Supplementary Material
MR imaging in the BiDirect Study
Technical details 2
Detailed description of the examination procedure 4
Visualisation of frequencies of contraindications and refusals 6
Analysis of further explanatory factors for MRI refusal 7
References10
Technical details
Magnetic resonance imaging (MRI) is performed on a single 3.0 T MRI scanner (Intera with Achieva update; Philips Medical Systems, Best, The Netherlands) operated by the Department of Clinical Radiology, University Hospital Münster. To reduce the number of potential contraindications a transmit/receive (T/R) head coil is used despite drawbacks with regard to acquisition time, as, for instance, no parallel imaging is possible. The limitation of the high frequency excitation to the head area (with leakage field in neck area) allows examining subjects with electrically conductive implants in lower body parts in most cases.
The 35-minute MRI protocol includes a T2* weighted FFE sequence, a TSE-FLAIR sequence, a T2 weighted TSE sequence (first follow-up examination and beyond), a T1 weighted 3D TFE sequence with preceding inversion pulse, a diffusion-weighted sequence and a resting state functional MRI (rs-fMRI) sequence. Subjects with a minimum visual acuity of 0.5 (measured with a Snellen chart; contact lenses permitted) underwent an additional functional MRI (fMRI) sequence with a visually presented emotion processing task (baseline examination only) [1–4]. Technical information on the protocol is provided in Table S1. No contrast agents were used.
1
Table S1: Protocol for cerebral MRI in the BiDirect Study
in-plane resolution / slicesTR
[ms] / TE
[ms] / TI
[ms] / FA / matrix / FOV
[mm x mm] / reconstructed
res. [mm x mm] / n / thickness
[mm] / gap
[mm] / orien-
tation
2D T2* weighted gradient echo sequence (FFE) / 574 / 16 / - / 18° / 256 x 164 / 230 x 183 / 0.45 x 0.45 / 27 / 4 / 1 / ax
2D fast dark fluid imaging sequence (TSE-FLAIR) / 11000 / 120 / 2600 / 90° / 352 x 206 / 230 x 186 / 0.45 x 0.45 / 27 / 4 / 1 / ax
2D T2 weighted fast spin echo sequence (TSE) a) / 3000 / 80 / - / 90° / 400 x 255 / 230 x 184 / 0.45 x 0.45 / 27 / 4 / 1 / ax
3D T1 weighted gradient echo sequence with inversion prepulse (3D TFE) / 7.26 / 3.56 / 404 / 9° / 256 x 256 / 256 x 256 / 1.00 x 1.00 / 160 / b) 2 / - / sag
2D diffusion-weighted sequence with echo planar imaging (Single Shot SE-EPI) c) / 5900 / 95 / - / 90° / 128 x 128 / 240 x 240 / 0.94 x 0.94 / 36 / 3.6 / - / ax
Stimulation-based fMRI sequence with echo planar imaging (Single Shot FFE-EPI), 82 volumes after five dummy scans d) / 2200 / 30 / - / 90° / 64 x 64 / 230 x 230 / 3.60 x 3.60 / 35 / 3.6 / - / ax
Resting State fMRI sequence with echo planar imaging (Single Shot FFE-EPI), 72 volumes after five dummy scans e) / 3000 / 38 / - / 90° / 64 x 64 / 230 x 230 / 3.60 x 3.60 / 36 / 3.6 / - / ax
TR: repetition time, TE: echo time, TI: inversion time, FA: flip angle, FOV: field of view, fMRI: functional magnetic resonance imaging
a) 1st follow-up examination and beyond b) reconstructed by zero filling in k-space to 1 mm slice thickness c) 20 gradient directions with b value of 1000 s/mm2, reference b value is 0 s/mm2
d) baseline examination only. The emotion processing task is a short version of a previously published fMRI paradigm investigating neural responsiveness to happy and sad facial expressions in patients with major depression [1-3]. Facial stimuli consist of happy, sad and neutral expressions according to Ekman & Friesen [4]. The passive viewing task with a presentation time of 3 minutes is subdivided into 6 blocks of 30 seconds each. During the first 20 seconds of a block, facial stimuli falling in the same category are presented for 500 ms each in a random sequence. The remaining 10 seconds of a block serve as no-face epoch. The order of blocks is sad-neutral-happy-sad-neutral-happy.
e) Prior to this sequence participants were instructed to remain motionless, mainly keep their eyes open, not to fall asleep and not to think of anything in particular.
1
Detailed description of the examination procedure
Prior to their appointment, all participants received written information on the planned MRI examination and contraindications (CIs) for MRI, giving them the opportunity to clarify potential CIs in advance along with their treating physician and/ or members of the study team.
As part of the written informed consent form, study participants answer an 18-point questionnaire regarding surgeries, implants, tattoos, metalworking activities, seizure disorders and pregnancy. Based on this information, members of the study team exclude participants with definite CI from the MRI examination. For the remaining subjects, provided the subject agrees, the completed questionnaire serves as a basis for the pre-examination discussion with the radiologist.
An MR physicist and the responsible radiologist make the final decision on all MRI examinations. Thereby, a strict regime of CI classification was obeyed, as the potential risk of imaging with implants in this volunteer study cannot be balanced by an individual diagnostic benefit. Precisely, all active implants (‘Active Implantable Medical Devices’, AIMDs) and implants in the head/ neck region are considered as CI. Implants labelled as ‘MR conditional’ according to ASTM 2503/ IEC 62570 are not considered as CI given that they are positioned at least 15 cm outside of the edge of the T/R head coil, since implant heating by the radio frequency field is not relevant under these conditions [5, 6]. The presence of more than one cardiac stent, however, is considered as CI, even if both stents are labelled ‘MR conditional’ and required conditions are met by the imaging procedure, as information on potential radio frequency amplification effects of the stent combination is not available. Moreover, implants of any type, for which information on MR safety could not be provided, are considered as CI.
Self-reported claustrophobia is not classified as CI, but understood as (voluntary) refusal. Administration of tranquillizers prior to the procedure is not offered, however, patients with depression are allowed to take their anxiolytic as-required medications (ingestion is recorded).
Study participants can decide on their agreement on MRI independent from participation in other study modules and separately for each examination wave. Potential CIs are re-determined at every appointment. Specifically for subjects of cohort 2, who wish to be scanned after having been excluded from MRI at baseline examination due to recently implanted cardiac stents, a questionnaire on safety-related data of the implantation (time and place of surgery, manufacturer, model, material and length of stent, number of stents, specification of complications) was designed. Participants are encouraged to collect this information from their treating physician to accelerate the decision on MRI approval.
A flowchart illustrating the consecutive stages that are passed by the participants in order to undergo MRI is shown in Figure S1.
Figure S1: Flowchart illustrating the consecutive stages that are passed by participants of the BiDirect Study in order to undergo magnetic resonance imaging (MRI) of the brain. For baseline (BL) as well as follow-up (FU) examination, the clarification and assessment of contraindications (CIs), the possibilities to refuse the examination and the disclosure of incidental findings (IFs) are presented according to the actual workflow. Outcomes that are discussed in the paper are highlighted in purple.
Illustration not to scale.
Visualisation of frequencies of contraindications and refusals
Figure S2:Flowchart illustrating the number of participants included in the evaluable study population of the BiDirect study (upper panel), the number of participants with and without contraindications (CIs, middle panel) and the number of participants, who agreed or refused to undergo magnetic resonance imaging (MRI) examination of the brain (bottom panel). Data came from the baseline examination wave. Additionally, the distribution of participants across cohorts is shown for each panel. Panel lengths are scaled according to the respective number of participants.
Analysis of further explanatory factors for MRI refusal
Beyond age, gender and depression, there are further potential explanatory factors for the participants’ tendency to refuse the MRI examination despite having no medical contraindication.In the BiDirect study we have information on education, anxiety and cognitive function, which are likely to play a role for refusal. So we here present an extended analysis of the proportion of refusal including these factors.
Education
The educational level of participants was categorised according to Table S2 with level definitionsabided by the International Standard Classification of Education (ISCED), which was developed by the United Nations Educational, Scientific and Cultural Organization (UNESCO) [7].
Table S2: Categorisation of educational level
0 / Low-skilled (participants without upper secondary education, neither general nor vocational; corresponds to ISCED-97 levels 1 and 2; typically 10 or less years of education)1 / Vocational qualification (participants with upper secondary or higher vocational education; corresponds to ISCED-97 levels 3, 4 and 5B; typically 13-16 years of education)
2 / University level (participants with general higher education; corresponds to ISCED-97 level 5A; typically 17-20 years of education)
ISCED: International Standard Classification of Education
Anxiety
The frequency of anxiety disorders was determined by asking the participants “Have you ever been diagnosed with an anxiety disorder by a physician?” during a computer-assisted interview. As this particular question was included in the interview two month after the study had begun, this information is missing in several cases.
Cognitive function
A cognitive functioning module including the assessment of cognitive processing speed (Trail Making Test A), verbal memory (12-item word list), verbal fluency (animal naming test) and executive functions (Trail Making Test B, Colour-Word-Interference Test) is included in the study’s examination program (please see [8] for a detailed description). Cognitive results (i.e. time needed for Trail Making Tests A and B, respectively, time difference between trial 3 and trial 2 of the Colour-Word-Interference Test, number of correctly recalled words immediately after presentation, number of correctly named animals within 60 seconds) were combined into a global cognitive score (standardised z-score) using principal component analysis.
For the evaluable study population, the distribution of educational level, frequency of formerly diagnosed anxiety disorders and global cognitive z-score across cohorts is shown in Table S3. The results of the logistic regression analysis based on the extended model (in comparison to the model, which was used to obtain Table 4) are summarised in Table S4.
In summary, although the distribution of educational level variedacross cohorts, we did not find any association between education and refusal in our data. Also, the former diagnosis of an anxiety disorder had no independent effect on refusal rates. However, anxiety constitutes—also from a statistical point of view—a comorbidity of depression (i.e. replacing depression with anxiety in the initial model for cohorts 1+3 yields practically the same regression coefficients, data not shown). For patients with depression (cohort 1), our data indicate that patients with a higher cognitive z-score, i.e. better cognitive performance, are less likely to refuse the MRI examination. As comorbid dementia was an exclusion criterion, we hypothesise that cognitive function in cohort 1 is indicativeof the severity of depression, i.e. that the risk of refusal increases with an increasing severity of depression.
Table S3: Distribution of educational level, frequency of formerly diagnosed anxiety disorders and global z-score of the cognitive functioning module across cohorts for the evaluable study population
Cohort 1(depression) / Cohort 2
(CVD) / Cohort 3
(control) / All
N / 988 / 345 / 843 / 2176
Educational levela
0 / 69 ( 7.0%) / 15 ( 4.3%) / 38 ( 4.5%) / 122 ( 5.6%)
1 / 679 (68.7%) / 230 (66.7%) / 464 (55.0%) / 1373 (63.1%)
2 / 237 (24.0%) / 99 (28.7%) / 341 (40.5%) / 677 (31.1%)
NAb / 3 ( 0.3%) / 1 ( 0.3%) / 0 ( 0.0%) / 4 ( 0.2%)
Diagnosis of anxiety disorder
ever / 370 (37.4%) / 15 ( 4.3%) / 49 ( 5.8%) / 434 (19.9%)
never / 589 (59.6%) / 328 (95.1%) / 748 (88.7%) / 1665 (76.5%)
NAc / 29 ( 2.9%) / 2 ( 0.6%) / 46 ( 5.5%) / 77 ( 3.5%)
Cognitive test results
z-score, mean ± SD / -0.09 ± 1.04 / -0.14 ± 0.89 / 0.17 ± 0.97 / 0.00 ± 1.00
NA d / 25 / 13 / 25 / 63
CVD: cardiovascular disease, N: number of participants, NA: not available, SD: standard deviation
Data are presented as frequencies unless specified otherwise.a Educational level is categorised according to Table S2.b Data are not available for subjects who terminated their examination in the study centre prior to interview or refused the interview.cData are not available for subjects who had their examination within the first two months of the baseline examination, who terminated their examination in the study centre prior to interview or refused the interview.dMissing data result from participants, who terminated their examination in the study centre prior to cognitive testing or refused cognitive testing or from participants with impaired hearing or non-native speakers.
Table S4:Odds Ratios with 95% confidence intervals (OR) for the association of age, gender, education, cognitive function, anxiety and depression with the proportion of refusal. a
refusalN / OR
Cohort 1 (depression) / 895 – 47 = 848
Age [years] / 1.01 / [0.99; 1.04]
Female / 1.56 / [1.07; 2.32]
Education [levels] b / 1.09 / [0.77; 1.53]
Cognitive function [SD] c / 0.82 / [0.68; 0.98]
Diagnosis of anxiety disorder d / 1.31 / [0.91; 1.87]
Cohort 2 (CVD) / 83 – 2 = 81
Age [years] / 0.97 / [0.90; 1.06]
Female / 0.49 / [0.12; 1.67]
Education [levels] b / 0.62 / [0.26; 1.42]
Cognitive function [SD] c / 1.54 / [0.80; 3.11]
Diagnosis of anxiety disorder d / 2.10 / [0.23; 19.34]
Cohort 3 (control) / 768 – 66 = 702
Age [years] / 1.03 / [1.00; 1.06]
Female / 1.89 / [1.16; 3.16]
Education [levels] b / 0.80 / [0.52; 1.23]
Cognitive function [SD] c / 1.21 / [0.92; 1.63]
Diagnosis of anxiety disorder d / 1.44 / [0.62; 3.03]
Cohorts 1+3 / 1663 – 113 = 1550
Age [years] / 1.02 / [1.00; 1.04]
Female / 1.73 / [1.28; 2.35]
Depression / 1.36 / [0.99; 1.87]
Education [levels] b / 0.97 / [0.74; 1.26]
Cognitive function [SD] c / 0.93 / [0.80; 1.08]
Diagnosis of anxiety disorder d / 1.35 / [0.97; 1.87]
N: number of participants – number of datasets with missing value in any explanatory variable = number of evaluated datasets, CVD: cardiovascular disease, SD: standard deviation
a All categorial variables are coded with 1=’yes’/ 0=’no’ unless specified otherwise. b The categorisation of educational levels is specified in Table S2. c Cognitive function is presented by a global standardised z-score reflecting global cognitive performance. d Diagnosis of anxiety disorder is coded with 1=’ever’ / 0=’never’.
References
1. Baune BT, Dannlowski U, Domschke K et al (2010) The interleukin 1 beta (IL1B) gene is associated with failure to achieve remission and impaired emotion processing in major depression. Biol Psychiatry 67:543–549
2. Dannlowski U, Ohrmann P, Bauer J et al (2007) Serotonergic genes modulate amygdala activity in major depression. Genes Brain Behav 6:672–676
3. Dannlowski U, Ohrmann P, Konrad C et al (2009) Reduced amygdala–prefrontal coupling in major depression: association with MAOA genotype and illness severity. Int J Neuropsychopharmacol 12:11–22
4. Ekman P, Friesen WV (1976) Pictures of Facial Affect. Consulting Psychologists Press, Palo Alto, CA
5. ASTM International (2013) ASTM F2503-13: Standard Practice for Marking Medical Devices and Other Items for Safety in the Magnetic Resonance Environment. West Conshohocken, PA, USA.
6. IEC (2014) IEC 62570 Ed.1.0: Standard practice for marking medical devices and other items for safety in the magnetic resonance environment.
7. UNESCO: ISCED: International Standard Classification of Education. Available via Accessed 5 Dec 2015
8. Teismann H, Wersching H, Nagel M et al (2014) Establishing the bidirectional relationship between depression and subclinical arteriosclerosis - rationale, design, and characteristics of the BiDirect Study. BMC Psychiatry 14:174
1