The risk of acute infections in new users of antidepressants: An observational cohort study.

BACKGROUND: Preclinical studies suggested that drugs that functionally inhibit acid sphingomyelinase (FIASMA)may enhance immune cell longevity and potentially offer protection against infections. Many antidepressants have shown FIASMA activity. METHODS: We conducted a cohort study using primary-care data from the UK-based Clinical Practice Research Datalink (2000-2021). We assessed the association of composite diagnosed acute infections in new users of fluoxetine, sertraline, paroxetine, or venlafaxine aged 18-80 years compared to citalopram. We compared SARS-CoV-2 infections between groups in a secondary analysis. We estimated incidence rates (IR) and IR ratios (IRR) of acute infections in four pairwise comparisons using negative binomial regression. We applied propensity score (PS) fine stratification to control for confounding. RESULTS: In the PS-weighted cohorts, we included 353,138 fluoxetine, 222,463 sertraline, 69,963 paroxetine, 32,608 venlafaxine, and between 515,996 and 516,583 new citalopram users. PS-weighted IRs ranged between 76.8 acute infections /1000 person-years (py) (sertraline) and 98.9 infections/1000 py (citalopram). We observed PS-weighted IRRs around unity for paroxetine (0.97, 95 % CI, 0.95-1.00), fluoxetine (0.94, 95 % CI, 0.92-0.95), and venlafaxine (0.90, 95 % CI, 0.87-0.94) vs citalopram. Reduced IRR for sertraline vs citalopram (0.84, 95 % CI, 0.82-0.85), became null within subgroups by cohort entry date. In the analysis of SARS-CoV-2 infection, no statistically relevant risk reduction was seen. LIMITATIONS: Analysis not limited to patients with diagnosed depression, possible underestimation of infection incidence, and unclear FIASMA activity of citalopram. CONCLUSIONS: Fluoxetine, sertraline, paroxetine, and venlafaxine were not associated with a reduced risk of acute infection when compared with the presumably weak FIASMA citalopram.

Risk of skin cancer in new users of thiazides and thiazide-like diuretics: a cohort study using an active comparator group.

BACKGROUND: Case-control studies report a dose-dependent increased risk of skin cancer in users of hydrochlorothiazide (HCTZ) vs. nonusers. The degree to which other thiazides and thiazide-like diuretics (TZs) are associated with skin cancer is less certain. OBJECTIVES: To assess the risk of skin cancer in new users of different TZs compared with new users of calcium channel blockers (CCBs). METHODS: We conducted a cohort study using a UK primary-care database (1998-2017), including 271 154 new TZ users [87·6% bendroflumethiazide (BFT), 5·8% indapamide and 3·6% HCTZ] and 275 263 CCB users. The outcomes were basal cell carcinoma (BCC), squamous cell carcinoma (SCC) and cutaneous malignant melanoma (CMM). We estimated incidence rates (IRs) and IR ratios (IRRs) in short-term (< 20 prescriptions) and long-term (≥ 20 prescriptions) users of TZs and CCBs using negative binomial regression, and calculated rate differences (RDs) for selected results. We used fine stratification on the propensity score (PS) to control for 23 baseline covariates. RESULTS: Long-term use of HCTZ increased absolute and relative risks of SCC [PS-weighted IRR 1·95; 95% confidence interval (CI) 1·87-2·02; RD per 100 000 person-years 87.4], but not of BCC or CMM. Long-term use of indapamide was associated with an increased incidence of CMM (IRR 1·43; 95% CI 1·35-1·50). BFT was not meaningfully associated with the risk of any type of skin cancer. CONCLUSIONS: Our results corroborate the previously reported increased risk of SCC (but not of BCC or CMM) for long-term use of HCTZ. BFT may be a safer alternative for patients at increased risk of skin cancer.

Hyperlipidaemia and incident osteoarthritis of the hand: a population-based case-control study.

OBJECTIVE: Preclinical evidence suggests that increased cholesterol levels might be involved in the pathophysiology of osteoarthritis of the hand (HOA), but evidence from observational studies remains scarce. We aimed to analyse the association between hyperlipidaemia and incident HOA. DESIGN: We conducted a matched (1:1) case-control study using the UK-based Clinical Practice Research Datalink (CPRD). Cases were patients aged 30-89 years with an incident diagnosis of HOA between 1995 and 2014. In multivariable conditional logistic regression analyses, we calculated odds ratios (OR) for incident HOA in patients with hyperlipidaemia, categorized by gender, age, previous duration of hyperlipidaemia, and recent statin treatment. RESULTS: Among 19,590 cases and 19,590 controls, we observed an increased risk of HOA in patients with hyperlipidaemia (OR 1.37, 95% confidence intervals (CI) 1.28-1.47), when compared to patients without hyperlipidaemia. Thus, of all HOA cases in our study population, 3.6% may have been attributable to the presence of hyperlipidaemia (population attributable risk). Most patients with HOA were elderly, but the strength of the association between HOA and hyperlipidaemia inversely correlated with increasing age, with the highest OR of 1.72 (95% CI 1.24-2.38) in patients aged 29-49 years. Categorization by previous hyperlipidaemia duration, as well as sub-classification of patients with hyperlipidaemia into those with and without recent statin use did not meaningfully change the effect estimate. CONCLUSIONS: Our results suggest that hyperlipidaemia may be an independent risk factor for new onset HOA.

Type II diabetes mellitus and incident osteoarthritis of the hand: a population-based case-control analysis.

OBJECTIVES: Emerging evidence suggests that diabetes may be a risk factor for osteoarthritis (OA). However, previous results on the association between diabetes and all OA were conflicting. We aimed to comprehensively analyse the association between type II diabetes mellitus (T2DM) and osteoarthritis of the hand (HOA) specifically. METHODS: We conducted a matched (1:1) case-control study using the UK-based Clinical Practice Research Datalink (CPRD) of cases aged 30-90 years with an incident diagnosis of HOA from 1995 to 2013. In multivariable conditional logistic regression analyses, we calculated odds ratios (OR) for incident HOA in patients with T2DM, categorized by T2DM severity (HbA1C), duration, and pharmacological treatment. We further performed sensitivity analyses in patients with and without other metabolic diseases (hypertension (HT), hyperlipidaemia (HL), obesity). RESULTS: Among 13,500 cases and 13,500 controls, we observed no statistically significant association between T2DM and HOA (OR 0.95, 95% confidence interval (CI) 0.87-1.04), regardless of T2DM severity, duration, or pharmacological treatment. Having HT did not change the OR. Although we observed slightly increased ORs in overweight T2DM patients with co-occurring HL with or without coexisting HT, none of these ORs were statistically significant. CONCLUSIONS: Our results provide evidence that T2DM is not an independent risk factor for HOA. Concurrence of T2DM with HT, HL, and/or obesity did not change this association significantly.

Antihypertensive drugs and the risk of incident rosacea.

BACKGROUND: Despite scarce evidence, use of calcium channel blockers is discouraged in patients with rosacea, whereas beta-blockers are recommended as an off-label treatment for erythematotelangiectatic rosacea. OBJECTIVES: To study the association of the use of calcium channel blockers, beta-blockers and other antihypertensive drugs with incident rosacea. METHODS: We conducted a matched case-control study of antihypertensive drugs and incident rosacea, using the U.K.-based General Practice Research Database. Cases had an incident diagnosis of rosacea recorded between 1995 and 2009. Each case was matched to one control on age, sex, general practice and years of history on the database before the index date. Drug use was stratified by timing (≤ or > 180 days before the index date) and duration (number of prescriptions) of drug exposure, in a multivariate conditional logistic regression model. RESULTS: Among 53 927 cases and 53 927 controls, we observed odds ratios (ORs) around unity for calcium channel blockers across all strata, with a slightly decreased OR of 0·77 (95% CI 0·69-0·86) for current users of dihydropyridine calcium channel blockers with ≥ 40 prescriptions. Among beta-blockers, atenolol and bisoprolol yielded slightly decreased ORs across all exposure strata, whereas propranolol revealed ORs around 1·0, irrespective of timing and duration of exposure. Neither angiotensin-converting-enzyme inhibitors nor angiotensin receptor blockers altered the relative rosacea risk. CONCLUSIONS: Our data contradict the prevailing notion that calcium channel blockers increase the risk of rosacea. Beta-blocker use was associated with a slightly decreased risk of rosacea, but the effect may be somewhat stronger in patients with erythematotelangiectatic rosacea.

The association between psychiatric diseases, psychotropic drugs and the risk of incident rosacea.

BACKGROUND: Psychological conditions, such as traumatic events or stress, have been discussed controversially as aetiological factors for rosacea. OBJECTIVES: To assess the association between diagnosed depression, other affective disorders or schizophrenia and subsequent incident rosacea. We further aimed at evaluating the possible role of various psychotropic drugs within this association. METHODS: We conducted a matched case-control study of psychiatric diseases and incident rosacea, stratified by exposure to various psychotropic drugs, using the UK-based General Practice Research Database. Cases had a first diagnosis of rosacea recorded between 1995 and 2009. Each case was matched to one control on age, sex, general practice and years of history on the database. RESULTS: A history of depression or other affective disorders was not associated with an increased risk of developing rosacea; lithium was the only antidepressant drug that significantly altered this association. Current long-term use of lithium was associated with a decreased odds ratio (OR) of 0·58 [95% confidence interval (CI) 0·38-0·88] among people without a schizophrenia diagnosis (with or without affective disorders), compared with people not exposed to lithium. Patients with diagnosed schizophrenia revealed a decreased rosacea risk (OR 0·71, 95% CI 0·60-0·91), independent of antipsychotic drug use. CONCLUSIONS: Depression or other affective disorders were not associated with incident rosacea, whereas patients with schizophrenia were at a decreased risk of this skin disease in our study population. The materially decreased risk of rosacea among people with chronic lithium exposure may lead to new insights into the pathomechanism of rosacea.

A study on the epidemiology of rosacea in the U.K.

BACKGROUND: Rosacea is a chronic facial skin disease of unclear origin. Epidemiological data are scarce and controversial, with reported prevalences ranging from 0·09% to 22%. To our knowledge, incidence rates have not been quantified before. OBJECTIVES: In this observational study we quantified incidence rates of diagnosed rosacea in the U.K. and described demographic characteristics and the prevalence of ocular symptoms in patients with rosacea. We compared lifestyle factors such as smoking and alcohol consumption between patients with rosacea and controls. METHODS: Using the U.K.-based General Practice Research Database, we identified patients with an incident diagnosis of rosacea between 1995 and 2009 and matched them (1:1) to rosacea-free control patients. We assessed person-time of all patients at risk and assessed incidence rates of rosacea, stratified by age, sex, year of diagnosis and region. RESULTS: We identified 60,042 rosacea cases and 60,042 controls (61·5% women). The overall incidence rate for diagnosed rosacea in the U.K. was 1·65 per 1000 person-years. Rosacea was diagnosed in some 80% of cases after the age of 30 years. Ocular symptoms were recorded in 20·8% of cases at the index date. We observed a significantly reduced relative risk of developing rosacea among current smokers (odds ratio 0·64, 95% confidence interval 0·62-0·67). Alcohol consumption was associated with a marginal risk increase. CONCLUSIONS: We quantified incidence rates and characteristics of patients with rosacea diagnosed in clinical practice in a large epidemiological study using primary care data from the U.K. Smoking was associated with a substantially reduced risk of developing rosacea.