of mood on three questionnaires by nine subjects with bronchial asthma (age range: 35–64) were unaffected by exposure to negative and positive ions for 6-hour exposure periods in a double-blind crossover study. In contrast, Gianinni et al.  used a double-blind crossover design to evaluate the influence of negative and positive air ions in 14 university-affiliated volunteers and found that positive air ionization significantly increased anxiety, excitement, and suspicion. In contrast, negative air ionization significantly lowered subjects’ extent of suspicion and excitement to those levels attained prior to positive air ion exposure.
Fifteen studies on activation, anxiety, and mood examined the effects of negative air ions only [5, 9, 10, 12, 14, 18, 26, 28–35]. Tom et al.  utilized a double-blind randomized controlled study to determine the impact of negative air ions on mood in 56 adults (age range: 17–61; mean age: 23). No significant differences were observed between experimental and control conditions. On the other hand, Buckalew and Rizzuto  conducted a double-blind randomized controlled trial (RCT) and identified a significant improvement in mood attributed to negative air ionization between experimental (n=12 men) and control (n=12 men) groups (age range: 20–30; mean age: 22.8).
Baron et al.  examined the effect of negative air ionization on mood, memory, and aggression as mediated by personality type among 71 male undergraduate students in a single-blind experiment. The authors found that exposure to moderate/high concentrations of negative air ions significantly heightened aggression among subjects classified as Type A, but not Type B. In addition, the authors reported that negative air ionization produced positive shifts in mood when not provoked by an accomplice, but negative shifts in mood when incited.
Deleanu and Stamatiu  conducted an experiment of 112 patients with mental disorders (blinding not reported). The overall study goal was to mitigate patients’ symptoms by exposing them to negative aeroionotherapy for 10 to 30 days. The findings suggested that in the majority of treated patients, attenuation or the complete disappearance of anxiety and depressive reactions, including insomnia and general disposition, were identified. In contrast, Hedge and Collis  examined the impact of negative air ionization on mood in a double-blind study conducted among 28 healthy women and found no significant benefit of exposure.
Misiaszek et al.  explored the influence of negative air ions on eight manic patients (age range: 22–49) in an experimental pilot study conducted in two phases of four subjects each. The first phase was non-blind and the second was double-blind involving collection of data using anxiety and psychiatric metrics. In phase two, three of the four subjects showed score reductions consistent with clinical improvement; however, inference of these findings was impossible due to the limited number of subjects examined. A more recent single-blind experiment by Reilly and Stevenson  evaluated anxiety levels among eight healthy men (age range: 19–25) who were exposed to negative air ionization. The results showed no significant effect of air ions on state anxiety pre- or post-exercise . In a single-blind study conducted by Watanabe et al. , 13 healthy adults (age range: 21–49; mean age: 26.4) rated their mood after entering a sauna system on two occasions—one with negative air ionization, the other without. The authors observed no significant difference in reported mood states between experimental and control conditions.
Nakane et al.  conducted a crossover study (blinding not reported) among 12 female undergraduates (age range: 18–22) to examine the effect of negative air ionization on anxiety and salivary chromogranin A-like immunoreactivity (CgA-like IR), a protein indicator of sympathetic nerve activity. The findings showed that exposure to negative air ions significantly reduced anxiety compared to the positive control while performing a computer-oriented task, but negative air ionization in the post-task period was associated with a non-significant reduction. Similar results were reported for CgA-like IR.
Goel and Etwaroo  performed a single-blind RCT to determine the immediate effects of bright light, auditory stimulus, and high-density (n=29) and low-density negative air ionization (n=30) on mood and attentiveness in 118 mildly depressed and non-depressed college students (mean age: 19.4). The results showed that exposure to high-density negative air ionization decreased depressive symptoms, total mood disturbance, or anger within 15 to 30 minutes of exposure; however, low-density exposure did not produce significant effects.
A double-blind crossover experiment by Gianinni et al.  exposed 24 manic men (age range: 23–29; mean age: 26.7) to high levels of ambient negative air ions and found a statistically significant reduction in subjects’ manic states. In contrast, Malcolm et al.  conducted a single-blind experiment among 30 healthy subjects (age range: 18–28) randomized to receive either high-density negative air ions or a control condition and found no effect of exposure on anxiety. Of note, the clinic that performed the Malcolm et al.  study subsequently performed a double-blind RCT of adults (21 patients with SAD and 21 controls) exposed to high-density negative air ions and also reported no effect on measures of visual analogue (mood) or State-Trait Anxiety Inventory ratings . When Malik et al.  induced stress in 20 adults (age range: 24–35; mean age: 28.9) in a single-blind study by performing a computer-oriented task, the subjects reported a significant decrease in computer-oriented stress and psychological stress following negative air ionization.
Gianinni et al.  researched the effects of positive air ions only in a double-blind crossover study conducted among 12 adult male volunteers and found that anxiety, excitement, and serum serotonin levels significantly increased when exposed.
Relaxation and sleep
Several studies examined the impact of negative and positive air ionization on relaxation and sleepiness. In the study by Silverman and Kornbleuh , more than half of their 12 subjects reported one or more symptoms of dryness of the mouth/upper respiratory tract, relaxation, or sleepiness when exposed to either negative or positive air ionization; however, these responses were more prevalent during negative air ionization. Yaglou  conducted a single-blind crossover study in 25 healthy adults (age range: 22–51) and a separate study in 6 arthritic patients (age range: 30–62) to examine the effects of negative and positive air ionization on relaxation. In the first study of 25 adults, 5% reported feeling relaxed when exposed to positive air ions; 17% reported feeling relaxed when exposed to negative air ions; and 21% reported feeling relaxed under control conditions . In the second study, a higher frequency of patients reported feeling relaxed or sleepy, or both, when exposed to negative versus positive air ions .
Albrechtsen et al.  conducted two single-blind experiments to evaluate the influence of negative and positive air ionization on subjective feelings among two groups: 6 randomly-selected women (age range: 20–30) and 12 adults (age range: 19–45) who appeared to be most sensitive to ionization. Outcomes included subjective assessments on feelings of self-exertion, stuffiness, the unpleasantness of cognitive tasks performed, and sleepiness. Across both studies, no significant effects were identified. Hawkins  examined the influence of negative and positive air ionization in an office environment on personal ratings of thermal comfort, stuffiness, alertness, and well-being in a double-blind crossover experiment conducted over 12 weeks. Subjects (n=106) were divided into groups based on areas of variable ionization levels. Hawkins observed that negative air ionization was associated with higher subjective ratings of alertness, atmospheric freshness, environmental/personal warmth, and a reduction in the overall complaint rate by 50%. Positive air ion effects were not explicitly discussed.
Twelve studies examined the association of negative air ions only with relaxation and sleepiness [5, 7, 8, 11–14, 22, 29, 32, 34],. Assael et al.  conducted a double-blind crossover study to examine the effects of negative air ions on relaxation and alertness among 10 healthy participants (age range: 20–65) and 10 subjects on tranquilizers. The authors found that all patients reported an initial relaxation followed by alertness when exposed to negative air ions.
Three previously mentioned studies on air ions and mood associations also evaluated ion effects on relaxation or sleepiness, or both [12, 29, 34]. The double-blind experiment conducted by Tom et al.  of 56 adults assessed the impact of negative air ions on relaxation (very tense versus very relaxed). Although reported feelings of relaxation were slightly elevated in the experimental compared to the control group, the findings were statistically non-significant. On the other hand, Buckalew and Rizzuto  identified a significant increase in relaxation attributed to negative air ionization between experimental and control groups in their double-blind study. In the work of Deleanu and Stamatiu , sleep normalization was achieved in 53 of 67 patients with insomnia who were exposed to negative air ions (blinding not reported).
Lips et al.  performed a double-blind crossover trial to examine the effect of negative air ions on alertness in 18 healthy adults. Subjects worked in either room one with windows (natural ventilation) or room two with no windows (mechanically ventilated). Lips et al.  observed that following exposure to enhanced negative air ions, subjects’ feelings of drowsiness were significantly reduced within both rooms. In the pilot study by Misiaszek et al. , all four subjects fell asleep and reported feeling calm following negative air ionization in the first phase of the study (non-blind). In the second phase (double-blind), three of the four subjects fell asleep and one subject appeared less agitated. In both phases, patients’ manic behavior reappeared 5 to 10 minutes post-treatment .
Terman et al.  conducted a double-blind crossover experiment to examine the effects of timed bright light and negative air ionization on sleep timing in 124 subjects (age range: 18–59; mean age: 39.4), with 20 subjects randomized to high-density and 19 subjects randomized to low-density negative air ionization. The findings showed that exposure to high-density versus low-density negative air ionization did not result in statistically significant differences in sleep patterns. On the other hand, Iwama et al.  conducted a double-blind experiment with 44 patients randomized to the control and 51 patients randomized to receive negative air ion treatment (mean age: 40). Five degrees of tension were defined: 1=relaxed; 2=normal tension; 3=mild tension; 4=moderate tension; and 5=severe tension. The authors found that treated patients’ tension reduced significantly and quicker.
Goel et al.  conducted a double-blind RCT to evaluate the efficacy of bright light and high-density negative air ionization for non-seasonal chronic depression and sleep in 32 patients (age range: 22–65; mean age: 43.7). The findings showed no significant change in sleep onset between high-density (n=12) and low-density (n=10) negative air ionization; but a significant alteration in sleep offset was noted among the high-density subjects. Similarly, in a single-blind study of light and air ion treatment for depression, Goel and Etwaroo  found no significant differences in subjects’ feelings of sedation, pleasantness, or intensity. In a double-blind RCT by Terman and Terman , 99 adults with SAD (age range: 19–63; mean age: 40.4) were followed to examine the effects of high- and low-density negative air ionization and light therapy during subjects’ final hours of sleep. Sleep disturbances in 3 of 16 patients in the low-density group were observed, but none in the high-density group.
In a single-blind experiment of 30 healthy subjects (age range: 18–28) randomized either to receive high-density negative air ionization or to a control condition, Malcolm et al.  found no effect of air ionization on subjects’ feelings of alertness or calmness. A subsequent double-blind RCT of SAD patients and controls reported no effect on patient alertness and found that negative air ion treatment increased vigilance to unmasked positive items in the visual dot-probe task regardless of patient group .
Personal comfort ratings
Three studies evaluated the impact of negative and positive air ionization on personal comfort [17, 37, 38]. McGurk  examined the effects of negative and positive air ions on self-reported feelings of comfort, ease of working on cognitive tasks, and reactions to the test room environment in 10 college-aged males undergoing a single-blind experimental assessment. All subjects were informed that on some days the air would be ionized; however, subjects remained uninformed about polarity. The findings showed that negative air ion exposure resulted in a notable increase in the proportion of subjects reporting more pleasant feelings, while positive air ion exposure versus the control condition resulted in a significantly higher reporting of unpleasantness.
Findings in the Albrechtsen et al.  study found no significant relationship between exposure to high concentrations of negative and positive air ions and feelings of self-exertion, stuffiness, or the unpleasantness of cognitive tasks among 25 healthy subjects or 6 arthritic patients. In contrast, Hawkins  observed that negative air ion exposure was associated with higher subjective ratings of alertness, atmospheric freshness, and environmental/personal warmth among office employees working in three different areas of variable air ionization levels (double-blind study).
Several more recent studies [13, 35, 40] examined the influence of exposure to negative air ions only on personal comfort among adults. Finnegan et al.  conducted a single-blind experiment and found no significant effect of negative air ionization on personal comfort among 26 adults working within 5 different rooms of an office building. On the other hand, Lips et al.  examined the effects of negative air ion exposure on personal comfort and well-being in a double-blind study of 18 healthy adults who worked in either a room with windows (normal environment) or one mechanically ventilated (ion-depleted environment). The findings showed that following exposure to enhanced negative air ions, subjects’ assessments of both their own well-being and their environments (room pleasantness and comfort) improved significantly at both sites, but failed to result in a significant difference in personal thermal comfort scores. In addition, subjects in the ion-depleted environment failed to experience an improvement in air freshness during negative air ion exposure. In the single-blind, ion-enhanced sauna study by Watanabe et al. , no significant differences in the reported feelings of pleasantness between exposure settings were observed.
All depression studies evaluated potential alterations only from exposure to negative air ions [4–8, 22, 24, 29, 31]. In the study of 112 psychiatric patients by Deleanu and Stamatiu , the findings showed that in over 50% of 45 treated patients diagnosed with depression, depressive reactions attenuated or completely disappeared with exposure to negative air ions (blinding not reported). Terman and Terman  performed a double-blind RCT among 25 patients (mean age: 38.2) to examine the effects of negative air ions on SAD. Subjects were randomized to low-density (n=13) or high-density (n=12) treatment. The authors found that depression severity decreased (determined using SIGH-SAD) more notably for the high- than the low-density treatment group. Applying a remission criterion of ≥50% reduction in symptom severity, 58% of patients reacted to high-density and 15% reacted to low-density air ion exposure. Terman et al.’s  double-blind study of the effects of timed bright light and negative air ionization on SAD in 124 adults showed that exposure to high-density air ionization provided subjects with clinically significant relief by producing a 50% reduction in depressive symptoms from baseline. In addition, the remission rate associated with high-density negative air ionization rose substantially with an additional 10 to 14 days of treatment after the first period, but low-density exposure showed no significant effect .
In their double-blind study evaluating the efficacy of bright light and high-density negative air ion exposure for non-seasonal chronic depression in 32 adults, Goel et al.  observed a score improvement on the SIGH-SAD of 51% for high-density exposure (remission rate 50%) compared to 17% for low-density exposure (remission rate 0%). Similarly, Goel and Etwaroo’s  single-blind study of the immediate effects of bright light (n=29), auditory stimulus (n=30), high-density (n=29), and low-density negative air ionization (n=30) in mildly depressed and non-depressed adults indicated that exposure to high-density negative air ions decreased depressive symptoms within 15 to 30 minutes; however, low-density exposure did not produce any significant effects.
In a double-blind RCT by Terman and Terman , 99 adults with SAD or bipolar II disorder were followed to examine the effects of high- and low-density negative air ionization and light therapy during the final hours of sleep. Study findings based on SIGH-SAD indicated that exposure to low-density negative air ions resulted in a significantly lower improvement (22.7%) in depression scores compared to improvement with high-density exposure (47.9%). Flory et al.  also investigated the effects of high- and low-density negative air ionization and light therapy on SAD among 73 university-affiliated women (age range: 18–51; mean age: 20.8) in a single-blind RCT and found that subjects in all study groups showed significant score decreases on the SIGH-SAD self-rating scale and the Beck Depression Inventory (BDI) scale. Dauphinais et al.  performed a double-blind RCT of adult patients with bipolar depression to examine the effect of negative air ions. Subjects were randomized to low-density (n=20), high density (n=2), or bright light (n=18) treatment for 8 weeks. Of note, the low-density group was considered the control and too few data were available for the high-density group to allow for a meaningful analysis; therefore, data among the high-density group were not reported. The authors found no significant difference between the depression severity scores (determined using SIGH-SAD) of the light and low-density treatment groups (52% vs. 47% reduction, respectively) or between the proportion of responders and remitters (light group—50% of subjects were either responders or remitters; low density ion group—55.6% of subjects in the low-density treatment group were either responders or remitters).
Harmer et al.  exposed 21 SAD patients and 21 controls in a double-blind RCT to high levels of negative air ions for 1.5 hours. Post-exposure measures of depression, as measured by the BDI scale, were unaffected by treatment. Additionally, SAD patients, but not controls, exhibited an increased recognition memory for positive words. The overlap in the results of this study with those of Malcolm et al. , and parallels between air high-density negative ion treatment and single-dose antidepressant administration on negative affective bias [41, 42], suggest a link between emotional processing of certain stimuli and depressive states.
Meta-analysis of depression studies
The forest plots and overall weighted differences in group means (i.e. pre- minus post-ion exposure mean scores) by ion concentration (high/low) are shown in Figures 1 and 2. Estimates of treatment effects for studies with multiple follow-up times [6–8] were examined by time point also. Utilizing the later post-baseline mean score where applicable, the weighted differences in group means for the Atypical symptom subscale, Hamilton subscale, and composite SIGH-SAD scale were 5.64 (95% CI: 4.44-6.85), 9.23 (95% CI: 8.52-9.94), and 14.28 (95% CI: 12.93-15.62), respectively (p for heterogeneity (SIGH-SAD) = 0.94); thus, the results were indicative of a beneficial effect of high-density negative air ion treatment on SAD and treatment effects were comparable between studies (Figure 1). The weighted differences in group means in the low-density negative air ion analysis for the Atypical symptom subscale, Hamilton subscale, and composite SIGH-SAD scale were 1.98 (95% CI: 0.57-3.40), 4.87 (95% CI: 0.96-8.77), and 7.23 (95% CI: 2.62-11.83), respectively (p for heterogeneity (SIGH-SAD) < 0.0001); thus the results were also statistically significant, but smaller in magnitude and were significantly different between studies (Figure 2).
The findings were similar when utilizing the earlier post-baseline mean score reported by Terman and Terman [6, 7] and Terman et al.  (results not shown); however, the magnitude of effect by subscale and overall was consistently smaller than those shown in Figures 1 and 2. Furthermore, the weighted group mean difference for the Atypical symptom subscale was statistically non-significant in the low-density negative ionization analysis (mean=1.54 (95% CI: -0.31-3.39)).
Sensitivity analyses were performed by removing the Terman and Terman  study since the data were presented in a figure and not explicitly reported. These analyses showed no alteration in the findings. An additional assessment of exposure duration (hours), within high- and low-density air ion levels, and each study’s score mean difference indicated no evidence of a dose–response relationship (Figure 3).
Publication bias was examined visually with funnel plots, which allow for a visual assessment of the estimated intervention effects from the individual studies plotted against a measure of treatment effect size. Separate plots were done for SIGH-SAD composite scores and SIGH-SAD subscales combined since Terman and Terman  reported estimates by subscale only and Terman et al.  reported estimates for the composite scale only. A clustering indicative of publication bias was not observed (Figure 4) (i.e., no marked asymmetry was evident). Statistical evidence of publication bias was not found (Begg rank correlation p=0.71; Egger regression p=0.37). These findings were supported by those observed when combining the Atypical and Hamilton subscales.