Relative risks of cardiovascular disease in people prescribed olanzapine, risperidone and quetiapine

Antipsychotics may confer long term benefits and risks, including cardiovascular disease (CVD) risk. Several studies using routine clinical data have reported associations between antipsychotics and CVD but potential confounding factors and unclear classification of drug exposure limits their interpretation.

METHOD

We used data from The Health Improvement Network, a large UK primary care database to determine relative risks of (CVD) comparing similar groups of people only prescribed olanzapine versus either risperidone or quetiapine. We included participants over 18 between 1995 and 2011. To assess confounding factors we created propensity scores for being prescribed each antipsychotic. We used propensity score matching and Poisson regression to calculate the CVD incidence rate ratios for olanzapine versus the other two drugs.

RESULTS

We identified 18,319 people who received a single antipsychotic during follow-up (n=5090 risperidone, 7797 olanzapine and 4613 quetiapine). In unmatched analyses, the CVD incidence rate ratio (IRR) for olanzapine versus risperidone was 0.63 (0.51-0.77) but the propensity score matched IRR was 0.78 (0.61-1.02). In the unmatched olanzapine versus quetiapine analysis the IRR adjusted for age and sex for olanzapine was 1.52 (1.16-1.98) but the propensity score matched analysis gave an IRR of 1.08 (0.79-1.46).

CONCLUSIONS

After propensity score matching, we found no statistical differences in CVD incidence between olanzapine and either risperidone or quetiapine. Analyses which did not account for confounding factors produced very different results. Researchers must address confounding factors when designing observational studies to assess adverse outcomes of drugs, including antipsychotics.

Statin prescribing for people with severe mental illnesses: a staggered cohort study of 'real-world' impacts

OBJECTIVES

To estimate the 'real-world effectiveness of statins for primary prevention of cardiovascular disease (CVD) and for lipid modification in people with severe mental illnesses (SMI), including schizophrenia and bipolar disorder.

DESIGN

Series of staggered cohorts. We estimated the effect of statin prescribing on CVD outcomes using a multivariable Poisson regression model or linear regression for cholesterol outcomes.

SETTING

587 general practice (GP) surgeries across the UK reporting data to The Health Improvement Network.

PARTICIPANTS

All permanently registered GP patients aged 40-84 years between 2002 and 2012 who had a diagnosis of SMI. Exclusion criteria were pre-existing CVD, statin-contraindicating conditions or a statin prescription within the 24 months prior to the study start.

EXPOSURE

One or more statin prescriptions during a 24-month 'baseline' period (vs no statin prescription during the same period).

MAIN OUTCOME MEASURES

The primary outcome was combined first myocardial infarction and stroke. All-cause mortality and total cholesterol concentration were secondary outcomes.

RESULTS

We identified 2944 statin users and 42 886 statin non-users across the staggered cohorts. Statin prescribing was not associated with significant reduction in CVD events (incident rate ratio 0.89; 95% CI 0.68 to 1.15) or all-cause mortality (0.89; 95% CI 0.78 to 1.02). Statin prescribing was, however, associated with statistically significant reductions in total cholesterol of 1.2 mmol/L (95% CI 1.1 to 1.3) for up to 2 years after adjusting for differences in baseline characteristics. On average, total cholesterol decreased from 6.3 to 4.6 in statin users and 5.4 to 5.3 mmol/L in non-users.

CONCLUSIONS

We found that statin prescribing to people with SMI in UK primary care was effective for lipid modification but not CVD events. The latter finding may reflect insufficient power to detect a smaller effect size than that observed in randomised controlled trials of statins in people without SMI.

Incorporating Functional Genomic Information in Genetic Association Studies Using an Empirical Bayes Approach

There is a large amount of functional genetic data available, which can be used to inform fine-mapping association studies (in diseases with well-characterised disease pathways). Single nucleotide polymorphism (SNP) prioritization via Bayes factors is attractive because prior information can inform the effect size or the prior probability of causal association. This approach requires the specification of the effect size. If the information needed to estimate a priori the probability density for the effect sizes for causal SNPs in a genomic region isn't consistent or isn't available, then specifying a prior variance for the effect sizes is challenging. We propose both an empirical method to estimate this prior variance, and a coherent approach to using SNP-level functional data, to inform the prior probability of causal association. Through simulation we show that when ranking SNPs by our empirical Bayes factor in a fine-mapping study, the causal SNP rank is generally as high or higher than the rank using Bayes factors with other plausible values of the prior variance. Importantly, we also show that assigning SNP-specific prior probabilities of association based on expert prior functional knowledge of the disease mechanism can lead to improved causal SNPs ranks compared to ranking with identical prior probabilities of association. We demonstrate the use of our methods by applying the methods to the fine mapping of the CASP8 region of chromosome 2 using genotype data from the Collaborative Oncological Gene-Environment Study (COGS) Consortium. The data we analysed included approximately 46,000 breast cancer case and 43,000 healthy control samples.

Effectiveness and cost-effectiveness of a cardiovascular risk prediction algorithm for people with severe mental illness

Introduction

Cardiovascular risk prediction tools are important for cardiovascular disease (CVD) prevention, however, which algorithms are appropriate for people with severe mental illness (SMI) is unclear.

Objectives/aims

To determine the cost-effectiveness using the net monetary benefit (NMB) approach of two bespoke SMI-specific risk algorithms compared to standard risk algorithms for primary CVD prevention in those with SMI, from an NHS perspective.

Methods

A microsimulation model was populated with 1000 individuals with SMI from The Health Improvement Network Database, aged 30–74 years without CVD. Four cardiovascular risk algorithms were assessed; (1) general population lipid, (2) general population BMI, (3) SMI-specific lipid and (4) SMI-specific BMI, compared against no algorithm. At baseline, each cardiovascular risk algorithm was applied and those high-risk (> 10%) were assumed to be prescribed statin therapy, others received usual care. Individuals entered the model in a ‘healthy’ free of CVD health state and with each year could retain their current health state, have cardiovascular events (non-fatal/fatal) or die from other causes according to transition probabilities.

Results

The SMI-specific BMI and general population lipid algorithms had the highest NMB of the four algorithms resulting in 12 additional QALYs and a cost saving of approximately £37,000 (US$ 58,000) per 1000 patients with SMI over 10 years.

Conclusions

The general population lipid and SMI-specific BMI algorithms performed equally well. The ease and acceptability of use of a SMI-specific BMI algorithm (blood tests not required) makes it an attractive algorithm to implement in clinical settings.

Disclosure of interest

The authors have not supplied their declaration of competing interest.

Predicting dementia risk in primary care: development and validation of the Dementia Risk Score using routinely collected data

BACKGROUND

Existing dementia risk scores require collection of additional data from patients, limiting their use in practice. Routinely collected healthcare data have the potential to assess dementia risk without the need to collect further information. Our objective was to develop and validate a 5-year dementia risk score derived from primary healthcare data.

METHODS

We used data from general practices in The Health Improvement Network (THIN) database from across the UK, randomly selecting 377 practices for a development cohort and identifying 930,395 patients aged 60-95 years without a recording of dementia, cognitive impairment or memory symptoms at baseline. We developed risk algorithm models for two age groups (60-79 and 80-95 years). An external validation was conducted by validating the model on a separate cohort of 264,224 patients from 95 randomly chosen THIN practices that did not contribute to the development cohort. Our main outcome was 5-year risk of first recorded dementia diagnosis. Potential predictors included sociodemographic, cardiovascular, lifestyle and mental health variables.

RESULTS

Dementia incidence was 1.88 (95% CI, 1.83-1.93) and 16.53 (95% CI, 16.15-16.92) per 1000 PYAR for those aged 60-79 (n = 6017) and 80-95 years (n = 7104), respectively. Predictors for those aged 60-79 included age, sex, social deprivation, smoking, BMI, heavy alcohol use, anti-hypertensive drugs, diabetes, stroke/TIA, atrial fibrillation, aspirin, depression. The discrimination and calibration of the risk algorithm were good for the 60-79 years model; D statistic 2.03 (95% CI, 1.95-2.11), C index 0.84 (95% CI, 0.81-0.87), and calibration slope 0.98 (95% CI, 0.93-1.02). The algorithm had a high negative predictive value, but lower positive predictive value at most risk thresholds. Discrimination and calibration were poor for the 80-95 years model.

CONCLUSIONS

Routinely collected data predicts 5-year risk of recorded diagnosis of dementia for those aged 60-79, but not those aged 80+. This algorithm can identify higher risk populations for dementia in primary care. The risk score has a high negative predictive value and may be most helpful in 'ruling out' those at very low risk from further testing or intensive preventative activities.

Severity of depressive symptoms as a predictor of mortality: the English longitudinal study of ageing

BACKGROUND

Major depressive disorder and subthreshold depression have been associated with premature mortality. We investigated the association between depressive symptoms and mortality across the full continuum of severity.

METHOD

We used Cox proportional hazards models to examine the association between depressive symptom severity, assessed using the eight-item Center for Epidemiological Studies Depression Scale (CES-D; range 0-8), and the risk of all-cause mortality over a 9-year follow-up, in 11 104 members of the English Longitudinal Study of Ageing.

RESULTS

During follow-up, one fifth of study members died (N = 2267). Depressive symptoms were associated with increased mortality across the full range of severity (p trend < 0.001). Relative to study members with no symptoms, an increased risk of mortality was found in people with depressive symptoms of a low [hazard ratio (HR) for a score of 2 was 1.59, 95% confidence interval (CI) 1.40-1.82], moderate (score of 4: HR 1.80, 95% CI 1.52-2.13) and high (score of 8: HR 2.27, 95% CI 1.69-3.04) severity, suggesting risk emerges at low levels but plateaus thereafter. A third of participants (36.4%, 95% CI 35.5-37.3) reported depressive symptoms associated with an increased mortality risk. Adjustment for physical activity, physical illnesses, and impairments in physical and cognitive functioning attenuated this association (p trend = 0.25).

CONCLUSIONS

Depressive symptoms are associated with an increased mortality risk even at low levels of symptom severity. This association is explained by physical activity, physical illnesses, and impairments in physical and cognitive functioning.

Socio-demographic characteristics, lifestyle factors and burden of morbidity associated with self-reported hearing and vision impairments in older British community-dwelling men: a cross-sectional study

BACKGROUND

Hearing and vision problems are common in older adults. We investigated the association of self-reported sensory impairment with lifestyle factors, chronic conditions, physical functioning, quality of life and social interaction.

METHODS

A population-based cross-sectional study of participants of the British Regional Heart Study aged 63-85 years.

RESULTS

A total of 3981 men (82% response rate) provided data. Twenty-seven per cent (n = 1074) reported hearing impairment including being able to hear with aid (n = 482), being unable to hear (no aid) (n = 424) and being unable to hear despite aid (n = 168). Three per cent (n = 124) reported vision impairment. Not being able to hear, irrespective of use of hearing aid, was associated with poor quality of life, poor social interaction and poor physical functioning. Men who could not hear despite hearing aid were more likely to report coronary heart disease (CHD) [age-adjusted odds ratios (ORs) 1.89 (95% confidence interval 1.36-2.63)]. Vision impairment was associated with symptoms of CHD including breathlessness [OR 2.06 (1.38-3.06)] and chest pain [OR 1.58 (1.07-2.35)]. Vision impairment was also associated with poor quality of life, poor social interaction and poor physical functioning.

CONCLUSIONS

Sensory impairment is associated with poor physical functioning, poor health and poor social interaction in older men. Further research is warranted on pathways underlying these associations.

A systematic review and meta-analysis of premature mortality in bipolar affective disorder

OBJECTIVE

To review and complete meta-analysis of studies estimating standardised mortality ratios (SMRs) in bipolar affective disorder (BPAD) for all-cause and cause-specific mortalities.

METHOD

Cause-specific mortality was grouped into natural and unnatural causes. These subgroups were further divided into circulatory, respiratory, neoplastic and infectious causes, and suicide and other violent deaths. Summary SMRs were calculated using random-effects meta-analysis. Heterogeneity was examined via subgroup analysis and meta-regression.

RESULTS

Systematic searching found 31 studies meeting inclusion criteria. Summary SMR for all-cause mortality = 2.05 (95% CI 1.89-2.23), but heterogeneity was high (I(2) = 96.2%). This heterogeneity could not be accounted for by date of publication, cohort size, mid-decade of data collection, population type or geographical region. Unnatural death summary SMR = 7.42 (95% CI 6.43-8.55) and natural death = 1.64 (95% CI 1.47-1.83). Specifically, suicide SMR = 14.44 (95% CI 12.43-16.78), other violent death SMR = 3.68 (95% CI 2.77-4.90), deaths from circulatory disease = 1.73 (95% CI 1.54-1.94), respiratory disease = 2.92 (95% CI 2.00-4.23), infection = 2.25 (95% CI 1.70-3.00) and neoplasm = 1.14 (95% CI 1.10-1.21).

CONCLUSION

Despite considerable heterogeneity, all summary SMR estimates and a large majority of individual studies showed elevated mortality in BPAD compared to the general population. This was true for all causes of mortality studied.

Novel bayes factors that capture expert uncertainty in prior density specification in genetic association studies

Bayes factors (BFs) are becoming increasingly important tools in genetic association studies, partly because they provide a natural framework for including prior information. The Wakefield BF (WBF) approximation is easy to calculate and assumes a normal prior on the log odds ratio (logOR) with a mean of zero. However, the prior variance (W) must be specified. Because of the potentially high sensitivity of the WBF to the choice of W, we propose several new BF approximations with logOR ∼N(0,W), but allow W to take a probability distribution rather than a fixed value. We provide several prior distributions for W which lead to BFs that can be calculated easily in freely available software packages. These priors allow a wide range of densities for W and provide considerable flexibility. We examine some properties of the priors and BFs and show how to determine the most appropriate prior based on elicited quantiles of the prior odds ratio (OR). We show by simulation that our novel BFs have superior true-positive rates at low false-positive rates compared to those from both P-value and WBF analyses across a range of sample sizes and ORs. We give an example of utilizing our BFs to fine-map the CASP8 region using genotype data on approximately 46,000 breast cancer case and 43,000 healthy control samples from the Collaborative Oncological Gene-environment Study (COGS) Consortium, and compare the single-nucleotide polymorphism ranks to those obtained using WBFs and P-values from univariate logistic regression.

Comparing the efficacy of SNP filtering methods for identifying a single causal SNP in a known association region

Genome-wide association studies have successfully identified associations between common diseases and a large number of single nucleotide polymorphisms (SNPs) across the genome. We investigate the effectiveness of several statistics, including p-values, likelihoods, genetic map distance and linkage disequilibrium between SNPs, in filtering SNPs in several disease-associated regions. We use simulated data to compare the efficacy of filters with different sample sizes and for causal SNPs with different minor allele frequencies (MAFs) and effect sizes, focusing on the small effect sizes and MAFs likely to represent the majority of unidentified causal SNPs. In our analyses, of all the methods investigated, filtering on the ranked likelihoods consistently retains the true causal SNP with the highest probability for a given false positive rate. This was the case for all the local linkage disequilibrium patterns investigated. Our results indicate that when using this method to retain only the top 5% of SNPs, even a causal SNP with an odds ratio of 1.1 and MAF of 0.08 can be retained with a probability exceeding 0.9 using an overall sample size of 50,000.

Bayesian inference supports a location and neighbour-dependent model of DNA methylation propagation at the MGMT gene promoter in lung tumours

We exploit model-based Bayesian inference methodologies to analyse lung tumour-derived methylation data from a CpG island in the O6-methylguanine-DNA methyltransferase (MGMT) promoter. Interest is in modelling the changes in methylation patterns in a CpG island in the first exon of the promoter during lung tumour development. We propose four competils of methylation state propagation based on two mechanisms. The first is the location-dependence mechanism in which the probability of a gain or loss of methylation at a CpG within the promoter depends upon its location in the CpG sequence. The second mechanism is that of neighbour-dependence in which gain or loss of methylation at a CpG depends upon the methylation status of the immediately preceding CpG. Our data comprises the methylation status at 12 CpGs near the 5' end of the CpG island in two lung tumour samples for both alleles of a nearby polymorphism. We use approximate Bayesian computation, a computationally intensive rejection-sampling algorithm to infer model parameters and compare models without the need to evaluate the likelihood function. We compare the four proposed models using two criteria: the approximate Bayes factors and the distribution of the Euclidean distance between the summary statistics of the observed and simulated datasets. Our model-based analysis demonstrates compelling evidence for both location and neighbour dependence in the process of aberrant DNA methylation of this MGMT promoter CpG island in lung tumours. We find equivocal evidence to support the hypothesis that the methylation patterns of the two alleles evolve independently.

The effectiveness of national guidance in changing analgesic prescribing in primary care from 2002 to 2009: an observational database study

Background

Numerous national guidelines have been issued to assist general practitioners’ safe analgesic prescribing. Their effectiveness is unclear. The objective of this study was to examine trends in general practitioners’ prescribing behaviour in relation to national guidelines.

Methods

This was a retrospective observational database study of registered adult patients prescribed an analgesic (2002–2009) from the Consultations in Primary Care Archive – 12 North Staffordshire general practices. Prescribing guidance from the UK Medicines Regulatory Health Authority (MHRA) regarding non-steroidal anti-inflammatory drugs (NSAIDs) and co-proxamol, and the National Institute for Health and Clinical Excellence (NICE) osteoarthritis (OA) management guidelines were considered. Analgesic prescribing rates were examined, arranged according to a classification of six equipotent medication groups: (1) basic analgesics; (2)–(5) increasingly potent opioids and (6) NSAIDs. In each quarter from 2002 to 2009, the number of patients per 10,000 registered population receiving a prescription for the first time from each group was determined. Quarters associated with significant changes in the underlying prescribing trend were determined using joinpoint regression.

Results

A significant decrease in incident co-proxamol and Cox-2 prescribing occurred around the time of the first MHRA advice to stop using them and were rarely prescribed thereafter. The new prescribing of weak analgesics (e.g., co-codamol 8/500) increased at this same time. Initiating topical NSAIDs significantly increased around the time of the NICE OA guidelines.

Conclusions

Significant prescribing changes occurred when national advice and guidelines were issued. The effectiveness of this advice may vary depending upon the content and method of dissemination. Further evaluation of the optimal methods for delivering prescribing guidance is required.

Cardiovascular risk assessment of South Asians in a religious setting: a feasibility study

AIMS

South Asians in the UK have high cardiovascular disease (CVD) mortality. Therefore, this population is likely to benefit from screening programmes. To address this issue, an initiative was set up between the Royal Free Hampstead NHS Trust, H.E.A.R.T. UK and two Hindu temples in North London to provide screening for CVD risk factors in the community.

METHODS

A total of 434 individuals of Gujarati Indian origin were screened. Measurements included anthropometry, blood pressure and lipid profiles. Three different scoring systems: Framingham, Joint British Societies' 2 and QRISK2 were used to estimate CVD risk.

RESULTS

At least one modifiable CVD risk factor was present in 92% of the individuals screened; 52% were hypertensive, 40% were obese, 75% had central adiposity and 10% had total cholesterol/high density lipoprotein cholesterol ratio > 6. In addition, 37% of a subset of 104 individuals with a fasting sample fulfilled the diagnostic criteria for metabolic syndrome. Overall, 15% of participants screened had a 10-year CV risk score > 20% using QRISK2. The three risk score calculators showed moderate agreement: QRISK2 and JBS2 (kappa 0.61, 95% CI 0.54-0.67), QRISK2 and Framingham (kappa 0.63, 95% CI 0.57-0.70) and JBS2 and Framingham (kappa 0.70, 95% CI 0.64-0.75).

CONCLUSIONS

A high prevalence of modifiable risk factors for CVD was detected in the population screened.

PTEN and PIK3CA gene copy numbers and poor outcomes in non-small cell lung cancer patients with gefitinib therapy

METHODS

Fluorescent in situ hybridisation analyses of PTEN, PIK3CA, EGFR and CEN7 were performed on tumour specimens from patients treated on the expanded access gefitinib trial. Progression-free survival (PFS) and overall survival (OS) were correlated with outcomes in all patients and EGFR wild-type patients.

RESULTS

Progression-free survival (hazard ratio=2.54, P<0.001) and OS (hazard ratio=4.04, P<0.001) were significantly shorter in patients whose tumours had all of the following molecular patterns: CEN7 <4 copies per cell, PTEN loss (<2 copies in at least 20% of cells), and PIK3CA gain (>2 copies in at least 40% of cells) both in all and EGFR wild-type only patients.

CONCLUSION

The combination of low CEN7 copy number, PTEN loss, and PI3KCA gain may be useful for identifying NSCLC patients unlikely to benefit from treatment with EGFR (TKIs), specifically in wild-type EGFR cases.

Two-stage method to remove population- and individual-level outliers from longitudinal data in a primary care database

PURPOSE: In the UK, primary care databases include repeated measurements of health indicators at the individual level. As these databases encompass a large population, some individuals have extreme values, but some values may also be recorded incorrectly. The challenge for researchers is to distinguish between records that are due to incorrect recording and those which represent true but extreme values. This study evaluated different methods to identify outliers. METHODS: Ten percent of practices were selected at random to evaluate the recording of 513,367 height measurements. Population-level outliers were identified using boundaries defined using Health Survey for England data. Individual-level outliers were identified by fitting a random-effects model with subject-specific slopes for height measurements adjusted for age and sex. Any height measurements with a patient-level standardised residual more extreme than ±10 were identified as an outlier and excluded. The model was subsequently refitted twice after removing outliers at each stage. This method was compared with existing methods of removing outliers. RESULTS: Most outliers were identified at the population level using the boundaries defined using Health Survey for England (1550 of 1643). Once these were removed from the database, fitting the random-effects model to the remaining data successfully identified only 75 further outliers. This method was more efficient at identifying true outliers compared with existing methods. CONCLUSIONS: We propose a new, two-stage approach in identifying outliers in longitudinal data and show that it can successfully identify outliers at both population and individual level. Copyright © 2011 John Wiley & Sons, Ltd.

Age-related loss of CpG methylation in the tumour necrosis factor promoter

BACKGROUND

Dysregulated production of TNF has been implicated in the pathogenesis and severity of inflammatory rheumatic diseases, many of which show age-related increased incidence. Ageing is also associated with changes in the immune system including higher systemic levels of pro-inflammatory cytokines. Methylation of DNA is an important regulator of gene expression and changes with age.

OBJECTIVE

In this study we investigated whether the DNA methylation status of the TNF promoter changed with age in peripheral blood leucocytes and macrophages.

METHODS AND RESULTS

Using pyrosequencing assays we detected age-related demethylation of CpG motifs (-304, -245 and -239) in the TNF promoter in human peripheral blood cells from 312 healthy controls (0.8% per decade, confidence interval (CI)=0.44-1.13%, p=1×10(-5)) and primary monocyte-derived macrophages (MDM) from a separate population of 78 healthy controls (1.4% per decade, CI=0.79-2.13%, p=7×10(-5)). Methylation a TNF promoter fragment (-345-+154) resulted in 78% reduction of reporter gene activity compared with the unmethylated promoter construct.

CONCLUSIONS

These data suggest a potential role of accrued changes in DNA methylation in the development of age-related inflammatory diseases, such as rheumatoid arthritis and polymyalgia rheumatica, in which TNF is a pivotal mediator.

Deep supercooling, vitrification and limited survival to –100°C in the Alaskan beetle Cucujus clavipes puniceus (Coleoptera: Cucujidae) larvae

Larvae of the freeze-avoiding beetle Cucujus clavipes puniceus (Coleoptera: Cucujidae) in Alaska have mean supercooling points in winter of –35 to –42°C, with the lowest supercooling point recorded for an individual of –58°C. We previously noted that some larvae did not freeze when cooled to –80°C, and we speculated that these larvae vitrified. Here we present evidence through differential scanning calorimetry that C. c. puniceus larvae transition into a glass-like state at temperatures &lt;–58°C and can avoid freezing to at least –150°C. This novel finding adds vitrification to the list of insect overwintering strategies. While overwintering beneath the bark of fallen trees, C. c. puniceus larvae may experience low ambient temperatures of around –40°C (and lower) when microhabitat is un-insulated because of low snow cover. Decreasing temperatures in winter are correlated with loss of body water from summer high levels near 2.0 to winter lows near 0.4 mg mg–1 dry mass and concomitant increases in glycerol concentrations (4–6 mol l–1) and thermal hysteresis. Finally, we provide direct evidence that Cucujus from Wiseman, Alaska, survive temperatures to –100°C.