Umeå University's logo

Objectives: Monocytes and dendritic cells (DCs) are essential players in the immune response to infections, involved in shaping innate and adaptive immunity. However, a complete understanding of their specific roles in respiratory infections, including SARS-CoV-2, remains elusive.

Methods: To investigate the dynamics of monocytes and DCs in blood as well as the upper and lower airways, we sampled 147 patients with varying degree of COVID-19 severity longitudinally during the spring of 2020.

Results: Using flow cytometry, proteomics and in vitro TLR stimulation, we found differences in the distribution and function of monocytes and DCs in patients compared with controls, and importantly, reduced levels of DCs in both blood and airways. In fact, lower frequencies of cDC2s (Lin− HLA-DR+ CD1c+) early after symptom onset predicted subsequent severe disease, and depletion of DC subsets lasted longer in patients with more severe disease. In contrast, severe COVID-19 was associated with increased frequencies of activated monocytes in the lower, but not the upper, airways. Proteomic analysis showed that monocyte and DC-related cytokines in plasma and airways associated with disease severity. During convalescence, cell frequencies and responses to TLR ligands normalised in blood, except for persistently low plasmacytoid DCs.

Conclusion: Our study reveals a distinct pattern of recruitment of monocytes but not DCs to the airways during severe COVID-19. Instead, decreased levels of DCs in both blood and airways were found, possibly contributing to more severe COVID-19. The connection between low blood DCs early in disease course and more severe outcomes provides insight into COVID-19 immunopathology, with possible therapeutic implications.

Objectives: COVID-19 impacts physical and respiratory health, and the clinical presentation ranges from asymptomatic cases to severe infections requiring hospitalisation. While the long-term effects on lung function and physical capacity are well-documented in moderate to severe cases, the long-term outcome for individuals with mild COVID-19 remains poorly understood. This study investigates the long-term recovery of physical capacity and breathlessness among both hospitalised and non-hospitalised individuals.

Methods: This prospective cohort study enrolled individuals with confirmed SARS-CoV-2 infection between April 2020 and May 2021 through the CoVUm-study. Participants underwent assessments of lung function at 3–6 months after infection and attended follow-ups up to three years post-infection. Physical capacity was evaluated at follow-ups, using the one-minute sit-to-stand test and the modified Medical Research Council scale to assess breathlessness.

Results: The cohort included 291 participants, 35% of whom were hospitalised during SARS-CoV-2 infection. At the 3-year follow-up, 191 participants completed the physical capacity test and 179 had an assessment of breathlessness. Physical capacity improved significantly in the total cohort up to two years post-infection, where improvement plateaued. Hospitalisation and impaired diffusing capacity were significantly associated with reduced physical capacity (beta –6.4, p < 0.001; beta –8.9, p < 0.001, respectively) and breathlessness (beta 3.9, p < 0.001; beta 1.6, p = 0.012, respectively). While non-hospitalised participants demonstrated improvements in physical capacity for up to two years, improvement for hospitalised individuals plateaued by six months.

Conclusion: Hospitalisation and impaired diffusing capacity are strong independent predictors of reduced physical capacity and persistent breathlessness up to three years post-infection. Non-hospitalised individuals also experience long-term reductions in physical capacity, underscoring the need for targeted rehabilitation strategies.

Introduction: Little is known of the long-term prognosis of patients with acute ischaemic stroke in the absence of standard modifiable stroke risk factors (SMoRFs). In acute coronary syndromes, patients without modifiable risk factors have a higher mortality rate. We analysed data from the Swedish Stroke Register to determine survival of patients without SMoRFs following an ischaemic stroke.

Patients and methods: We identified adult patients with first-presentation acute ischaemic stroke between 2010 and 2020. Patients were considered to possess a SMoRF if they had one of: hypertension, diabetes, hyperlipidaemia, atrial fibrillation or an active smoking history. We compared mortality in patients with and without SMoRFs following first-presentation ischaemic stroke using cox regression models. We also assessed the combined endpoint death and dependency (mRS 3–6) at 3 months via logistic regression models.

Results: Of 152,588 patients with ischaemic stroke, hypertension (58.7%) and atrial fibrillation (27.3%) were the most common risk factors. 34,019 patients (22.3%) had no SMoRFs. After a first-presentation ischaemic stroke, patients without SMoRFs had a lower risk of death than patients with one or more SMoRFs (HR 0.58 [95% CI 0.57–0.59]). The absence of SMoRFs was associated with lower odds of death and dependency at 3 months in logistic regression models (OR 0·60 [95% CI 0.58–0.62]).

Conclusion: One in five patients with acute ischaemic stroke had no standard modifiable stroke risk factors. These patients have lower risk of death compared to patients with one or more SMoRFs.

Background: Olfactory impairment has been associated with adverse health outcomes, particularly in older populations, including cognitive decline, malnutrition, and frailty. The COVID-19 pandemic highlighted olfactory impairment as a key symptom affecting individuals across all age groups, raising concerns about its long-term impacts. This study investigates the association between post-acute olfactory impairment and long-term physical capacity in COVID-19 patients, hypothesizing that impaired olfaction is linked to reduced physical performance.

Methods: This prospective cohort study included 63 hospitalized and non-hospitalized COVID-19 patients (38.1 % women; median age 51 years, IQR 47.0–60.0) who underwent olfactory testing 1–3 months post-infection. Olfactory assessments included threshold screening, supra-threshold intensity ratings, and an odour identification test. Physical capacity was assessed using the 1-min sit-to-stand test at follow-ups (3, 6, 12, and 24 months). Partial correlation analysis and linear mixed models were used to analyse the data, adjusting for covariates such as age, sex, BMI, comorbidities, smoking status, and severity of infection.

Results: In the early post-acute phase, 36.5 % of participants exhibited olfactory impairment. We identified a significant, negative correlation between objectively tested olfactory impairment and physical capacity at all follow-ups. In a linear mixed model adjusted for relevant covariates, olfactory impairment was associated with reduced physical capacity up to 24 months after infection. The association strengthened over time, reflected by the increasing beta values for the interaction term: 0.09 (p = 0.200) at 6 months, 0.13 (p = 0.053) at 12 months, and 0.23 (p = 0.001) at 24 months.

Conclusion: Individuals with olfactory impairment in the early post-acute phase of COVID-19 infection were more likely to exhibit diminished physical capacity 24 months later. This study highlights the broader implications of olfactory impairment, previously noted mainly in older populations, demonstrating its relevance across age groups. The COVID-19 pandemic presented a unique opportunity to investigate this relationship, enhancing our understanding of how olfactory impairments relate to long-term physical performance. These findings emphasize the need for further research with larger, more diverse cohorts and objective longitudinal assessments to confirm and extend these observations.

Background: Current evidence indicates that Post COVID-19 Condition (PCC) is multifaceted with distinct phenotypes. While previous studies have identified symptom clusters—commonly featuring fatigue, respiratory symptoms, and cognitive impairment—findings have been inconsistent, and no clear consensus exists. Moreover, how these symptom clusters evolve over time, particularly beyond the first year post-infection, remains poorly understood.

Methods: This multicentre prospective cohort study included 470 hospitalised and non-hospitalised adult individuals from the CoVUm study across four sites in Sweden between 2020 and 2021. Follow-ups were conducted up to 3 years after infection to assess persistent symptoms, health-related quality of life (HRQoL), and work capacity. Symptom clusters at 6 months were identified via hierarchical cluster analysis, and participants were tracked using a k-nearest neighbour algorithm.

Results: The most common symptoms at 6 months were fatigue (33 %), dyspnoea (32 %), mental fatigue (30 %), and concentration difficulties (28 %), with a median EQ-5D-5L index of 0.98 (IQR 0.93–1). Four distinct symptom clusters were identified: (i) “Few Symptoms” (n = 265, 57 %), (ii) “Respiratory Symptoms” (n = 66, 14 %), (iii) “Neurocognitive Symptoms” (n = 75, 16 %), and (iv) “Multisystem Symptoms” (n = 52, 11 %). Participants in the latter three clusters were older, had more comorbidities, and were more often hospitalised during primary COVID-19 infection. These clusters also had significantly lower HRQoL compared to the “Few Symptoms” cluster. Over time, more than half of participants transitioned to a cluster with fewer or no symptoms, with significant perceived HRQoL improvement in the “Multisystem Symptoms” cluster.

Conclusion: While many patients with PCC improved over time, a subset had persistent symptoms at 3 years, especially if primary infection required hospitalisation. The identification of symptom clusters and their trajectories over time contributes to a better understanding of PCC heterogeneity, ultimately bringing the field closer to consensus on the classification and long-term impact of PCC.

Bayesian Gaussian Graphical Models (BGGMs) are tools of growing popularity and interest in network psychometrics and probabilistic graphical modeling. However, some of the existing models are derived from different modeling principles that do not easily allow for extensions and combinations into new models. More specifically, the implementation of some models may not be flexible enough to test different priors or likelihoods. In this paper, we present a new approach to BGGMs that overcomes this limitation by allowing for the estimation of regularized partial correlations between any type of variables while also having an intuitive approach on how to decide about the priors. Our approach is based on using a transformation of the lower diagonal values of the Cholesky (or LDL) decomposition matrix as the parameters of the models, which can receive any zero-centered symmetric distribution as a prior, as well as to include moderators. We have developed the gbggm R package to implement some models based on this approach, and the potentials of the approach are demonstrated with a toy simulation and an empirical example. This new approach expands the range of applications and enhances the flexibility of BGGMs, making them more useful in a variety of contexts.

The objective of this thesis is to examine some challenges that may emerge when conducting time-to-event studies based on observational data. Time-to-event (also called survival) is a setting that involves analyzing how different factors may influence the length of time until an individual experiences the event of interest. This type of analysis is commonly applied in fields such as medical research and epidemiology. In this thesis, which focuses on stroke, we are interested in the time to a recurrent stroke or the death of a patient who survived a first stroke.

Hazard ratios are one of the main parameters estimated in time-to-event studies. Hazard ratios involve comparing the risk of experiencing the event between two groups, usually a treated group and an untreated group.  They can also involve other factors, such as different age groups. Hazard ratios can be estimated from the data by using the Cox regression model.

Observational data, in contrast to experimental data, involves data collected without any intervention or random assignment of treatment to the individuals. Confounders, that is, variables that distort or obscure the true relationship between treatment and outcome, are always present and need to be controlled for in observational studies.

National registers are an important source of observational data. A national registry is a centralized database or system that collects, stores, and maintains information about a specific population or group of individuals within a country. Sweden is known for its detailed and complete national registers. In this thesis, data from the Swedish Stroke Register (Riksstroke) is used to study factors related to stroke.

In time-to-event studies involving observational data, several challenges may arise for the researcher during data analysis. Some individuals may not experience the event during the observation period and thus the information about their time until the event is incomplete. These individuals are considered as censored. Some individuals may experience another event rather than the one of interest, a competing risk. Additionally, models must be properly constructed, with researchers selecting variables and determining the suitable functional form.

Four papers are included in the thesis. Paper I demonstrates how to handle competing risks in survival analysis. The study involves comparing individuals with and without standard modifiable risk factors and their risks of a recurrent stroke or death using data from the Swedish Stroke Register.

The estimation of marginal hazard ratios is a common theme in the other three papers. All involve simulation studies in order to extend methods and explore best practices when estimating marginal hazard ratios.

Paper II explores non-parametric methods that can be used as alternatives to more traditional parametric methods when balancing datasets in order to estimate a marginal hazard ratio. A case study was also conducted using data from the Swedish Stroke Register involving the prescription of anticoagulants at hospital discharge after a stroke.

Paper III is about how censoring affects marginal hazard ratio estimation, even with perfect balancing of the dataset. We study this issue, taking into consideration varying effect sizes and censoring rates. A procedure to attenuate the problem is also studied.

Paper IV concerns covariate selection in the case of high-dimensional data. High-dimensional data involves cases in which the number of covariates in the study is comparable to the number of individuals, and therefore covariate selection methods are needed. In the paper, we explore some of these methods and suggest a best-performing procedure. As Paper II, Paper IV involves a case study of anticoagulant prescription using data from the Swedish Stroke Register.

In observational studies weighting techniques are often used to overcome bias due to confounding. Modeling approaches, such as inverse propensity score weighting, are popular, but often rely on the correct specification of a parametric model wherein neither balance nor stability are targeted. More recently, balancing approach methods that directly target covariate imbalances have been proposed, and these allow the researcher to explicitly set the desired balance constraints. In this study, we evaluate the finite sample properties of different modeling and balancing approach methods, when estimating the marginal hazard ratio, through Monte Carlo simulations. The use of the different methods is also illustrated by analyzing data from the Swedish stroke register to estimate the effect of prescribing oral anticoagulants on time to recurrent stroke or death in stroke patients with atrial fibrillation. In simulated scenarios with good overlap and low or no model misspecification the balancing approach methods performed similarly to the modeling approach methods. In scenarios with bad overlap and model misspecification, the modeling approach method incorporating variable selection performed better than the other methods. The results indicate that it is valuable to use methods that target covariate balance when estimating marginal hazard ratios, but this does not in itself guarantee good performance in situations with, e.g., poor overlap, high censoring, or misspecified models/balance constraints.

Reduction of graphs is a class of procedures used to decrease the dimensionality of a given graphin which the properties of the reduced graph are to be induced from the properties of the largeroriginal graph. This paper introduces both a new method for reducing chain graphs to simplerdirected acyclic graphs (DAGs), that we call power chain graphs (PCG), as well as a procedure forstructure learning of this new type of graph from correlational data of a Gaussian graphical model.Adefinitionfor PCGs is given, directly followed by the reduction method. The structure learningprocedure is a two-step approach:first,the correlation matrix is used to cluster the variables; andthen, the averaged correlation matrix is used to discover the DAGs using the PC-stable algorithm.The results of simulations are provided to illustrate the theoretical proposal, which demonstrateinitial evidence for the validity of our procedure to recover the structure of power chain graphs.The paper ends with a discussion regarding suggestions for future studies as well as some practicalimplications

Hazard ratios are frequently reported in time-to-event and epidemiological studies to assess treatment effects. In observational studies, the combination of propensity score weights with the Cox proportional hazards model facilitates the estimation of the marginal hazard ratio (MHR). The methods for estimating MHR are analogous to those employed for estimating common causal parameters, such as the average treatment effect. However, MHR estimation in the context of high-dimensional data remain unexplored. This paper seeks to address this gap through a simulation study that consider variable selection methods from causal inference combined with a recently proposed multiply robust approach for MHR estimation. Additionally, a case study utilizing stroke register data is conducted to demonstrate the application of these methods. The results from the simulation study indicate that the double selection covariate selection method is preferable to several other strategies when estimating MHR. Nevertheless, the estimation can be further improved by employing the multiply robust approach to the set of propensity score models obtained during the double selection process.