Several diagnostic tests are often adopted into diagnostic pathways for specific indications without strong evidence to support their use. In this context, real-world prospective cohort studies in combination with decision modelling can generate evidence to support decision-making. The Early Detection of neovascular Age-Related Macular Degeneration (EDNA) study was a prospective cohort designed to assess the diagnostic accuracy and cost-effectiveness of several diagnostic monitoring tests used in routine practice for the detection of neovascular age-related macular degeneration (nAMD) in the second eye of patients being treated for unilateral disease.

Five-hundred and fifty-two participants with newly diagnosed unilateral nAMD were monitored for up to 3 years in 24 UK eye clinics. The diagnostic monitoring performance of five index tests was compared: self-reported change in visual function, Amsler test, clinic measured change in visual acuity, fundus assessment by clinical examination or colour photography, and spectral-domain optical coherence tomography (SD-OCT). The reference standard was fundus fluorescein angiography (FFA). A patient-level state transition model was used to simulate the onset of nAMD in the second eye, and assess the impact of different tests on the timing of detection and treatment, and associated costs and quality adjusted life years (QALYs) over a 25-year time-horizon.

One hundred and forty-five (26.3%) patients developed active nAMD in the study eye, of whom 120 had an FFA at detection. SD-OCT had the highest sensitivity (91.7 percent (95% CI: 85.2-95.6) and provided high specificity (87.8% (95% CI: 83.8-90.9)). It generated more QALYs and lower health and personal social care costs compared to all other monitoring tests. The combination of SD-OCT with fundus-examination provided a marginal increase in sensitivity over OCT alone, but the associated incremental cost-effectiveness ratios was >GBP 100,000 per QALY.

The efficiency of diagnostic pathways for nAMD may be improved by using SD-OCT alone to monitor the second eye of people being treated for unilateral disease. Prospective cohort studies embedded into routine practice offer value for informing decisions surrounding the use of technologies already in routine use.

Data suggest poorer bereavement outcomes for lesbian, gay and bisexual people, but this has not been estimated in population-based research. This study compared bereavement outcomes for partners of same-gender and different-gender decedents.

In this population-based, cross-sectional survey of people bereaved of a civil partner or spouse 6–10 months previously, we used adjusted logistic and linear regression to investigate outcomes of interest: (1) positive screen on Inventory of Complicated Grief (ICG), (2) positive screen on General Health Questionnaire (GHQ), (3) grief intensity (ICG) and (4) psychiatric symptoms (GHQ-12).

Among 233 same-gender partners and 329 of different-gender partners, 66.1% [95% confidence interval (CI) 60.0–72.2] and 59.2% [95% CI (53.9–64.6)] respectively screened positive for complicated grief on the ICG, whilst 76.0% [95% CI (70.5–81.5)] and 69.3% [95% CI (64.3–74.3)] respectively screened positive on the GHQ-12. Same-gender bereaved partners were not significantly more likely to screen positive for complicated grief than different-gender partners [adjusted odds ratio (aOR) 1.56, 95% CI (0.98–2.47)], p = 0.059, but same-gender bereaved partners were significantly more likely to screen for psychiatric caseness [aOR 1.67 (1.02, 2.71) p = 0.043]. We similarly found no significant association of partner gender with grief intensity [B = 1.86, 95% CI (−0.91to 4.63), p = 0.188], but significantly greater psychological distress for same-gender partners [B = 1.54, 95% CI (−0.69–2.40), p < 0.001].

Same-gender bereaved partners report significantly more psychological distress. In view of their poorer sub-clinical mental health, clinical and bereavement services should refine screening processes to identify those at risk of poor mental health outcomes.

Clarifying the relationship between depression symptoms and cardiometabolic and related health could clarify risk factors and treatment targets. The objective of this study was to assess whether depression symptoms in midlife are associated with the subsequent onset of cardiometabolic health problems.

The study sample comprised 787 male twin veterans with polygenic risk score data who participated in the Harvard Twin Study of Substance Abuse (‘baseline’) and the longitudinal Vietnam Era Twin Study of Aging (‘follow-up’). Depression symptoms were assessed at baseline [mean age 41.42 years (s.d. = 2.34)] using the Diagnostic Interview Schedule, Version III, Revised. The onset of eight cardiometabolic conditions (atrial fibrillation, diabetes, erectile dysfunction, hypercholesterolemia, hypertension, myocardial infarction, sleep apnea, and stroke) was assessed via self-reported doctor diagnosis at follow-up [mean age 67.59 years (s.d. = 2.41)].

Total depression symptoms were longitudinally associated with incident diabetes (OR 1.29, 95% CI 1.07–1.57), erectile dysfunction (OR 1.32, 95% CI 1.10–1.59), hypercholesterolemia (OR 1.26, 95% CI 1.04–1.53), and sleep apnea (OR 1.40, 95% CI 1.13–1.74) over 27 years after controlling for age, alcohol consumption, smoking, body mass index, C-reactive protein, and polygenic risk for specific health conditions. In sensitivity analyses that excluded somatic depression symptoms, only the association with sleep apnea remained significant (OR 1.32, 95% CI 1.09–1.60).

A history of depression symptoms by early midlife is associated with an elevated risk for subsequent development of several self-reported health conditions. When isolated, non-somatic depression symptoms are associated with incident self-reported sleep apnea. Depression symptom history may be a predictor or marker of cardiometabolic risk over decades.

Despite evidence of gender differences in bipolar disorder characteristics and comorbidity, there is little research on the differences in treatment and service use between men and women with bipolar disorder.

To use routine data to describe specialist mental health service contact for bipolar disorder, including in-patient, community and support service contacts; to compare clinical characteristics and mental health service use between men and women in contact with secondary services for bipolar disorder.

Cross-sectional analysis of mental health patients with bipolar disorder in New Zealand, based on complete national routine health data.

A total of 3639 individuals were in contact with specialist mental health services with a current diagnosis of bipolar disorder in 2015. Of these 58% were women and 46% were aged 45 and over. The 1-year prevalence rate of bipolar disorder leading to contact with specialist mental health services was 1.56 (95% CI 1.50–1.63) per 100 000 women and 1.20 (95% CI 1.14–1.26) per 100 000 men. Rates of bipolar disorder leading to service contact were 30% higher in women than men (rate ratio 1.30, 95% CI 1.22–1.39). The majority (68%) had a diagnosis of bipolar I disorder. Women were more likely to receive only out-patient treatment and have comorbid anxiety whereas more men had substance use disorder, were convicted for crimes when unwell, received compulsory treatment orders and received in-patient treatment.

Although the prevalence of bipolar disorder is equal between men and women in the population, women were more likely to have contact with specialist services for bipolar disorder but had a lower intensity of service interaction.

Heavy alcohol consumption is associated with poorer cognitive function in older adults. Although understudied in middle-aged adults, the relationship between alcohol and cognition may also be influenced by genetics such as the apolipoprotein (ApoE) ε4 allele, a risk factor for Alzheimer’s disease. We examined the relationship between alcohol consumption, ApoE genotype, and cognition in middle-aged adults and hypothesized that light and/or moderate drinkers (≤2 drinks per day) would show better cognitive performance than heavy drinkers or non-drinkers. Additionally, we hypothesized that the association between alcohol use and cognitive function would differ by ApoE genotype (ε4+ vs. ε4−).

Participants were 1266 men from the Vietnam Era Twin Study of Aging (VETSA; M age = 56; range 51–60) who completed a neuropsychological battery assessing seven cognitive abilities: general cognitive ability (GCA), episodic memory, processing speed, executive function, abstract reasoning, verbal fluency, and visuospatial ability. Alcohol consumption was categorized into five groups: never, former, light, moderate, and heavy.

In fully adjusted models, there was no significant main effect of alcohol consumption on cognitive functions. However, there was a significant interaction between alcohol consumption and ApoE ε4 status for GCA and episodic memory, such that the relationship of alcohol consumption and cognition was stronger in ε4 carriers. The ε4+ heavy drinking subgroup had the poorest GCA and episodic memory.

Presence of the ε4 allele may increase vulnerability to the deleterious effects of heavy alcohol consumption. Beneficial effects of light or moderate alcohol consumption were not observed.

Optimal stroke care requires access to resources such as neuroimaging, acute revascularization, rehabilitation, and stroke prevention services, which may not be available in rural areas. We aimed to determine geographic access to stroke care for residents of rural communities in the province of Ontario, Canada.

We used the Ontario Road Network File database linked with the 2016 Ontario Acute Stroke Care Resource Inventory to estimate the proportion of people in rural communities, defined as those with a population size <10,000, who were within 30, 60, and 240 minutes of travel time by car from stroke care services, including brain imaging, thrombolysis treatment centers, stroke units, stroke prevention clinics, inpatient rehabilitation facilities, and endovascular treatment centers.

Of the 1,496,262 people residing in rural communities, the majority resided within 60 minutes of driving time to a center with computed tomography (85%), thrombolysis (81%), a stroke unit (68%), a stroke prevention clinic (74%), or inpatient rehabilitation (77.0%), but a much lower proportion (32%) were within 60 minutes of driving time to a center capable of providing endovascular thrombectomy (EVT).

Most rural Ontario residents have appropriate geographic access to stroke services, with the exception of EVT. This information may be useful for jurisdictions seeking to optimize the regional organization of stroke care services.

Studies involving clinically recruited samples show that genetic liability to schizophrenia overlaps with that for several psychiatric disorders including bipolar disorder, major depression and, in a population study, anxiety disorder and negative symptoms in adolescence.

We examined whether, at a population level, association between schizophrenia liability and anxiety disorders continues into adulthood, for specific anxiety disorders and as a group. We explored in an epidemiologically based cohort the nature of adult psychopathology sharing liability to schizophrenia.

Schizophrenia polygenic risk scores (PRSs) were calculated for 590 European-descent individuals from the Christchurch Health and Development Study. Logistic regression was used to examine associations between schizophrenia PRS and four anxiety disorders (social phobia, specific phobia, panic disorder and generalised anxiety disorder), schizophrenia/schizophreniform disorder, manic/hypomanic episode, alcohol dependence, major depression, and – using linear regression – total number of anxiety disorders. A novel population-level association with hypomania was tested in a UK birth cohort (Avon Longitudinal Study of Parents and Children).

Schizophrenia PRS was associated with total number of anxiety disorders and with generalised anxiety disorder and panic disorder. We show a novel population-level association between schizophrenia PRS and manic/hypomanic episode.

The relationship between schizophrenia liability and anxiety disorders is not restricted to psychopathology in adolescence but is present in adulthood and specifically linked to generalised anxiety disorder and panic disorder. We suggest that the association between schizophrenia liability and hypomanic/manic episodes found in clinical samples may not be due to bias.

None.

The field of psychiatry would benefit significantly from developing objective biomarkers that could facilitate the early identification of heterogeneous subtypes of illness. Critically, although machine learning pattern recognition methods have been applied recently to predict many psychiatric disorders, these techniques have not been utilized to predict subtypes of posttraumatic stress disorder (PTSD), including the dissociative subtype of PTSD (PTSD + DS).

Using Multiclass Gaussian Process Classification within PRoNTo, we examined the classification accuracy of: (i) the mean amplitude of low-frequency fluctuations (mALFF; reflecting spontaneous neural activity during rest); and (ii) seed-based amygdala complex functional connectivity within 181 participants [PTSD (n = 81); PTSD + DS (n = 49); and age-matched healthy trauma-unexposed controls (n = 51)]. We also computed mass-univariate analyses in order to observe regional group differences [false-discovery-rate (FDR)-cluster corrected p < 0.05, k = 20].

We found that extracted features could predict accurately the classification of PTSD, PTSD + DS, and healthy controls, using both resting-state mALFF (91.63% balanced accuracy, p < 0.001) and amygdala complex connectivity maps (85.00% balanced accuracy, p < 0.001). These results were replicated using independent machine learning algorithms/cross-validation procedures. Moreover, areas weighted as being most important for group classification also displayed significant group differences at the univariate level. Here, whereas the PTSD + DS group displayed increased activation within emotion regulation regions, the PTSD group showed increased activation within the amygdala, globus pallidus, and motor/somatosensory regions.

The current study has significant implications for advancing machine learning applications within the field of psychiatry, as well as for developing objective biomarkers indicative of diagnostic heterogeneity.