Abstract
Introduction
Resting and postural heart rate (HR) is an important component of assessing physical health and haemodynamic stability in eating disorders. Postural tachycardia is defined as a HR difference of more than 20 beats per a minute (BPM) from lying/sitting to standing position and is criteria for medical admission within major hospitals (Hoo et al., 2023). Postural tachycardia is further included in the Royal Australian and New Zealand College of Psychiatrists (RANZCP) medical admission guidelines for eating disorders on the basis that it reflects illness severity and medical compromise (Hay et al., 2014). These guidelines are followed in general practice, where community medical monitoring often occurs (Rowe, 2017). Medical admission criteria include body mass index (BMI) < 12 kg/m2, systolic blood pressure (SBP) < 80 mmHg, postural blood pressure (BP) drop of >20 mmHg upon standing, HR ≤40 BPM, HR >120 BPM, postural tachycardia (>20 BPM difference), temperature <35°C, or concerning electrocardiogram (ECG) or pathology results (Hay et al., 2014).
Cardiovascular complications may present in up to 80% of patients with anorexia nervosa (AN), and at least one-third of deaths from AN are attributed to cardiovascular causes (Di Cola et al., 2014). Complications include reduced cardiac functioning, bradycardia, electrolyte abnormalities which may precipitate long QT syndrome or arrhythmias, and issues related to refeeding syndrome (McCallum et al., 2006; Sekaninova et al., 2020). AN is more often associated with bradycardia due to increased parasympathetic nervous system activation (Sachs et al., 2016). Tachycardia in AN therefore necessitates close medical monitoring and may reflect sympathetic nervous system activation and severe autonomic nervous system dysfunction, electrolyte disturbances, or precede arrhythmias and even sudden death (Kessel and Mehler, 2018; Siegel et al., 1995). Resting tachycardia in AN should also provoke suspicion of causes other than malnutrition (Krantz and Mehler, 2004) such as medication side effects, infections, or refeeding syndrome complications e.g., impending cardiac failure (Mehler et al., 2014; Mehler, 2017).
Postural tachycardia has been observed in eating disorders (Jenkins et al., 2021, 2022) and likely relates to impaired haemodynamic adaptation and baroreflex response to postural change (Sekaninova et al., 2020). Additionally, starvation is associated with atrophic peripheral muscles which can reduce venous return to the heart. Standing upright results in peripheral venous pooling which reduces pre-load and cardiac output, and this further reduces BP which subsequently stimulates carotid, aortic, and cardiopulmonary baroreceptors. This leads to increased sympathetic outflow, parasympathetic inhibition, HR, and vascular resistance. These changes maintain systemic arterial pressure during upright posture and may be exaggerated in starvation (Shamim et al., 2003). To the best of our knowledge, there has been no previous data in relation to other contributors to postural tachycardia in eating disorders, including the consideration of psychotropic medications and medical or psychiatric comorbidities.
Orthostatic hypotension may be reciprocally linked to postural tachycardia. Orthostatic hypotension involves a decreased SBP of ≥20 mmHg, decreased diastolic BP of ≥10 mmHg, or clinical symptoms of cerebral hypoperfusion (Mehler et al., 2014). Orthostatic hypotension and tachycardia are increasingly prevalent with greater AN severity (Di Cola et al., 2014). A specialized eating disorders adolescent unit study showed that postural tachycardia was more common than BP changes and that compared to orthostatic BP changes, orthostatic HR changes were more sensitive as indicators of haemodynamic instability and took greater time to resolve during nutritional rehabilitation (Shamim et al., 2003).
This study aimed to determine the following among a cohort of patients admitted to the Eating Disorders Unit (EDU) at Austin Health, a metropolitan tertiary hospital in Melbourne, Victoria, Australia: (a) the prevalence of postural tachycardia among eating disorder inpatients, and (b) the associated demographic, psychiatric, and clinical features.
Methods
Setting and participants
This was a retrospective cohort study based on electronic patient files of psychiatric inpatients who were admitted to the EDU at a metropolitan public hospital in Melbourne, Victoria, Australia. Admissions between January 01, 2021 and October 30, 2023 were included. Further inclusion criteria related to EDU admission criteria, which included patients aged 18–65 years with a psychiatric eating disorder diagnosis who resided within the Body Image & Eating Disorders Treatment & Recovery Service (BETRS) catchment area. Exclusion criteria included patients below 18 or above 65 years, residents outside the catchment area, referrals where the patient declined admission or self-discharged on the day of admission, and patients too medically unwell to be transferred to the psychiatric unit, such as severe medical instability in terms of vital signs or pathology results.
The EDU is an adult eating disorders psychiatric ward which accepts admissions through medical ward transfers or from the community through BETRS, the outpatient arm of EDU. Medical monitoring includes ECG, vital signs, pathology tests, physical examination, weights, and imaging as indicated. Malnutrition diagnosis and severity is determined by the dietitian on admission and re-assessed regularly using the Global Clinical Leadership Initiative on Malnutrition (GLIM) criteria (Cederholm et al., 2019). The GLIM criteria assign a rating of mild/moderate or severe malnutrition based on comprehensive dietetics assessment which considers weight loss percentage, BMI, oral intake, and muscle mass assessment. If the patient does not qualify for the GLIM malnutrition criteria, their nutritional diagnosis is listed as disordered eating. Malnutrition assessments inform initial meal plans at EDU. The refeeding meal plan provides the lowest caloric and carbohydrate intake to mitigate risks of refeeding syndrome and patients not deemed to be at high refeeding risks may be started on meal plan 1. Meal plans and independence are progressed based on mental state, physical stability, and weight parameters.
Emergency medical reviews (EMRs) are called when predefined criteria are met to elicit urgent medical response for unwell patients. Criteria includes obstructed airway, difficulty breathing, respiratory rate <8 or >25 breaths per/minute, oxygen saturation ≤ 90% or requiring 8 litres per/minute of oxygen, HR <40 or >110 BPM, SBP <90 mmHg, urine output <100 mL over 4 h, sudden change in conscious state, seizures, first episode chest pain, pain crisis, active bleeding, or staff concern. EMRs are attended by the psychiatric, intensive care unit, and medical emergency teams.
Design and data collection
Admission notes, progress notes, medical results, and discharge summaries for each eligible patient file were reviewed by a trained medical doctor. Postural tachycardia was defined by whether a postural tachycardia diagnosis was documented by medical practitioners involved in care and if present all vital signs were reviewed, and the diagnosis confirmed by at least two vital signs during admission which showed a HR difference of >20 BPM from sitting to standing. For patients positive for postural tachycardia, qualitative data was collected regarding potential contributors as documented in medical reviews. Medications were classified according to the primary indication of their drug class. For instance, antipsychotics were not recorded as anxiolytics although this was the primary indication in our cohort, and antidepressants were not recorded as anxiolytics although they can be used to treat anxiety disorders. Outcomes of interest included the prevalence of postural tachycardia, clinical assessments of postural tachycardia from medical reviews, and demographic and clinical factors associated with postural tachycardia. Data was extracted and recorded in deidentified format on secure servers.
Statistical methods
Descriptive statistics were used for the prevalence of demographic and clinical factors in this population. Logistics regression was used to assess for any associations between variables and the group that showed postural tachycardia versus the group which did not. Statistical analysis used first admissions only (n = 130) to account for the number of readmissions in the sample of 177 admissions. Significance level was set at P < 0.05. IBM SPSS Statistics (Version 27) was used for statistical analysis (Corp, 2020).
Clinical and demographic factors which were statistically analysed to assess for association with postural tachycardia included age, state of malnutrition, admission weight or BMI, haemoglobin, EMRs or medical transfers during admission, number of medical or psychiatric conditions, postural orthostatic tachycardia syndrome (POTS) diagnosis, and number of discharge medications, antipsychotics, or previous EDU/psychiatric (excluding EDU)/medical admissions.
Ethics
The study was approved by the Austin Health Human Research Ethics Committee (HREC), and waiver of consent was obtained (HREC/104211/Austin-2023).
Results
Demographic and clinical characteristics
There were 179 inpatient admissions at EDU between January 01, 2021 and October 30, 2023. Of these, two files were excluded as admission length was <1 day. In total, 177 inpatient admissions from a sample of 130 participants were included in the final descriptive analysis (Table 1). The average age was 29.5 ± 11.3 years, with the majority being female and Australian born. The most common diagnosis was AN – restrictive subtype and restrictive eating was the most common eating disorder behaviour. The mean eating disorder duration was 9.5 ± 10.4 years, and in most admissions, malnutrition was severe. The mean length of admission was 22.0 ± 10.8 days. 88.1% of admissions had at least one medical comorbidity, 81.4% had at least one psychiatric comorbidity, and approximately two-thirds had at least one prior medical admission related to eating disorder sequelae. Approximately half of the cohort had no prior EDU or other psychiatric admissions (Table 1).
Demographic and clinical history for inpatient eating disorder psychiatric admissions.
EMR: emergency medical review.
The mean admission weight was 47.4 ± 10.6 kg (R = 29.9–94.2 kg) and mean admission BMI was 17.4 ± 3.7 kg/m2. The mean discharge weight was 50.1 ± 10.3 kg (R = 33–96.9 kg) and mean discharge BMI was 18.4 ± 3.6 kg/m2. Mean pathology results were within normal limits however this was in the setting of inpatient supplementation. The most common discharge medications included vitamin/mineral supplementation, antidepressants, antipsychotics, and anxiolytics (Table 2). The most common antidepressants included selective serotonin reuptake inhibitors, and the most common antipsychotics included olanzapine and quetiapine, which were used to manage anxiety and distress. The most common anxiolytics were benzodiazepines.
Medications at discharge for inpatient eating disorder psychiatric admissions.
Postural tachycardia was documented in 56 (31.6%) of 177 admissions. On average, patients with postural tachycardia compared to patients without postural tachycardia were younger, had a higher mean weight/BMI, were more likely to be prescribed an antipsychotic/s on discharge, and to be prescribed more antipsychotics overall (Table 3). Mean sitting/standing HR and postural difference were higher in the group with postural tachycardia however mean sitting/standing BP was within normal limits with minimal postural change for both groups (Table 4).
Clinical characteristics of eating disorder admissions with postural tachycardia compared to admissions without postural tachycardia.
BMI: body mass index. *p < 0.05.
Heart rate and blood pressure on admission and discharge for eating disorder admissions.
BPM: beats per a minute.
In terms of qualitative data from medical reviews during the inpatient setting, postural tachycardia was frequently described as multifactorial as assessed by psychiatric or general medical doctors involved in patient care. Among 48 patients and 56 admissions positive for postural tachycardia, it was documented for 42 patients and 47 admissions that postural tachycardia was at least partially attributable to malnutrition. It was documented among 11 patients and 16 admissions that POTS was potentially a contributor however in many patients this diagnosis had not been confirmed by investigations and was primarily based on clinical history and examination. Medications were documented as a potential cause in five admissions including olanzapine, quetiapine, dothiepin, or zuclopenthixol. Possible contributors which were less common included dehydration, cardiac history, prolonged bed rest, or viral illness.
Statistical analysis
Antipsychotic medication was significantly positively associated (p = 0.029) (CI: 1.066–3.352) with postural tachycardia. For every increase in one antipsychotic medication, there was an 89% increase in the likelihood of a postural tachycardia diagnosis (B = 0.637) (Exp(‘B’)) = 1.890). Admission BMI was also significantly positively associated with postural tachycardia (p = 0.009) (CI: 1.034–1.261). For each increase in a BMI value of 1.0 kg/m2, there was a 14.2% increase in the likelihood of a postural tachycardia diagnosis (B = 0.132) (Exp(‘B’)) = 1.142). Similarly, admission weight was significantly positively associated with postural tachycardia (p = 0.014) (CI: 1.009–1.079) (B = 0.042) (Exp(‘B’)) = 1.043). EMR(s) during admission (p = 0.007) (CI: 1.496–13.312) (B = 1.496) (Exp(‘B’)) = 4.462), a history of 1–3 previous medical admissions in relation to eating disorder symptoms (p = 0.018) (CI: 1.194 −6.451) (B = 1.021) (Exp{‘B’)) = 2.776), and a history of ≥10 psychiatric admissions (p = 0.025) (CI: 1.207–17.029) (B = 1.512) (Exp(‘B’)) = 4.534) were also significantly positively associated with a postural tachycardia diagnosis.
Demographic and clinical factors that were not significantly associated with a postural tachycardia diagnosis in this study included age, malnutrition severity, number of active medical conditions, POTS diagnosis, Hb level, iron level, ferritin level, number of psychiatric comorbidities, medical transfers during admission, number of discharge medications, less than 10 psychiatric admissions, previous EDU admissions, or previous medical admissions.
Discussion
In this cohort of 130 psychiatric inpatients and 177 inpatient admissions, postural tachycardia was documented in 56 (31.6%) of admissions and on average, these patients were younger with a higher weight and BMI, and more likely to be prescribed an antipsychotic, as well as multiple antipsychotics. Antipsychotic medication, increased weight/BMI, EMR(s), a history of 1–3 previous medical admissions due to eating disorder complications, and a history of ≥10 psychiatric admissions were significantly positively associated with a postural tachycardia diagnosis. The documented impression from medical reviews for most admissions with postural tachycardia was that malnutrition was a primary contributor, despite the group without postural tachycardia showing greater prevalence of a severe malnutrition diagnosis as per dietitian reviews.
Although the most frequently documented impressions from medical professionals suggested malnutrition was the primary cause of postural tachycardia during admissions, our findings show that severe malnutrition as assessed by dietitian reviews was more prevalent in patients without postural tachycardia. This occurred in the setting of a higher mean weight and BMI among the group who had postural tachycardia. This finding suggests that malnutrition severity may be underestimated in patients with a higher weight and BMI, such as in atypical anorexia nervosa (AAN), or that recent percentage of body weight loss may be an important indicator in malnutrition assessments. Our results indicate that despite this, significant medical instability and clinical indications for EMRs can occur at higher weights and BMIs. Malnutrition assessment should include comprehensive evaluation of vital signs, pathology results, weight trajectory, and physical symptoms, instead of relying on body weight. Ongoing investigation is needed in relation to whether postural tachycardia may be more prevalent in AAN or recent rapid weight loss presentations, and the implications for nutritional status and medical stability.
Higher weight and BMI are common antipsychotic side effects (Bak et al., 2014), and our findings also suggest that postural tachycardia in eating disorders may have underlying contributors that are independent of malnutrition, such as medications. Medications were only documented as a potential cause during medical reviews in five admissions. However, antipsychotics were positively associated with postural tachycardia among inpatients with eating disorders in this study. Antipsychotics have been strongly associated with severe postural tachycardia and EMRs among psychiatric inpatients previously (Lim et al., 2021) and can cause tachycardia through anticholinergic or α1-adrenoceptor blockade, in addition to complications such as QT prolongation and polymorphic ventricular tachycardia (Buckley and Sanders, 2000). This association with antipsychotics may also explain our finding that a history of ≥10 previous psychiatric admissions was associated with postural tachycardia. Psychotropic medications are commonly used in eating disorders (Fazeli et al., 2012), and antipsychotics may be used to encourage weight gain and reduce obsessive-compulsive symptoms (Sánchez et al., 2016) or as anxiolytics (Depping et al., 2010). Our findings would suggest that prescribing should be undertaken with caution in this population who are sensitized to cardiovascular abnormalities. Non-pharmacological strategies for anxiety management should be prioritized where possible. If antipsychotics or other psychotropics are prescribed, starting at lower doses with careful monitoring and cautious uptitration as clinically indicated may be necessary.
Factors other than malnutrition, such as medications, should be considered when assessing postural tachycardia among eating disorder patients, particularly if other medical parameters are within normal limits. Although postural tachycardia was associated with EMRs and previous medical admissions, in our study, there were only modest improvements in mean postural tachycardia at discharge from the inpatient setting. Postural tachycardia in eating disorders may persist after weight restoration and often improves over time with ongoing nutrition (McCallum et al., 2006) or may relate to external factors which are independent from malnutrition. Further longitudinal studies are needed regarding the trajectory and duration of postural tachycardia following discharge from acute inpatient refeeding treatment.
The notable prevalence of postural tachycardia with a normal mean BP in our sample further suggests that external factors may have contributed to postural tachycardia. While POTS is a common comorbidity with AN, distinguishing POTS from medical sequelae of AN is challenging as dysautonomia presents in both POTS and malnutrition (Mehler, 2017; Sachs et al., 2016). POTS was not associated with postural tachycardia in this study. In our methods, whether a patient had POTS was based on provisional clinical impressions documented in reviews by medical doctors. However, most documented diagnoses were not confirmed by investigations such as tilt-table testing and instead often relied on clinical assessment and judgment. Our findings reinforce the notion that clinical impressions may be vulnerable to subjectivity and the need to exclude other differentials including reflex tachycardia, involve specialty input including cardiology services, and use objective measures such as tilt-table testing prior to POTS diagnosis or treatment. Orthostatic intolerance may resolve months after weight restoration, and therefore POTS assessment is often recommended after malnutrition resolution if symptoms or signs persist (Mehler et al., 2014; Mehler, 2017). Abnormal vital signs, dizziness and lightheadedness may alert patients to the ‘warning signs’ of malnutrition, and dismissing or misattributing these symptoms may further reinforce underestimating the impact of eating disorders.
Strengths and limitations
To our knowledge, this paper is the first to describe the prevalence of postural tachycardia and its demographic and clinical associations in an inpatient eating disorder unit population. Limitations included the retrospective nature of the study, including the possibility of missing documentation and risk of biases. Further limitations include the generalizability of our findings, which would benefit from replication in more heterogeneous cohorts, including males. AN was the primary diagnosis in the majority, and results may not be generalizable to other eating disorder presentations such as bulimia nervosa. Our statistical analysis was also limited by the cohort percentage with postural tachycardia in addition to readmissions that were excluded to prevent resampling bias, which impacted statistical power. Future studies would benefit from pooled data from multiple hospital services.
Participants were allocated to the group with postural tachycardia versus without postural tachycardia based on whether a diagnosis was documented in their medical notes, and this is vulnerable to underreporting. Additionally, because only the first and last vital signs for each admission were included, this may have impacted our data collection. A retrospective analysis of all sitting and standing HR readings in the group with postural tachycardia showed that the total average HR and postural differences in this group were similar to the results in our study design. However, calculating the mean did not represent the incidence of postural tachycardia and progression over the course of the admission. Finally, as outpatient follow-up was not undertaken, it cannot be confirmed when postural tachycardia would have resolved with nutritional restoration in these patients.
Conclusion
Postural tachycardia is present in almost one-third of this inpatient eating disorder population. Our study found associations between postural tachycardia and antipsychotic medication, higher BMI and weight, EMRs, a history of 1–3 previous medical admissions related to malnutrition, and a history of ≥10 psychiatric admissions. Although lower BMI is regarded as a marker of eating disorder severity, AAN may present more often with postural tachycardia. Our findings also suggest that medical reviews should include an awareness of other potential contributors to vital sign abnormalities in eating disorder presentations, including psychotropic medications. Cautious and evidence-based prescribing is required in this medically vulnerable population who are sensitized to autonomic dysfunction.
Footnotes
Author contributions
Natalie Seiler was involved in conceptualization, data curation, and manuscript writing and revision. Benjamin Ziegeler led the formal analysis and investigation. Natalie Sacco was involved in data curation and manuscript review, and Andrew Teh and Karen Gwee led the conceptualization, manuscript review, and supervision
Consent for publication
Not applicable.
Data availability statement
The data that support the findings of this study are available from the corresponding author, Natalie Seiler, upon reasonable request.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical approval
The study was approved by the Austin Health Human Research Ethics Committee (HREC), and waiver of consent was obtained (HREC/104211/Austin-2023).
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
