You are here

  • Home
  • Archive
  • Volume 24, Issue 4
  • Prevalences of comorbid anxiety disorder and daily smartphone-based self-reported anxiety in patients with newly diagnosed bipolar disorder

Article Text

Article menu

Download PDFPDF

Digital mental health
Original research
Prevalences of comorbid anxiety disorder and daily smartphone-based self-reported anxiety in patients with newly diagnosed bipolar disorder
Free
  1. Sharleny Stanislaus1,
  2. Klara Coello1,
  3. Hanne Lie Kjærstad1,
  4. Kimie Stefanie Ormstrup Sletved1,
  5. Ida Seeberg1,
  6. Mads Frost2,
  7. Jakob Eyvind Bardram3,
  8. Rasmus Nejst Jensen4,
  9. Maj Vinberg4,5,
  10. Maria Faurholt-Jepsen1,
  11. Lars Vedel Kessing1,5
  1. 1 Psychiatric Center Copenhagen, Rigshospitalet, Region Hovedstadens Psykiatri, Kobenhavn, Hovedstaden, Denmark
  2. 2 Monsenso, Monsenso Aps, Copenhagen, Denmark
  3. 3 Department of Health Technology, Technical University of Denmark, Lyngby, Denmark
  4. 4 Psychiatric Centre North Zealand, Region Hovedstadens Psykiatri, Hilleroed, Hovedstaden, Denmark
  5. 5 Faculty of Health Sciences, University of Copenhagen, Kobenhavn, Denmark
  1. Correspondence to Dr Sharleny Stanislaus, Psychiatric Center Copenhagen, Rigshospitalet, Region Hovedstadens Psykiatri, 2100 Kobenhavn, Hovedstaden, Denmark; sharleny.stanislaus.01{at}regionh.dk

Abstract

Background Around 40% of patients with bipolar disorder (BD) additionally have anxiety disorder. The prevalence of anxiety in patients with newly diagnosed BD and their first-degree relatives (UR) has not been investigated.

Objective

To investigate (1) the prevalence of a comorbid anxiety diagnosis in patients with newly diagnosed BD and their UR, (2) sociodemographic and clinical differences between patients with and without a comorbid anxiety diagnosis and (3) the association between smartphone-based patient-reported anxiety and observer-based ratings of anxiety and functioning, respectively.

Methods We recruited 372 patients with BD and 116 of their UR. Daily smartphone-based data were provided from 125 patients. SCAN was used to assess comorbid anxiety diagnoses.

Findings In patients with BD, the prevalence of a comorbid anxiety disorder was 11.3% (N=42) and 10.3% and 5.9% in partial and full remission, respectively. In UR, the prevalence was 6.9%. Patients with a comorbid anxiety disorder had longer illness duration (p=0.016) and higher number of affective episodes (p=0.011). Smartphone-based patient-reported anxiety symptoms were associated with ratings of anxiety and impaired functioning (p<0.001).

Limitations The SCAN interviews to diagnose comorbid anxiety disorder were carried out regardless of the participants’ mood state.

Clinical implications

The lower prevalence of anxiety in newly diagnosed BD than in later stages of BD indicates that anxiety increases with progression of BD. Comorbid anxiety seems associated with poorer clinical outcomes and functioning and smartphones are clinically useful for monitoring anxiety symptoms.

Trial registration number ClinicalTrials.gov Registry (NCT02888262).

  • anxiety disorders
  • depression & mood disorders
  • adult psychiatry

Data availability statement

No data are available. The study is ongoing; therefore, the research data are not shared.

https://doi.org/10.1136/ebmental-2021-300259

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Bipolar disorder (BD) is characterised by alterations in mood and energy as well as a range of symptoms, including anxiety symptoms. In BD, the lifetime prevalence of anxiety disorders has in two meta-analyses been estimated to be 40.5% (39 studies including 13 409 patients) and 45% (40 studies including 14 914 patients with BD, ranging from 8% to 88%), respectively.1 2 In comparison, the current global prevalence of any anxiety disorder in the general population is approximately 7.3% (ranging from 0.9% to 28.3%).3 The high variability in the prevalence of anxiety diagnosis most likely represents the heterogeneous study populations and methods applied for assessing anxiety across institutions and nations. Further a proportion of patients with BD may experience subsyndromal anxiety symptoms during remitted phases and during episodes without fulfilling the criteria for anxiety diagnosis.4 5 The co-occurrence of anxiety symptoms is related to worsening of clinical outcomes in patients with BD, such as increased self-reported stress,4 shorter periods with remission6 impaired functioning, reduced quality of life,4 more severe mood episodes7 8 and a higher risk of suicide attempts.8 9 Also, anxiety disorders and anxiety symptoms in offspring of patients with BD have been associated with increased risk of onset of a mood disorder.10 Consequently, anxiety is important to acknowledge both in patients in remission and in individuals with risk of developing BD. Smartphones offer a unique method to monitor daily anxiety symptoms for a long term and unobtrusively and are likely to be a feasible method to detect anxiety symptoms during remission and in high-risk individuals.

    No study has previously investigated the prevalence of a comorbid anxiety diagnosis and daily self-reported anxiety symptoms in patients with newly diagnosed BD and their first-degree relatives and the association with clinical characteristics and functioning.

    This study aimed to investigate (1) the prevalence of anxiety diagnosis in patients with newly diagnosed BD and their first-degree relatives, (2) the differences in sociodemographic data and clinical characteristics in patients with and without a comorbid anxiety diagnosis and with and without smartphone-based self-reported anxiety, (3) the association between smartphone-based patient-reported anxiety and anxiety ratings on validated observer-based rating scales and (4) finally, we aimed to investigate the association between smartphone-based patient-reported anxiety and impairment in functioning in patients with newly diagnosed BD.

    We predicted that: (1) the prevalence of a comorbid anxiety diagnosis and symptoms would be lower in patients with newly diagnosed BD than in previously reported studies with samples of patients with varying duration of illness; (2) the prevalence of a comorbid anxiety diagnosis would be associated with the severity of mood symptoms; (3) patients with newly diagnosed BD and a comorbid anxiety diagnosis or self-reported anxiety would have more severe illness characteristics and impaired functioning; (4) smartphone-based patient-reported anxiety level would be associated with the observer-based rating of anxiety levels and (5) higher level of smartphone-based patient-reported anxiety would be associated with a lower level of functioning.

    Methods

    Study design

    The present study is part of the larger ongoing Bipolar Illness Onset (BIO) Study.11 Participants included in this report were included in the study from June 2015 to February 2020. All participants were assessed at baseline and then annually and additionally, if they experienced new episodes.

    Study participants

    Patients with newly diagnosed BD were recruited from the Copenhagen Affective Disorder Clinic, Denmark. The clinic offers a 2-year programme to all newly diagnosed patients with BD in the entire Capital Region of Denmark.12 All patients in the Capital Region are referred to the clinic when the diagnosis of a single manic episode or bipolar is made for the first time. The clinic receives patients from general practitioners, private psychiatrists in primary care and psychiatrists in secondary care following hospitalisation and from outpatient treatment. The criteria for inclusion were newly diagnosis with BD or a single manic episode according to the International Classification of Diseases (ICD) criteria and an age of 18–70 years. With permission from the patients with BD researchers contacted their first-degree relatives (siblings and children). The inclusion criterion for first-degree relatives was the age of 15–70 years and exclusion criteria were organic mental disorders, mental and behavioural disorders due to psychoactive substance use, schizophrenia or other psychotic disorders or BD.

    Healthy control (HC) individuals without any current or prior treatment requiring psychiatric disorder in the individual or among the individual’s first-degree relatives were recruited among blood donors from the blood bank at Rigshospitalet, Copenhagen University Hospital. The HC included in this paper was both from the BIO cohort and another study completed by our group, which used the same inclusion and exclusion criteria.13 The control group from the other study was used because all participants provided smartphone-based daily recordings of anxiety. The HC was only used in the smartphone-based analyses.

    Ratings of anxiety and questionnaires

    At inclusion, a medical doctor or psychologist trained in diagnosing BD confirmed the diagnosis of BD and assessed any psychiatric comorbidities, including anxiety disorders, according to the ICD 10th version (ICD-10)14 using the Schedules of Clinical Assessment in Neuropsychiatry (SCAN) interview.15 Anxiety diagnoses included any diagnosis from F40.0 (agoraphobia) to F42.9 (obsessive-compulsive disorder). During the clinical interview, number of previous episodes was estimated based on retrospective patient reports. Previous episode includes hypomanic, manic, mixed and depressive episodes. The severity of depressive and manic symptoms for the past 3 days was clinically evaluated using the Hamilton Depression Rating Scale 17 items (HAMD-17)16 and the Young Mania Rating Scale (YMRS),17 respectively. For the HAMD-17, we used subitem 10 to assess psychiatric anxiety and subitem 11 to assess somatic anxiety. The Functional Assessment Short Test (FAST) was included to investigate whether changes in daily smartphone-based patient-reported anxiety symptoms were reflected in changes in functioning, as assessed by clinical researchers. FAST is specifically developed for BD and addresses six areas of functioning for the past 14 days: autonomy, occupational functioning, cognitive functioning, financial issues, interpersonal relationship and leisure time.18 FAST scores 12, 20 and 40 are used as cut-offs for mild, moderate and severe functional impairment.19

    The EuroQol-5 Domain (EQ-5D), a standardised measure of health status, was used to provide a simple and easily accessible measure of health across several disease areas.20

    Smartphone-based patient-reported anxiety

    Participants recruited after September 2016 were invited to download a smartphone-based application, Monsenso, on their smartphones.21 It was voluntary to use the application and therefore smartphone-based recordings were not provided from all participants. The Monsenso app is available on both iPhone and Android smartphones. The patients with BD self-reported their mood, activity and sleep on a daily basis. In addition to these three measures, the participants could choose to rate daily anxiety symptoms and stress symptoms. Anxiety symptoms were rated on a scale from 0 (‘no anxiety symptoms’) to 4 (‘high levels of anxiety symptoms’) (0–4). Likewise, the level of overall daily stress was rated on a scale from 0 (‘no stress’) to 4 (‘high levels of stress’). It was optional to report daily anxiety symptoms. Consequently, it was only a subpopulation of the patients with BD who provided smartphone-based recordings that provided daily self-reports of anxiety symptoms. Only three participants in the first-degree relative group provided daily self-reports of anxiety symptoms, due to the low number of participants we did not include data from this group in the analysis. Likewise, only a few of the HCs in the BIO study provided daily anxiety ratings and data from this group were therefore pooled with data from a previous study using a similar control group with similar inclusion and exclusion criteria.13 The Monsenso system has a daily reminder function and self-reported data can be entered retrospectively for up to 2 days.

    Statistical methods

    All predictions and statistical analyses were planned a priori. Differences in demographic variables between two groups (patients with BD vs first-degree relatives, patients with BD with and without a comorbid anxiety diagnosis, patients with BD with and without daily patient-reported anxiety) were analysed using χ2 for categorical variables and t-test or Mann-Whitney test for continuous variables depending on whether the assumptions of normality were met or not met. Continuous variables are presented as median (IQR) or mean (95% CI) and categorical data are presented as % (n) in the tables.

    We employed a linear mixed-effect regression model with smartphone-based anxiety as the dependent variable and observer-based ratings and other smartphone measures (patient-reported mood, activity and stress) as fixed factors. Due to the longitudinal design of the study, some participants provided repeated measurements of HAMD-17, YMRS and FAST and smartphone-based self-reported symptoms. To account for this, the participants’ identification numbers were included as a random factor in a linear mixed-effect model. In this way, this model enables us to account for both the variation of the variable of interest within the participants (intraindividual variation) and between individuals (interindividual variation). For all analyses, an unadjusted model and a model adjusted for age and sex were employed. Also, model assumptions were checked using QQ plot and histogram of residuals for all analyses.

    Smartphone-based recordings of anxiety symptoms are a new area of research; therefore, due to the exploratory nature of the study, adjustment for multiple testing was not applied. The level of statistical significance was set to p values below 0.05. All statistical analyses were performed using SPSS software (V.25.0; IBM Corp., Armonk, New York, USA).

    Findings

    Sociodemographic and clinical characteristics

    We included 372 patients with BD. A total of 11.3% (N=42) of the patients with newly diagnosed BD and 6.9% (N=8) of their first-degree relatives had a SCAN verified anxiety diagnosis at baseline. For 93.2% of the patients, BD diagnoses were made within the preceding 2 years. Out of the 42 patients with BD and a comorbid anxiety diagnosis, 29 patients had one anxiety diagnosis and 13 patients with BD had 2–4 anxiety diagnoses at baseline. When including only participants without a mood episode, ie, in partial or full remission based on scores on the HAMD-17 and YMRS, the percentages of patients with BD with an anxiety disorder were 10.3% and 5.9%, respectively, and for first-degree relatives the percentages were 5.4% and 4.4%, respectively (table 1). There was no statistically significant difference in the prevalence of comorbid anxiety diagnosis between patients with newly diagnosed BD and their first-degree relatives neither in analysis including all participants regardless of mood episodes nor in partial or full remission (all p values>0.05).

    View this table:
    Table 1

    Prevalence of comorbid anxiety diagnosis in patients with newly diagnosed bipolar disorder (BD) and first-degree relatives at baseline

    As seen from table 2, comparing patients with and without an anxiety diagnosis, the patients with an anxiety diagnosis had an earlier onset of BD (median (IQR) age 19 (17; 22) vs age 21 (17; 27), p=0.044), more years with untreated BD (7 (2; 15) years vs 4 (1; 10), p=0.043), longer illness duration (14 (8; 19) years vs 9 (5; 15), p=0.016), more days on sick leave the previous year (days 163 (58; 308) vs 60 (15–173), p=0.009), lower functioning (p=0.001) and reported a lower level of health status (p=0.032). There was no difference between patients with and without anxiety diagnosis in relation to prescribed medication, BD type I or II, education level or civil status. When adjusting for age and sex, the difference in self-reported health status (EQ-5D, p=0.066) was no longer statistically significant (results not presented in the tables). In post hoc exploratory logistic regression analyses, we included anxiety diagnosis as a dependent variable and BD type, age at onset, illness duration, number of prior total episodes, prior psychosis and stressful life events as covariates. Longer illness duration (OR=1.097, 95% CI=1.015 to 1.186, p=0.019) and BD type II (OR=2.450, 95% CI=1.009 to 5.951, p=0.048) predicted the presence of a comorbid anxiety diagnosis.

    View this table:
    Table 2

    Differences in sociodemographic and clinical characteristics in patients with newly diagnosed bipolar disorder (BD) with and without anxiety diagnosis

    Table 3 illustrates the difference in sociodemographic and clinical characteristics between patients with and without smartphone-based daily patient-reported anxiety symptoms. In total 252 patients with BD provided self-reported data. Of these, 132 patients with BD self-reported their anxiety symptoms for a total of 19 945 days with a median number of days for each participant on 94 days (43; 178) (median(IQR)). Seven of the patients with BD rated 0 equaling to no anxiety symptoms. Therefore, these seven patients were pooled together with the patients who did not provide information about daily self-reported anxiety. Patients with self-reported anxiety symptoms had more frequently an anxiety diagnosis according to the SCAN interview, more days with sick leave the previous year (68 (20; 210) days vs 40 (10; 120) days, p=0.019), longer illness duration (11.0 (6.3; 16.0) years vs 8.0 (3.0; 13.0) years p=0.001) and were less often in remission at baseline (23.4% vs 36.2%, p=0.026) compared with patients without self-reported anxiety symptoms. When adjusting for age and sex, there was no longer a statistically significant difference in the YMRS score between the two groups (p=0.09).

    View this table:
    Table 3

    Differences in sociodemographic and clinical characteristics between patients with and without smartphone-based self-reported anxiety symptoms

    The mean level of smartphone-based self-reported anxiety symptoms for patients with BD was 0.89 (0.77; 1.02) and when only looking at days where patients had self-reported neutral mood (−0.5, 0 or 0.5, a total of 15 869 days), mean self-reported anxiety was 0.73 (0.62; 0.84). In contrast, in the HC group, the mean level of self-reported anxiety symptoms was 0.01 (0.01; 0.01) and the percentage of days with a self-reported anxiety score >0 was 0.02% which is markedly lower than in patients where it was 49.5%.

    Table 4 presents the association between smartphone-based patient-reported anxiety and clinically rated parameters of anxiety and functioning. We found that the mean anxiety level, calculated by taking the average score of daily self-reported anxiety the past 3 days, was positively associated with anxiety items on the HAMD-17 and functioning according to FAST (B=0.013, 95% CI: 0.004 to 0.023, p<0.001). A higher patient-reported anxiety score was associated with a 0.46 (−0.48; −0.45) decrease in mood score, 0.08 (−0.09; −0.07) decrease in activity score and a 0.42 (0.41; 0.43) increase in stress score, all at a statistically significant level (all p values<0.001).

    View this table:
    Table 4

    Associations between smartphone-based self-reported anxiety* and clinically rated anxiety using the Hamilton Depression Rating Scale (HAMD) and functioning according to the Functional Assessment Short Test (FAST) and smartphone-based self-reported measurements

    Discussion

    Overall, we found that the current prevalence of comorbid anxiety disorder in patients with newly diagnosed BD was 11.3% and 6.9% in first-degree relatives. Additionally, illness characteristics were more severe in patients with a SCAN verified anxiety diagnosis or with daily smartphone-based patient-reported anxiety symptoms compared with patients without an anxiety diagnosis or daily self-reported anxiety symptoms. Finally, we found that smartphone-based anxiety symptoms were associated with observer-based ratings of anxiety, patient-reported mood and stress and poorer functioning.

    Current prevalence of an anxiety diagnosis in patients with newly diagnosed BD

    The prevalence of a comorbid anxiety diagnosis, verified by SCAN interview, in patients with newly diagnosed BD was substantially lower than previously reported prevalences above 40% in the two recent meta-analyses.1 2 Further, patients with a comorbid anxiety diagnosis had statistically significantly longer illness duration, untreated BD and higher number of prior affective episodes than patients without a comorbid anxiety diagnosis. In line with prior studies,4 22 23 patients with a comorbid anxiety diagnosis reported statistically significantly more days on sick leave, lower functioning and lower self-reported health status compared with patients without a comorbid anxiety diagnosis. Overall, these findings show that the prevalence of comorbid anxiety disorders is low in newly diagnosed BD compared with other studies investigating anxiety diagnosis in later stages of BD, supporting that anxiety may increase with the progression of BD.24 Also, the findings indicate that comorbid anxiety diagnosis should be routinely assessed in patients with BD to aid full remission and functioning between episodes.

    Anxiety symptoms are often a prominent symptom during depressive episodes25 26 and may decrease with remission of episodes.27 Accordingly, we also found that the prevalence of comorbid anxiety diagnosis was lower in patients who at baseline were in partial (10.3%) or in full remission (5.9%). Further, in a meta-analysis investigating the prevalence of comorbid anxiety disorder, the prevalence of comorbid anxiety diagnosis was similarly lower in euthymic patients with BD (34.7%)28 compared with the life-time prevalence in patients with BD (45%).2 These findings stress the importance of aiming to treat patients with BD into remission as this may decrease anxiety symptoms as well.

    In contrast to a recent meta-analysis,1 we found a relatively low prevalence of generalised anxiety disorder and social phobia. This can partly be explained by that approximately one-third of the patients were not in remission when the SCAN interview was performed. Patients in an elevated mood state may likely under-report previous anxiety symptoms. Additionally, in patients with depression it can be clinically difficult to separate whether the reported anxiety symptoms are solely related to mood episodes or present between mood episodes. If the anxiety symptoms were solely related to mood episodes, it was not considered as a separate diagnosis even if the anxiety symptoms during the mood episode fulfilled the criteria for an anxiety diagnosis. This might have led to under-reporting of some of the anxiety disorders in our study since 30% of the patients were in a depressive state at the time of inclusion. Therefore, it is of interest to repeat the assessment of anxiety symptoms again in the present sample optimally when all the participants are in remission.

    Current prevalence of anxiety diagnosis in first-degree relatives

    In the first-degree relatives of patients with BD, the current prevalence of SCAN verified anxiety diagnosis was 6.9%. The prevalence is lower than what we expected for the group, this might be due to the method applied to verify anxiety diagnosis or that relatives with anxiety symptoms declined to participate in the study or already had developed a BD disorder or a psychotic disorder. Anxiety disorders and anxiety symptoms in offsprings of patients with BD are associated with increased risk of onset and earlier onset of a mood disorder.10 29–31 Early recognition of anxiety symptoms in individuals at risk seems crucial creating a possibility that both early diagnosis and prevention may be possible.32 33 Although promising results indicate that early intervention for high-risk individuals will be beneficial, further research is required over longer follow-up periods to determine who will benefit from which types of interventions. Also, systematic monitoring of high-risk individuals using clinical interviews is expensive and not feasible. It has been suggested by us34 and others35 that remote electronic monitoring of symptoms in this group of individuals may be feasible and an unobtrusive way to identify those individuals who would benefit from clinical care. Unfortunately, our present sample size was too low to investigate sociodemographic differences in this group and too few first-degree relatives provided smartphone-based anxiety self-reports.

    Smartphone-based monitoring of anxiety symptoms

    Smartphone-based symptoms monitoring is a reliable method to monitor changes in symptoms continuously and fine-grained. It gives a unique opportunity to measure the daily fluctuations in symptoms and thereby increase our understanding of the nature and heterogeneity of BD. We found that daily smartphone-based patient reports of anxiety reflected observer-based ratings of anxiety using the anxiety-items on HAMD-17. This indicates that daily patient reports of anxiety symptoms may be a useful measure of anxiety symptoms in patients with newly diagnosed BD.

    The adherence to smartphone-based daily self-reports may decline during depressive and manic episodes,36 therefore, the majority of patients reporting anxiety symptoms on smartphones will most likely be in partial or full remission. It is notable that despite being in partial or full remission most of the time, patients self-reported anxiety symptoms scored 1 or more 50% of the time. Interestingly, we also found that smartphone-based patient-reported anxiety was associated with an increased level of patient-reported stress and impaired functioning. Also, patients with self-reported anxiety symptoms had more days on sick leave, longer illness duration and fewer were in remission at inclusion compared with patients without daily self-reports of anxiety. These findings are of clinical importance and support previous findings concluding that anxiety symptoms are important to assess routinely during the clinical interview, in order to initiate treatment of anxiety symptoms and thereby potentially prevent worsening of the illness.4 8 23 37

    Strengths

    First, this is a large longitudinal observational study comprising 372 systematically recruited patients with newly diagnosed BD and 116 of their first-degree relatives. Second, the patients with BD were diagnosed at a specialised mood disorder clinic and all participants subsequently underwent a SCAN interview, by trained SCAN certified Ph.D. students with an MD or MSc psychology degree. Third, participants daily provided smartphone-based self-reports of mood, anxiety and stress—this is a unique way of monitoring anxiety symptoms daily and continuously, in real time and in naturalistic settings. Fourth, the system used in this study is a well-validated application suitable for long-term monitoring of symptoms.21 Finally, the results from this study may be generalised to all patients with newly diagnosed BD because our patients were referred from the entire Capital Region of Denmark.

    Limitations

    First, the SCAN interview was carried out at baseline regardless of the participants’ mood state. It would have been ideal to repeat parts of the SCAN interview during remission to investigate the prevalence of anxiety diagnosis at this point. Second, in table 2 when comparing clinical and sociodemographic differences between patients with and without a comorbid anxiety diagnosis, we included all anxiety diagnoses and did not differentiate between the different anxiety disorders. It is plausible that some anxiety diagnoses, such as social phobia, may impact functioning more than others. However, the sample of patients with comorbid anxiety did not allow for such subanalyses. Also, we used the SCAN interview and smartphone-based anxiety self-reports to assess anxiety. There are numerous ways to investigate the prevalence of anxiety diagnosis and symptoms in patients with BD, and it would have been interesting to include other parameters as well to verify anxiety diagnosis and to validate the smartphone-based anxiety measurement. Third, we excluded first-degree relatives with schizophrenia or other psychotic disorders or BD and therefore, the prevalence of anxiety disorder found in this study may not be representative of all first-degree relatives of patients with BD. Fourth, unfortunately, we could not include the control group in the BIO Study for this paper due to treatment requiring anxiety diagnosis being an exclusion criterion. It would have been preferable to have investigated the prevalence of anxiety diagnosis in a control group as well. Also, HC recruited among blood donors might represent a ‘super healthy’ control group, since there are strict criteria that blood donors must comply with to be eligible for blood donation.38 Fifth, it was not mandatory to self-report anxiety symptoms on the smartphone application. It is therefore a limitation that we do not know why some participants filled out self-reported anxiety and some did not. In the present study, we assume that patient reports of anxiety were provided by patients who regularly or intermittently experience anxiety and therefore, we compared clinical and sociodemographic differences between participants with and without anxiety self-reports (or anxiety self-reports at 0 at all times). Half of the time the patients who self-reported anxiety scored 0. It is probable that if we had included only participants with more severe anxiety symptoms the differences between the groups would have been larger. Also, we cannot draw any conclusions regarding adherence to daily anxiety reports, since it was solely on the participant’s initiation that anxiety symptoms were reported. Finally, we cannot draw any conclusion regarding the causality between a comorbid anxiety diagnosis and clinical and sociodemographic differences, as comorbid anxiety diagnoses were assessed at inclusion only.

    Conclusions

    Overall, the study shows that the prevalence of anxiety was lower in patients newly diagnosed with BD compared with findings from other studies investigating later stages of BD. Further, these findings stress the importance of aiming to treat patients with BD into remission as this may decrease anxiety symptoms as well. The study supports smartphones as a useful tool for remotely monitoring anxiety symptoms on a daily basis both in patients with newly diagnosed BD and in first-degree relatives. Also, the smartphone-based recordings reveal that although patients are in partial or full remission, a substantial part of patients still experience anxiety symptoms which are important to address and treat since comorbid anxiety symptoms are associated with lower functioning and worsening of illness characteristics.

    Data availability statement

    No data are available. The study is ongoing; therefore, the research data are not shared.

    Ethics statements

    Patient consent for publication

    Ethics approval

    The Bipolar Illness Onset study has been approved by the ethics committee in the Capital Region, Copenhagen, Denmark (ref. nr. H-7-2014-007) and the Danish Data Protection Agency, Capital Region of Copenhagen (protocol no.: RHP-2015–023). The study was conducted in accordance with the Declaration of Helsinki and all participants provided written informed consent.

    Acknowledgments

    The authors would like to specially thank all the participants in the study and the Copenhagen Affective Disorder Clinic, Psychiatric Center Copenhagen, Copenhagen University Hospital, Rigshospitalet, Denmark.

    References

    1. ↵
      2. Yapici Eser H ,
      3. Kacar AS ,
      4. Kilciksiz CM , et al
      . Prevalence and associated features of anxiety disorder comorbidity in bipolar disorder: a meta-analysis and meta-regression study. Front Psychiatry 2018;9:229.doi:10.3389/fpsyt.2018.00229 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29997527
      OpenUrlPubMed
    2. ↵
      2. Pavlova B ,
      3. Perlis RH ,
      4. Alda M , et al
      . Lifetime prevalence of anxiety disorders in people with bipolar disorder: a systematic review and meta-analysis. Lancet Psychiatry 2015;2:710–7.doi:10.1016/S2215-0366(15)00112-1 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26249302
      OpenUrlPubMed
    3. ↵
      2. Baxter AJ ,
      3. Scott KM ,
      4. Vos T , et al
      . Global prevalence of anxiety disorders: a systematic review and meta-regression. Psychol Med 2013;43:897–910.doi:10.1017/S003329171200147X pmid:http://www.ncbi.nlm.nih.gov/pubmed/22781489
      OpenUrlCrossRefPubMed
    4. ↵
      2. Faurholt-Jepsen M ,
      3. Frost M ,
      4. Christensen EM , et al
      . The validity of daily patient-reported anxiety measured using smartphones and the association with stress, quality of life and functioning in patients with bipolar disorder. J Affect Disord 2019;257:100–7.doi:10.1016/j.jad.2019.07.029 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31301609
      OpenUrlCrossRefPubMed
    5. ↵
      2. Provencher MD ,
      3. Hawke LD ,
      4. Thienot E
      . Psychotherapies for comorbid anxiety in bipolar spectrum disorders. J Affect Disord 2011;133:371–80.doi:10.1016/j.jad.2010.10.040 pmid:http://www.ncbi.nlm.nih.gov/pubmed/21093062
      OpenUrlCrossRefPubMed
    6. ↵
      2. Goghari VM ,
      3. Harrow M
      . Anxiety symptoms across twenty-years in schizoaffective disorder, bipolar disorder, and major depressive disorder. Psychiatry Res 2019;275:310–4.doi:10.1016/j.psychres.2019.03.050 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30953876
      OpenUrlCrossRefPubMed
    7. ↵
      2. Gaudiano BA ,
      3. Miller IW
      . Anxiety disorder comobidity in bipolar I disorder: relationship to depression severity and treatment outcome. Depress Anxiety 2005;21:71–7.doi:10.1002/da.20053 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15786484
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      2. Simon NM ,
      3. Otto MW ,
      4. Wisniewski SR , et al
      . Anxiety disorder comorbidity in bipolar disorder patients: data from the first 500 participants in the systematic treatment enhancement program for bipolar disorder (STEP-BD). Am J Psychiatry 2004;161:2222–9.doi:10.1176/appi.ajp.161.12.2222 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15569893
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      2. Goes FS ,
      3. McCusker MG ,
      4. Bienvenu OJ , et al
      . Co-morbid anxiety disorders in bipolar disorder and major depression: familial aggregation and clinical characteristics of co-morbid panic disorder, social phobia, specific phobia and obsessive-compulsive disorder. Psychol Med 2012;42:1449–59.doi:10.1017/S0033291711002637 pmid:http://www.ncbi.nlm.nih.gov/pubmed/22099954
      OpenUrlCrossRefPubMed
    10. ↵
      2. Duffy A ,
      3. Horrocks J ,
      4. Doucette S , et al
      . Childhood anxiety: an early predictor of mood disorders in offspring of bipolar parents. J Affect Disord 2013;150:363–9.doi:10.1016/j.jad.2013.04.021 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23707033
      OpenUrlCrossRefPubMed
    11. ↵
      2. Kessing LV ,
      3. Munkholm K ,
      4. Faurholt-Jepsen M , et al
      . The bipolar illness onset study: research protocol for the bio cohort study. BMJ Open 2017;7:e015462. doi:10.1136/bmjopen-2016-015462 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28645967
      OpenUrlAbstract/FREE Full Text
    12. ↵
      2. Kessing LV ,
      3. Hansen HV ,
      4. Hvenegaard A , et al
      . Treatment in a specialised out-patient mood disorder clinic V. standard out-patient treatment in the early course of bipolar disorder: randomised clinical trial. Br J Psychiatry 2013;202:212–9.doi:10.1192/bjp.bp.112.113548 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23349295
      OpenUrlAbstract/FREE Full Text
    13. ↵
      2. Faurholt-Jepsen M ,
      3. Þórarinsdóttir H ,
      4. Vinberg M , et al
      . Automatically generated smartphone data and subjective stress in healthy individuals - a pilot study. Nord J Psychiatry 2020;74:293–300.doi:10.1080/08039488.2019.1705904 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31880486
      OpenUrlCrossRefPubMed
    14. ↵
      2. WHO WHO
      . The ICD-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. World Health Organisation: Genova, 1992.
    15. ↵
      2. Wing JK ,
      3. Babor T ,
      4. Brugha T , et al
      . Scan. schedules for clinical assessment in neuropsychiatry. Arch Gen Psychiatry 1990;47:589–93.doi:10.1001/archpsyc.1990.01810180089012 pmid:http://www.ncbi.nlm.nih.gov/pubmed/2190539
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      2. Hamilton M
      . Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 1967;6:278–96.doi:10.1111/j.2044-8260.1967.tb00530.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/6080235
      OpenUrlCrossRefPubMed
    17. ↵
      2. Young RC ,
      3. Biggs JT ,
      4. Ziegler VE , et al
      . A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry 1978;133:429–35.doi:10.1192/bjp.133.5.429 pmid:http://www.ncbi.nlm.nih.gov/pubmed/728692
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Rosa AR ,
      3. Sánchez-Moreno J ,
      4. Martínez-Aran A , et al
      . Validity and reliability of the functioning assessment short test (fast) in bipolar disorder. Clin Pract Epidemiol Ment Health 2007;3:5.doi:10.1186/1745-0179-3-5 pmid:http://www.ncbi.nlm.nih.gov/pubmed/17555558
      OpenUrlCrossRefPubMed
    19. ↵
      2. Bonnín CM ,
      3. Martínez-Arán A ,
      4. Reinares M , et al
      . Thresholds for severity, remission and recovery using the functioning assessment short test (fast) in bipolar disorder. J Affect Disord 2018;240:57–62.doi:10.1016/j.jad.2018.07.045 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30053684
      OpenUrlCrossRefPubMed
    20. ↵
      Group E. Available: https://euroqol.org/home.html
    21. ↵
      2. Bardram JE ,
      3. Frost M ,
      4. Szántó K
      . Designing mobile health technology for bipolar disorder: a field trial of the monarca system. In: Proceedings of the SIGCHI conference on human factors in computing systems. ACM, 2013: :2627–36.
    22. ↵
      2. Albert U ,
      3. Rosso G ,
      4. Maina G , et al
      . Impact of anxiety disorder comorbidity on quality of life in euthymic bipolar disorder patients: differences between bipolar I and II subtypes. J Affect Disord 2008;105:297–303.doi:10.1016/j.jad.2007.05.020 pmid:http://www.ncbi.nlm.nih.gov/pubmed/17617468
      OpenUrlCrossRefPubMed
    23. ↵
      2. Kauer-Sant'Anna M ,
      3. Frey BN ,
      4. Andreazza AC , et al
      . Anxiety comorbidity and quality of life in bipolar disorder patients. Can J Psychiatry 2007;52:175–81.doi:10.1177/070674370705200309 pmid:http://www.ncbi.nlm.nih.gov/pubmed/17479526
      OpenUrlCrossRefPubMed
    24. ↵
      2. Kessing LV ,
      3. Andersen PK
      . Evidence for clinical progression of unipolar and bipolar disorders. Acta Psychiatr Scand 2017;135:51–64.doi:10.1111/acps.12667 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27858964
      OpenUrlCrossRefPubMed
    25. ↵
      2. Kessler RC ,
      3. Chiu WT ,
      4. Demler O , et al
      . Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National comorbidity survey replication. Arch Gen Psychiatry 2005;62:617–27.doi:10.1001/archpsyc.62.6.617 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15939839
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      2. Gorman JM
      . Comorbid depression and anxiety spectrum disorders. Depress Anxiety 1996;4:160–8.doi:10.1002/(SICI)1520-6394(1996)4:4&lt;160::AID-DA2&gt;3.0.CO;2-J pmid:http://www.ncbi.nlm.nih.gov/pubmed/9166648
      OpenUrlCrossRefPubMed
    27. ↵
      2. Mantere O ,
      3. Isometsä E ,
      4. Ketokivi M , et al
      . A prospective latent analyses study of psychiatric comorbidity of DSM-IV bipolar I and II disorders. Bipolar Disord 2010;12:271–84.doi:10.1111/j.1399-5618.2010.00810.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/20565434
      OpenUrlPubMed
    28. ↵
      2. Pavlova B ,
      3. Perlis RH ,
      4. Mantere O , et al
      . Prevalence of current anxiety disorders in people with bipolar disorder during euthymia: a meta-analysis. Psychol Med 2017;47:1107–15.doi:10.1017/S0033291716003135 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27995827
      OpenUrlCrossRefPubMed
    29. ↵
      2. Duffy A ,
      3. Alda M ,
      4. Crawford L ,
      5. Milin L.: , et al
      . The early manifestations of bipolar disorder: a longitudinal prospective study of the offspring of bipolar parents. Bipolar Disord 2007;9:828–38.doi:10.1111/j.1399-5618.2007.00421.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/18076532
      OpenUrlCrossRefPubMed
    30. ↵
      2. Nurnberger JI ,
      3. McInnis M ,
      4. Reich W , et al
      . A high-risk study of bipolar disorder. childhood clinical phenotypes as precursors of major mood disorders. Arch Gen Psychiatry 2011;68:1012–20.doi:10.1001/archgenpsychiatry.2011.126 pmid:http://www.ncbi.nlm.nih.gov/pubmed/21969459
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      2. Yapıcı Eser H ,
      3. Taşkıran AS ,
      4. Ertınmaz B , et al
      . Anxiety disorders comorbidity in pediatric bipolar disorder: a meta-analysis and meta-regression study. Acta Psychiatr Scand 2020;141:327–39.doi:10.1111/acps.13146 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31899546
      OpenUrlCrossRefPubMed
    32. ↵
      2. Perich T ,
      3. Mitchell PB
      . Psychological interventions for young people at risk for bipolar disorder: a systematic review. J Affect Disord 2019;252:84–91.doi:10.1016/j.jad.2019.04.058 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30981060
      OpenUrlCrossRefPubMed
    33. ↵
      2. Lambert M ,
      3. Niehaus V ,
      4. Correll C
      . Pharmacotherapy in children and adolescents at Clinical-High risk for psychosis and bipolar disorder. Pharmacopsychiatry 2016;49:229–44.doi:10.1055/s-0042-116668 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27737475
      OpenUrlCrossRefPubMed
    34. ↵
      2. Stanislaus S ,
      3. Faurholt-Jepsen M ,
      4. Vinberg M , et al
      . Mood instability in patients with newly diagnosed bipolar disorder, unaffected relatives, and healthy control individuals measured daily using smartphones. J Affect Disord 2020;271:336–44.doi:10.1016/j.jad.2020.03.049 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32479333
      OpenUrlCrossRefPubMed
    35. ↵
      2. Duffy A ,
      3. Keown-Stoneman CD ,
      4. Goodday SM , et al
      . Daily and weekly mood ratings using a remote capture method in high-risk offspring of bipolar parents: compliance and symptom monitoring. Bipolar Disord 2019;21:159–67.doi:10.1111/bdi.12721 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30422376
      OpenUrlPubMed
    36. ↵
      2. Gershon A ,
      3. Kaufmann CN ,
      4. Torous J , et al
      . Electronic ecological Momentary assessment (EMA) in youth with bipolar disorder: demographic and clinical predictors of electronic EMA adherence. J Psychiatr Res 2019;116:14–18.doi:10.1016/j.jpsychires.2019.05.026 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31176107
      OpenUrlCrossRefPubMed
    37. ↵
      2. Reinholdt-Dunne M ,
      3. Seeberg I ,
      4. Blicher A
      . Residual anxiety in patients with bipolar disorder in full or partial remission: metacognitive beliefs and neurocognitive function. Cognit Ther Res 2020:1–11.doi:10.1007/s10608-020-10148-7
    38. ↵
      2. Burgdorf KS ,
      3. Simonsen J ,
      4. Sundby A , et al
      . Socio-demographic characteristics of Danish blood donors. PLoS One 2017;12:e0169112. doi:10.1371/journal.pone.0169112 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28182624
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors LVK, RNJ and MF-J conceived the study. LVK and JEB obtained the required funding for the study and wrote the study protocol. LVK, MF-J, MV, JEB and MF were involved in optimising the study protocol. SS, KC, KSOS, HLK, IS and MV have been responsible for the recruitment of participants and have carried out the assessment and data collection. MF and SS have been responsible for data processing. Data analyses were done by SS and supervised by MF-J and LVK. Interpretation of the data has been done by SS under the supervision of LVK and MF-J. All authors have read, contributed to and approved the final version of the manuscript.

    • Funding The study was funded by grants from the Mental Health Services, Capital Region of Denmark, the Danish Council for Independent Research, Medical Sciences (DFF–4183–00570), Weimans Fund, Markedmodningsfonden (the market development fund, 2015-310), Gangstedfonden (A29594), Helsefonden (16-B-0063), Innovation Fund Denmark (the Innovation Fund, Denmark, 5164-00001B), Copenhagen Center for Health Technology, EU H2020 ITN (EU project 722561), Augustinusfonden (16-0083) and Lundbeck Foundation (R215-2015-4121).

    • Competing interests HLK, KSOS, IS, RNJ and MF-J declare no competing interests. LVK, SS and KC have within recent 3 years been a consultant for Lundbeck. MV has within the last 3 years been a consultant for Lundbeck, Sunovion and Janssen. JEB and MF are cofounders and shareholders of Monsenso A/S.

    • Provenance and peer review Not commissioned; externally peer reviewed.