Clinical neuroscience 1081
Women are more strongly affected by dizziness in static magnetic fields of magnetic resonance imaging scanners Angela Heinricha, Anne Szosteka, Patric Meyera, Iris Reinhardb, Maria Gillesc, Georgios Paslakisc, Jaane Rauschenbergd, Jens Gröbnerd, Wolfhard Semmlerd, Michael Deuschlec, Andreas Meyer-Lindenbergc, Herta Flora and Frauke Neesa Increasing field strengths in MRI necessitate the examination of potential side effects. Previously reported results have been contradictory, possibly caused by imbalanced samples. We aimed to examine whether special groups of people are more prone to develop side effects that might have led to contradictory results in previous studies. We examined the occurrence of sensory side effects in static magnetic fields of MRI scanners of 1.5, 3, and 7 T and a mock scanner in 41 healthy participants. The contribution of field strength, sex, age, and attention to bodily processes, and stress hormone levels to the sensation of dizziness was examined in separate univariate analyses and in a joint analysis that included all variables. Field strength and sex were significant factors in the joint analysis (P = 0.001), with women being more strongly affected than men by dizziness in higher static magnetic fields. This effect was not mediated by the other variables such as attention to bodily symptoms or stress hormones. Further research needs to elucidate the underlying factors
of increased dizziness in women in static magnetic fields in MRI. We hypothesize that imbalanced samples of earlier studies might be one reason for previous contradictory results on the side effects of static magnetic fields. NeuroReport 25:1081–1084 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins.
Introduction
that is they disappeared after a habituation time of several minutes [3,7]. However, it would be interesting to determine whether certain groups of individuals, who might then need special care during examinations and longer habituation times, are more likely to develop such sensory side effects. In addition, one can assume that an MRI examination per se is stressful for participants and that the perceived dizziness might be a result of increased stress or increased stress reactivity. Although the quantity of cortisol and catecholamines, which are associated with higher levels of stress, does not appear to be associated with higher field strengths in MRI [3], one could hypothesize that higher stress reactivity and increased levels of cortisol and catecholamines might contribute toward the magnitude of perceived dizziness. Furthermore, we hypothesized that increased attention to bodily symptoms, which is associated with bodily disorders that cannot be related to an organic disorder, might be predictive of sensory symptoms. To test the hypothesized relationships, we examined the relationship of the variables field strength, bodily symptoms, cortisol, epinephrine, and norepinephrine and the demographic variables sex and age and the development of side effects in static magnetic fields of MRI.
During the last few years, it has been discussed whether the ever-increasing static magnetic fields in MRI influence the health, perception, and cognition of patients and health personnel. As MRI examinations in comparison with computed tomography involve advantages such as the lack of ionizing radiation, it has to be carefully analyzed whether MRI can be harmful for either patients or health personnel. To date, it has been shown that static magnetic fields appear to have no negative impact on physiological functions [1–3]. For cognition, several studies reported small, transient effects such as on recognition memory, which were, in most cases, not replicable [2,4,5] (for a review, see Heinrich et al. [6]). More recent studies reported no effects on cognition [7,8]. However, individuals exposed to static magnetic fields in MRI report diverse sensory effects such as metallic taste, head ringing, dizziness, nystagmus, vertigo, and nausea [1,2,7,9]. In a recent study, we found significantly more dizziness, head ringing, nystagmus, and phosphenes in higher (up to 7 T) compared with lower field strengths [7]. A prime sensory symptom was the presence and strength of dizziness, which was associated with increasing field strengths. Sensory effects were transient, 0959-4965 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
NeuroReport 2014, 25:1081–1084 Keywords: dizziness, magnetic resonance imaging, safety, sensory effects, static magnetic fields Departments of aCognitive and Clinical Neuroscience, bBiostatistics, cPsychiatry and Psychotherapy, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Mannheim and dDepartment of Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Germany Correspondence to Angela Heinrich, PhD, Department of Cognitive and Clinical Neuroscience, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Square J 5, D-68159 Mannheim, Germany Tel: + 49 621 1703 6302; fax: + 49 621 1703 6305; e-mail: [email protected] Received 6 May 2014 accepted 5 June 2014
DOI: 10.1097/WNR.0000000000000225
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1082 NeuroReport 2014, Vol 25 No 14
Table 1
Sensory perceptions
Perception Dizziness Nystagmus Phosphenes Head ringing
0T
1.5 T
3T
2 (2) 2 (2) 0 0
5 (2) 1 (0) 1 (1) 0
10 (6) 1 (1) 1 (1) 0
7T 27 7 6 3
(18) (5) (4) (2)
Note: Numbers of participants who indicated the specific side effect (number of women shown in brackets). Per field strength and complaint 41 (20 for women) would be the maximum number (= all participants would have indicated the particular effect).
Methods This study was carried out in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the University of Heidelberg. Participants were informed about the study and provided signed informed consent. We examined 41 healthy individuals (21 men, 20 women), mean age 26.30 years (SD = 3.74, range 18.86–34.40). All participants were examined in three different MRI scanners of 1.5, 3, and 7 T and a mock scanner with no magnetic field, all separated by 1 week (previously reported in Heinrich et al. [7]). The participants were not aware of the actual field strength. The order of scanners was pseudorandomized to exclude order effects (10 participants started with 0, 1.5, and 3 T and 11 participants started with 7 T, respectively). Perceived sensory side effects were assessed by a questionnaire at the end of each examination day (equating one field strength). Prompted side effects were previously reported side effects of MRI but also fake effects were asked for to control for participants’ reliability. The prompted effects were dizziness, runny nose, nausea, nystagmus, forgetfulness, diplopia, phosphenes, metallic taste, itch, headache, sneezing, concentration problems, fatigue, head ringing, watery eyes, irregular heartbeat, and subsultus. One week before the first examination, participants came for an instruction and examination session, where they were medically examined and completed a questionnaire on bodily symptoms [further also referred to as SOMS – screening for somatoform disorders [10,11]]. This questionnaire examined sensations and pain that affected participants seriously, but no cause could be found by a physician. Cortisol and catecholamine
Cortisol levels were obtained in saliva samples directly in the scanner [3,7]. Blood samples (for the measurement of norepinephrine/epinephrine) were taken directly before the participants were brought into the scanner [3]. Statistical analysis
Dizziness, nystagmus, phosphenes, and head ringing have previously been shown to appear with increasing field strengths in the sample used for the present study (Table 1). We analyzed the impact of different factors on the field strength-related occurrence of dizziness as this side effect occurred in all four conditions (0, 1.5, 3, and
7 T). The other side effects of nystagmus, head ringing, and phosphenes are reported on a descriptive level as their distribution (values of 0) does not allow an analysis of their influence on field strength-related changes in sensory perception. We included sex, age, bodily symptoms, level of cortisol, epinephrine, and norepinephrine together with field strength, first in univariate analyses and then in a joint analysis including all variables. To analyze their effects on the binary outcome dizziness (yes vs. no), we performed logistic regressions with correlated data in a generalized linear model, where we fitted a marginal, population-averaged (GEE-type) model, which accounts for the correlations between measurements. We used the procedure GLIMMIX in SAS 9.2 (SAS Corporation, Cary, North Carolina, USA). We first estimated models with one predictor at a time and finally a model that included all potentially influencing factors. Furthermore, to determine whether effects of sex might be mediated by the other control variables, we performed partial correlations for any field strength separately. We abstained from adjusting for multiple testing because we had clear hypotheses about the main factors of interest including sex, age, and stress hormones, whereas the SOMS was considered a control variable.
Results Univariate analyses showed sex to be associated significantly with the development of dizziness (P = 0.03; Table 2). Women were more prone to feel dizzy than men in higher field strengths, with 90% of them indicating dizziness in 7 T, whereas only 43% of the men reported this side effect (P = 0.003, χ2; Fig. 1). In addition, higher values in the SOMS correlated with a higher risk for dizziness (P = 0.02). As expected, field strength itself had a strong effect on the occurrence of dizziness (P < 0.0001), with a significantly higher risk for dizziness in 7 T (even after adjusting for multiple comparisons) compared with all the remaining field strengths (0, 1.5, and 3 T). Similarly, 3 T led to a higher risk compared with 0 T. Age (P = 0.40), cortisol (P = 0.80), epinephrine (P = 0.53), and norepinephrine (P = 0.52) were not significantly related to dizziness. In a joint analysis (containing field strength together with sex, age, SOMS score, cortisol, epinephrine, and norepinephrine), a Table 2
Results of the logistic regressions
Factors
P-value (single)
F-value (joint)
P-value (joint)
< 0.0001 0.03 0.40 0.02 0.80 0.53 0.52
10.60 12.33 0.47 1.48 0.61 3.58 0.18
< 0.0001 0.0011 0.50 0.23 0.44 0.06 0.67
Field strength Sex Age SOMS score Cortisol Epinephrine Norepinephrine
Note: Results of the logistic regressions with one predictor at a time (single) and all potentially influencing factors at once (joint). SOMS, screening for somatoform disorders.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Influence on sensory perceptions in MRI Heinrich et al. 1083
dizziness in static magnetic fields of MRI. Several previous studies have reported side effects such as metallic taste, head ringing, dizziness, nystagmus, vertigo, and nausea [1,2,9], but usually did not address the question of predictors and did not match by sex.
Fig. 1
Number of participants
20
15
10
5
0 0T
1.5 T
3T
7T
Field strength Number of participants indicating dizziness in all distinct field strengths. White bars, women; black bars, men.
similar risk pattern became obvious (Table 2). Again, we found a significantly higher risk for dizziness with increasing field strengths and for women. However, the SOMS score was not significant in the joint model (Table 2), suggesting stronger effects for field strength and sex and no additional information gain for the SOMS score, cortisol, epinephrine, and norepinephrine. Partial correlations controlling for the SOMS score, performed for all field strengths separately, showed no significant correlations between sex and dizziness for 0, 1.5, and 3 T (all P > 0.05). However, the correlation between sex and dizziness was significant in 7 T when controlling for the SOMS score, even after correcting for multiple testing (P = 0.002). The same pattern occurred for cortisol, epinephrine, and norepinephrine: there were no significant correlations in 0, 1.5, and 3 T, but highly significant correlations between sex and dizziness controlling for cortisol (P = 0.001), epinephrine (P = 0.001), and norepinephrine (P
Women are more strongly affected by dizziness in static magnetic fields of magnetic resonance imaging scanners Angela Heinricha, Anne Szosteka, Patric Meyera, Iris Reinhardb, Maria Gillesc, Georgios Paslakisc, Jaane Rauschenbergd, Jens Gröbnerd, Wolfhard Semmlerd, Michael Deuschlec, Andreas Meyer-Lindenbergc, Herta Flora and Frauke Neesa Increasing field strengths in MRI necessitate the examination of potential side effects. Previously reported results have been contradictory, possibly caused by imbalanced samples. We aimed to examine whether special groups of people are more prone to develop side effects that might have led to contradictory results in previous studies. We examined the occurrence of sensory side effects in static magnetic fields of MRI scanners of 1.5, 3, and 7 T and a mock scanner in 41 healthy participants. The contribution of field strength, sex, age, and attention to bodily processes, and stress hormone levels to the sensation of dizziness was examined in separate univariate analyses and in a joint analysis that included all variables. Field strength and sex were significant factors in the joint analysis (P = 0.001), with women being more strongly affected than men by dizziness in higher static magnetic fields. This effect was not mediated by the other variables such as attention to bodily symptoms or stress hormones. Further research needs to elucidate the underlying factors
of increased dizziness in women in static magnetic fields in MRI. We hypothesize that imbalanced samples of earlier studies might be one reason for previous contradictory results on the side effects of static magnetic fields. NeuroReport 25:1081–1084 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins.
Introduction
that is they disappeared after a habituation time of several minutes [3,7]. However, it would be interesting to determine whether certain groups of individuals, who might then need special care during examinations and longer habituation times, are more likely to develop such sensory side effects. In addition, one can assume that an MRI examination per se is stressful for participants and that the perceived dizziness might be a result of increased stress or increased stress reactivity. Although the quantity of cortisol and catecholamines, which are associated with higher levels of stress, does not appear to be associated with higher field strengths in MRI [3], one could hypothesize that higher stress reactivity and increased levels of cortisol and catecholamines might contribute toward the magnitude of perceived dizziness. Furthermore, we hypothesized that increased attention to bodily symptoms, which is associated with bodily disorders that cannot be related to an organic disorder, might be predictive of sensory symptoms. To test the hypothesized relationships, we examined the relationship of the variables field strength, bodily symptoms, cortisol, epinephrine, and norepinephrine and the demographic variables sex and age and the development of side effects in static magnetic fields of MRI.
During the last few years, it has been discussed whether the ever-increasing static magnetic fields in MRI influence the health, perception, and cognition of patients and health personnel. As MRI examinations in comparison with computed tomography involve advantages such as the lack of ionizing radiation, it has to be carefully analyzed whether MRI can be harmful for either patients or health personnel. To date, it has been shown that static magnetic fields appear to have no negative impact on physiological functions [1–3]. For cognition, several studies reported small, transient effects such as on recognition memory, which were, in most cases, not replicable [2,4,5] (for a review, see Heinrich et al. [6]). More recent studies reported no effects on cognition [7,8]. However, individuals exposed to static magnetic fields in MRI report diverse sensory effects such as metallic taste, head ringing, dizziness, nystagmus, vertigo, and nausea [1,2,7,9]. In a recent study, we found significantly more dizziness, head ringing, nystagmus, and phosphenes in higher (up to 7 T) compared with lower field strengths [7]. A prime sensory symptom was the presence and strength of dizziness, which was associated with increasing field strengths. Sensory effects were transient, 0959-4965 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
NeuroReport 2014, 25:1081–1084 Keywords: dizziness, magnetic resonance imaging, safety, sensory effects, static magnetic fields Departments of aCognitive and Clinical Neuroscience, bBiostatistics, cPsychiatry and Psychotherapy, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Mannheim and dDepartment of Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Germany Correspondence to Angela Heinrich, PhD, Department of Cognitive and Clinical Neuroscience, Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Square J 5, D-68159 Mannheim, Germany Tel: + 49 621 1703 6302; fax: + 49 621 1703 6305; e-mail: [email protected] Received 6 May 2014 accepted 5 June 2014
DOI: 10.1097/WNR.0000000000000225
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1082 NeuroReport 2014, Vol 25 No 14
Table 1
Sensory perceptions
Perception Dizziness Nystagmus Phosphenes Head ringing
0T
1.5 T
3T
2 (2) 2 (2) 0 0
5 (2) 1 (0) 1 (1) 0
10 (6) 1 (1) 1 (1) 0
7T 27 7 6 3
(18) (5) (4) (2)
Note: Numbers of participants who indicated the specific side effect (number of women shown in brackets). Per field strength and complaint 41 (20 for women) would be the maximum number (= all participants would have indicated the particular effect).
Methods This study was carried out in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the University of Heidelberg. Participants were informed about the study and provided signed informed consent. We examined 41 healthy individuals (21 men, 20 women), mean age 26.30 years (SD = 3.74, range 18.86–34.40). All participants were examined in three different MRI scanners of 1.5, 3, and 7 T and a mock scanner with no magnetic field, all separated by 1 week (previously reported in Heinrich et al. [7]). The participants were not aware of the actual field strength. The order of scanners was pseudorandomized to exclude order effects (10 participants started with 0, 1.5, and 3 T and 11 participants started with 7 T, respectively). Perceived sensory side effects were assessed by a questionnaire at the end of each examination day (equating one field strength). Prompted side effects were previously reported side effects of MRI but also fake effects were asked for to control for participants’ reliability. The prompted effects were dizziness, runny nose, nausea, nystagmus, forgetfulness, diplopia, phosphenes, metallic taste, itch, headache, sneezing, concentration problems, fatigue, head ringing, watery eyes, irregular heartbeat, and subsultus. One week before the first examination, participants came for an instruction and examination session, where they were medically examined and completed a questionnaire on bodily symptoms [further also referred to as SOMS – screening for somatoform disorders [10,11]]. This questionnaire examined sensations and pain that affected participants seriously, but no cause could be found by a physician. Cortisol and catecholamine
Cortisol levels were obtained in saliva samples directly in the scanner [3,7]. Blood samples (for the measurement of norepinephrine/epinephrine) were taken directly before the participants were brought into the scanner [3]. Statistical analysis
Dizziness, nystagmus, phosphenes, and head ringing have previously been shown to appear with increasing field strengths in the sample used for the present study (Table 1). We analyzed the impact of different factors on the field strength-related occurrence of dizziness as this side effect occurred in all four conditions (0, 1.5, 3, and
7 T). The other side effects of nystagmus, head ringing, and phosphenes are reported on a descriptive level as their distribution (values of 0) does not allow an analysis of their influence on field strength-related changes in sensory perception. We included sex, age, bodily symptoms, level of cortisol, epinephrine, and norepinephrine together with field strength, first in univariate analyses and then in a joint analysis including all variables. To analyze their effects on the binary outcome dizziness (yes vs. no), we performed logistic regressions with correlated data in a generalized linear model, where we fitted a marginal, population-averaged (GEE-type) model, which accounts for the correlations between measurements. We used the procedure GLIMMIX in SAS 9.2 (SAS Corporation, Cary, North Carolina, USA). We first estimated models with one predictor at a time and finally a model that included all potentially influencing factors. Furthermore, to determine whether effects of sex might be mediated by the other control variables, we performed partial correlations for any field strength separately. We abstained from adjusting for multiple testing because we had clear hypotheses about the main factors of interest including sex, age, and stress hormones, whereas the SOMS was considered a control variable.
Results Univariate analyses showed sex to be associated significantly with the development of dizziness (P = 0.03; Table 2). Women were more prone to feel dizzy than men in higher field strengths, with 90% of them indicating dizziness in 7 T, whereas only 43% of the men reported this side effect (P = 0.003, χ2; Fig. 1). In addition, higher values in the SOMS correlated with a higher risk for dizziness (P = 0.02). As expected, field strength itself had a strong effect on the occurrence of dizziness (P < 0.0001), with a significantly higher risk for dizziness in 7 T (even after adjusting for multiple comparisons) compared with all the remaining field strengths (0, 1.5, and 3 T). Similarly, 3 T led to a higher risk compared with 0 T. Age (P = 0.40), cortisol (P = 0.80), epinephrine (P = 0.53), and norepinephrine (P = 0.52) were not significantly related to dizziness. In a joint analysis (containing field strength together with sex, age, SOMS score, cortisol, epinephrine, and norepinephrine), a Table 2
Results of the logistic regressions
Factors
P-value (single)
F-value (joint)
P-value (joint)
< 0.0001 0.03 0.40 0.02 0.80 0.53 0.52
10.60 12.33 0.47 1.48 0.61 3.58 0.18
< 0.0001 0.0011 0.50 0.23 0.44 0.06 0.67
Field strength Sex Age SOMS score Cortisol Epinephrine Norepinephrine
Note: Results of the logistic regressions with one predictor at a time (single) and all potentially influencing factors at once (joint). SOMS, screening for somatoform disorders.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Influence on sensory perceptions in MRI Heinrich et al. 1083
dizziness in static magnetic fields of MRI. Several previous studies have reported side effects such as metallic taste, head ringing, dizziness, nystagmus, vertigo, and nausea [1,2,9], but usually did not address the question of predictors and did not match by sex.
Fig. 1
Number of participants
20
15
10
5
0 0T
1.5 T
3T
7T
Field strength Number of participants indicating dizziness in all distinct field strengths. White bars, women; black bars, men.
similar risk pattern became obvious (Table 2). Again, we found a significantly higher risk for dizziness with increasing field strengths and for women. However, the SOMS score was not significant in the joint model (Table 2), suggesting stronger effects for field strength and sex and no additional information gain for the SOMS score, cortisol, epinephrine, and norepinephrine. Partial correlations controlling for the SOMS score, performed for all field strengths separately, showed no significant correlations between sex and dizziness for 0, 1.5, and 3 T (all P > 0.05). However, the correlation between sex and dizziness was significant in 7 T when controlling for the SOMS score, even after correcting for multiple testing (P = 0.002). The same pattern occurred for cortisol, epinephrine, and norepinephrine: there were no significant correlations in 0, 1.5, and 3 T, but highly significant correlations between sex and dizziness controlling for cortisol (P = 0.001), epinephrine (P = 0.001), and norepinephrine (P
