Fast Calibrated Additive Quantile Regression

Proteomic Profiles of Maternal Plasma Extracellular Vesicles for Prediction of Preeclampsia

PROBLEM

Preeclampsia is a heterogeneous syndrome of diverse etiologies and molecular pathways leading to distinct clinical subtypes. Herein, we aimed to characterize the extracellular vesicle (EV)-associated and soluble fractions of the maternal plasma proteome in patients with preeclampsia and to assess their value for disease prediction.

METHOD OF STUDY

This case-control study included 24 women with term preeclampsia, 23 women with preterm preeclampsia, and 94 healthy pregnant controls. Blood samples were collected from cases on average 7 weeks before the diagnosis of preeclampsia and were matched to control samples. Soluble and EV fractions were separated from maternal plasma; EVs were confirmed by cryo-EM, NanoSight, and flow cytometry; and 82 proteins were analyzed with bead-based, multiplexed immunoassays. Quantile regression analysis and random forest models were implemented to evaluate protein concentration differences and their predictive accuracy. Preeclampsia subgroups defined by molecular profiles were identified by hierarchical cluster analysis. Significance was set at p < 0.05 or false discovery rate-adjusted q < 0.1.

RESULTS

In preterm preeclampsia, PlGF, PTX3, and VEGFR-1 displayed differential abundance in both soluble and EV fractions, whereas angiogenin, CD40L, endoglin, galectin-1, IL-27, CCL19, and TIMP1 were changed only in the soluble fraction (q < 0.1). The direction of changes in the EV fraction was consistent with that in the soluble fraction for nine proteins. In term preeclampsia, CCL3 had increased abundance in both fractions (q < 0.1). The combined EV and soluble fraction proteomic profiles predicted preterm and term preeclampsia with an AUC of 78% (95% CI, 66%-90%) and 68% (95% CI, 56%-80%), respectively. Three clusters of preeclampsia featuring distinct clinical characteristics and placental pathology were identified based on combined protein data.

CONCLUSIONS

Our findings reveal distinct alterations of the maternal EV-associated and soluble plasma proteome in preterm and term preeclampsia and identify molecular subgroups of patients with distinct clinical and placental histopathologic features.

The role of general cognitive skills in integrating visual and linguistic information during sentence comprehension: individual differences across the lifespan

Individuals exhibit massive variability in general cognitive skills that affect language processing. This variability is partly developmental. Here, we recruited a large sample of participants (N = 487), ranging from 9 to 90 years of age, and examined the involvement of nonverbal processing speed (assessed using visual and auditory reaction time tasks) and working memory (assessed using forward and backward Digit Span tasks) in a visual world task. Participants saw two objects on the screen and heard a sentence that referred to one of them. In half of the sentences, the target object could be predicted based on verb-selectional restrictions. We observed evidence for anticipatory processing on predictable compared to non-predictable trials. Visual and auditory processing speed had main effects on sentence comprehension and facilitated predictive processing, as evidenced by an interaction. We observed only weak evidence for the involvement of working memory in predictive sentence comprehension. Age had a nonlinear main effect (younger adults responded faster than children and older adults), but it did not differentially modulate predictive and non-predictive processing, nor did it modulate the involvement of processing speed and working memory. Our results contribute to delineating the cognitive skills that are involved in language-vision interactions.

Prelingually Deaf Children With Cochlear Implants Show Better Perception of Voice Cues and Speech in Competing Speech Than Postlingually Deaf Adults With Cochlear Implants

OBJECTIVES

Postlingually deaf adults with cochlear implants (CIs) have difficulties with perceiving differences in speakers' voice characteristics and benefit little from voice differences for the perception of speech in competing speech. However, not much is known yet about the perception and use of voice characteristics in prelingually deaf implanted children with CIs. Unlike CI adults, most CI children became deaf during the acquisition of language. Extensive neuroplastic changes during childhood could make CI children better at using the available acoustic cues than CI adults, or the lack of exposure to a normal acoustic speech signal could make it more difficult for them to learn which acoustic cues they should attend to. This study aimed to examine to what degree CI children can perceive voice cues and benefit from voice differences for perceiving speech in competing speech, comparing their abilities to those of normal-hearing (NH) children and CI adults.

DESIGN

CI children's voice cue discrimination (experiment 1), voice gender categorization (experiment 2), and benefit from target-masker voice differences for perceiving speech in competing speech (experiment 3) were examined in three experiments. The main focus was on the perception of mean fundamental frequency (F0) and vocal-tract length (VTL), the primary acoustic cues related to speakers' anatomy and perceived voice characteristics, such as voice gender.

RESULTS

CI children's F0 and VTL discrimination thresholds indicated lower sensitivity to differences compared with their NH-age-equivalent peers, but their mean discrimination thresholds of 5.92 semitones (st) for F0 and 4.10 st for VTL indicated higher sensitivity than postlingually deaf CI adults with mean thresholds of 9.19 st for F0 and 7.19 st for VTL. Furthermore, CI children's perceptual weighting of F0 and VTL cues for voice gender categorization closely resembled that of their NH-age-equivalent peers, in contrast with CI adults. Finally, CI children had more difficulties in perceiving speech in competing speech than their NH-age-equivalent peers, but they performed better than CI adults. Unlike CI adults, CI children showed a benefit from target-masker voice differences in F0 and VTL, similar to NH children.

CONCLUSION

Although CI children's F0 and VTL voice discrimination scores were overall lower than those of NH children, their weighting of F0 and VTL cues for voice gender categorization and their benefit from target-masker differences in F0 and VTL resembled that of NH children. Together, these results suggest that prelingually deaf implanted CI children can effectively utilize spectrotemporally degraded F0 and VTL cues for voice and speech perception, generally outperforming postlingually deaf CI adults in comparable tasks. These findings underscore the presence of F0 and VTL cues in the CI signal to a certain degree and suggest other factors contributing to the perception challenges faced by CI adults.

Age at death during the Covid-19 lockdown in French metropolitan regions: a non parametric quantile regression approach

BACKGROUND

Lockdowns have been implemented to limit the number of hospitalisations and deaths during the first wave of 2019 coronavirus disease. These measures may have affected differently death characteristics, such age and sex. France was one of the hardest hit countries in Europe with a decreasing east-west gradient in excess mortality. This study aimed at describing the evolution of age at death quantiles during the lockdown in spring 2020 (17 March-11 May 2020) in the French metropolitan regions focusing on 3 representatives of the epidemic variations in the country: Bretagne, Ile-de-France (IDF) and Bourgogne-Franche-Comté (BFC).

METHODS

Data were extracted from the French public mortality database from 1 January 2011 to 31 August 2020. The age distribution of mortality observed during the lockdown period (based on each decile, plus quantiles 1, 5, 95 and 99) was compared with the expected one using Bayesian non-parametric quantile regression.

RESULTS

During the lockdown, 5457, 5917 and 22 346 deaths were reported in Bretagne, BFC and IDF, respectively. An excess mortality from + 3% in Bretagne to + 102% in IDF was observed during lockdown compared to the 3 previous years. Lockdown led to an important increase in the first quantiles of age at death, irrespective of the region, while the increase was more gradual for older age groups. It corresponded to fewer young people, mainly males, dying during the lockdown, with an increase in the age at death in the first quantile of about 7 years across regions. In females, a less significant shift in the first quantiles and a greater heterogeneity between regions were shown. A greater shift was observed in eastern region and IDF, which may also represent excess mortality among the elderly.

CONCLUSIONS

This study focused on the innovative outcome of the age distribution at death. It shows the first quantiles of age at death increased differentially according to sex during the lockdown period, overall shift seems to depend on prior epidemic intensity before lockdown and complements studies on excess mortality during lockdowns.

Non-parametric quantile regression-based modelling of additive effects to solar irradiation in Southern Africa

Modelling of solar irradiation is paramount to renewable energy management. This warrants the inclusion of additive effects to predict solar irradiation. Modelling of additive effects to solar irradiation can improve the forecasting accuracy of prediction frameworks. To help develop the frameworks, this current study modelled the additive effects using non-parametric quantile regression (QR). The approach applies quantile splines to approximate non-parametric components when finding the best relationships between covariates and the response variable. However, some additive effects are perceived as linear. Thus, the study included the partial linearly additive quantile regression model (PLAQR) in the quest to find how best the additive effects can be modelled. As a result, a comparative investigation on the forecasting performances of the PLAQR, an additive quantile regression (AQR) model and the new quantile generalised additive model (QGAM) using out-of-sample and probabilistic forecasting metric evaluations was done. Forecasted density plots, Murphy diagrams and results from the Diebold-Mariano (DM) hypothesis test were also analysed. The density plot, the curves on the Murphy diagram and most metric scores computed for the QGAM were slightly better than for the PLAQR and AQR models. That is, even though the DM test indicates that the PLAQR and AQR models are less accurate than the QGAM, we could not conclude an outright greater forecasting performance of the QGAM than the PLAQR or AQR models. However, in situations of probabilistic forecasting metric preferences, each model can be prioritised to be applied to the metric where it performed slightly the best. The three models performed differently in different locations, but the location was not a significant factor in their performances. In contrast, forecasting horizon and sample size influenced model performance differently in the three additive models. The performance variations also depended on the metric being evaluated. Therefore, the study has established the best forecasting horizons and sample sizes for the different metrics. It was finally concluded that a 20% forecasting horizon and a minimum sample size of 10000 data points are ideal when modelling additive effects of solar irradiation using non-parametric QR.

Elevated Southern Hemisphere moisture availability during glacial periods

Late Pleistocene ice-age climates are routinely characterized as having imposed moisture stress on low- to mid-latitude ecosystems1-5. This idea is largely based on fossil pollen evidence for widespread, low-biomass glacial vegetation, interpreted as indicating climatic dryness6. However, woody plant growth is inhibited under low atmospheric CO2 (refs. 7,8), so understanding glacial environments requires the development of new palaeoclimate indicators that are independent of vegetation9. Here we show that, contrary to expectations, during the past 350 kyr, peaks in southern Australian climatic moisture availability were largely confined to glacial periods, including the Last Glacial Maximum, whereas warm interglacials were relatively dry. By measuring the timing of speleothem growth in the Southern Hemisphere subtropics, which today has a predominantly negative annual moisture balance, we developed a record of climatic moisture availability that is independent of vegetation and extends through multiple glacial-interglacial cycles. Our results demonstrate that a cool-moist response is consistent across the austral subtropics and, in part, may result from reduced evaporation under cool glacial temperatures. Insofar as cold glacial environments in the Southern Hemisphere subtropics have been portrayed as uniformly arid3,10,11, our findings suggest that their characterization as evolutionary or physiological obstacles to movement and expansion of animal, plant and, potentially, human populations10 should be reconsidered.

Assessing the external validity and clinical relevance of umbilical doppler resistance index references in daily practice

OBJECTIVE

To assess the external validity and clinical relevance of current references for umbilical artery resistance index (UA RI) in daily practice.

METHODS

Retrospective cross-sectional single center study including all UA RI measurements between 22 and 40 gestational weeks (GW) from distinct patients between 2014 and 2022. Patients with normal pregnancies and normal neonatal outcomes that had an UA RI measurement between 2014 and 2019 were used to calculate reference ranges. The established reference for the 95th centile was compared to two current references. The clinical relevance of the established reference was tested by comparing neonatal outcomes according to the 95th percentile among the consecutive distinct patients between 2020 and 2022.

RESULTS

Among the 13342 consecutive distinct patients with a singleton pregnancy that had an UA RI measurement between 22 and 40 GW between 2014 and 2022, 5298 patients were included to establish the reference ranges, and 3634 patients to validate these ranges. For each gestational age, the established references were similar to current references. Using the established references, the proportion of patients presenting an UA RI>95th percentile among the patients with normal pregnancies in the validation population was comparable to the proportion when using the two current references. Among the validation population, 268 patients (7.4 %) (95%CI[6.5-8.2]) presented an UA RI ≥ 95th percentile. Of these 268 patients, 67.9% had a SGA newborn (versus 19.2%, p<0.001) and 59% a preterm birth (versus 13.9%, p<0.001).

CONCLUSIONS

The reference range obtained from daily practice is clinically relevant and similar to current references.

Soil organic carbon stocks in European croplands and grasslands: How much have we lost in the past decade?

The EU Soil Strategy 2030 aims to increase soil organic carbon (SOC) in agricultural land to enhance soil health and support biodiversity as well as to offset greenhouse gas emissions through soil carbon sequestration. Therefore, the quantification of current SOC stocks and the spatial identification of the main drivers of SOC changes is paramount in the preparation of agricultural policies aimed at enhancing the resilience of agricultural systems in the EU. In this context, changes of SOC stocks (Δ SOCs) for the EU + UK between 2009 and 2018 were estimated by fitting a quantile generalized additive model (qGAM) on data obtained from the revisited points of the Land Use/Land Cover Area Frame Survey (LUCAS) performed in 2009, 2015 and 2018. The analysis of the partial effects derived from the fitted qGAM model shows that land use and land use change observed in the 2009, 2015 and 2018 LUCAS campaigns (i.e. continuous grassland [GGG] or cropland [CCC], conversion grassland to cropland (GGC or GCC) and vice versa [CGG or CCG]) was one of the main drivers of SOC changes. The CCC was the factor that contributed to the lowest negative change on Δ SOC with an estimated partial effect of -0.04 ± 0.01 g C kg-1  year-1 , while the GGG the highest positive change with an estimated partial effect of 0.49 ± 0.02 g C kg-1  year-1 . This confirms the C sequestration potential of converting cropland to grassland. However, it is important to consider that local soil and environmental conditions may either diminish or enhance the grassland's positive effect on soil C storage. In the EU + UK, the estimated current (2018) topsoil (0-20 cm) SOC stock in agricultural land below 1000 m a.s.l was 9.3 Gt, with a Δ SOC of -0.75% in the period 2009-2018. The highest estimated SOC losses were concentrated in central-northern countries, while marginal losses were observed in the southeast.

Phytodiversity is associated with habitat heterogeneity from Eurasia to the Hengduan Mountains

The geographic distribution of plant diversity matches the gradient of habitat heterogeneity from lowlands to mountain regions. However, little is known about how much this relationship is conserved across scales. Using the World Checklist of Vascular Plants and high-resolution biodiversity maps developed by species distribution models, we investigated the associations between species richness and habitat heterogeneity at the scales of Eurasia and the Hengduan Mountains (HDM) in China. Habitat heterogeneity explains seed plant species richness across Eurasia, but the plant species richness of 41/97 HDM families is even higher than expected from fitted statistical relationships. A habitat heterogeneity index combining growing degree days, site water balance, and bedrock type performs better than heterogeneity based on single variables in explaining species richness. In the HDM, the association between heterogeneity and species richness is stronger at larger scales. Our findings suggest that high environmental heterogeneity provides suitable conditions for the diversification of lineages in the HDM. Nevertheless, habitat heterogeneity alone cannot fully explain the distribution of species richness in the HDM, especially in the western HDM, and complementary mechanisms, such as the complex geological history of the region, may have contributed to shaping this exceptional biodiversity hotspot.

Frugal day-ahead forecasting of multiple local electricity loads by aggregating adaptive models

This paper focuses on day-ahead electricity load forecasting for substations of the distribution network in France; therefore, the corresponding problem lies between the instability of a single consumption and the stability of a countrywide total demand. Moreover, this problem requires to forecast the loads of over one thousand substations; consequently, it belongs to the field of multiple time series forecasting. To that end, the paper applies an adaptive methodology that provided excellent results at a national scale; the idea is to combine generalized additive models with state-space representations. However, extending this methodology to the prediction of over a thousand time series raises a computational issue. It is solved by developing a frugal variant that reduces the number of estimated parameters: forecasting models are estimated only for a few time series and transfer learning is achieved by relying on aggregation of experts. This approach yields a reduction of computational needs and their associated emissions. Several variants are built, corresponding to different levels of parameter transfer, to find the best trade-off between accuracy and frugality. The selected method achieves competitive results compared to individual models. Finally, the paper highlights the interpretability of the models, which is important for operational applications.

Recognizing non-native spoken words in background noise increases interference from the native language

Listeners frequently recognize spoken words in the presence of background noise. Previous research has shown that noise reduces phoneme intelligibility and hampers spoken-word recognition - especially for non-native listeners. In the present study, we investigated how noise influences lexical competition in both the non-native and the native language, reflecting the degree to which both languages are co-activated. We recorded the eye movements of native Dutch participants as they listened to English sentences containing a target word while looking at displays containing four objects. On target-present trials, the visual referent depicting the target word was present, along with three unrelated distractors. On target-absent trials, the target object (e.g., wizard) was absent. Instead, the display contained an English competitor, overlapping with the English target in phonological onset (e.g., window), a Dutch competitor, overlapping with the English target in phonological onset (e.g., wimpel, pennant), and two unrelated distractors. Half of the sentences was masked by speech-shaped noise; the other half was presented in quiet. Compared to speech in quiet, noise delayed fixations to the target objects on target-present trials. For target-absent trials, we observed that the likelihood for fixation biases towards the English and Dutch onset competitors (over the unrelated distractors) was larger in noise than in quiet. Our data thus show that the presence of background noise increases lexical competition in the task-relevant non-native (English) and in the task-irrelevant native (Dutch) language. The latter reflects stronger interference of one's native language during non-native spoken-word recognition under adverse conditions.

An interpretable measure of semantic similarity for predicting eye movements in reading

Predictions about upcoming content play an important role during language comprehension and processing. Semantic similarity as a metric has been used to predict how words are processed in context in language comprehension and processing tasks. This study proposes a novel, dynamic approach for computing contextual semantic similarity, evaluates the extent to which the semantic similarity measures computed using this approach can predict fixation durations in reading tasks recorded in a corpus of eye-tracking data, and compares the performance of these measures to that of semantic similarity measures computed using the cosine and Euclidean methods. Our results reveal that the semantic similarity measures generated by our approach are significantly predictive of fixation durations on reading and outperform those generated by the two existing approaches. The findings of this study contribute to a better understanding of how humans process words in context and make predictions in language comprehension and processing. The effective and interpretable approach to computing contextual semantic similarity proposed in this study can also facilitate further explorations of other experimental data on language comprehension and processing.

Status of udder health performance indicators and implementation of on farm monitoring on German dairy cow farms: results from a large scale cross-sectional study

Regional benchmarking data enables farmers to compare their animal health situation to that of other herds and identify areas with improvement potential. For the udder health status of German dairy cow farms, such data were incomplete. Therefore, the aim of this study was (1) to describe the incidence of clinical mastitis (CM), (2) to describe cell count based udder health indicators [annual mean test day average of the proportion of animals without indication of mastitis (aWIM), new infection risk during lactation (aNIR), and proportion of cows with low chance of cure (aLCC); heifer mastitis rate (HM)] and their seasonal variation, and (3) to evaluate the level of implementation of selected measures of mastitis monitoring. Herds in three German regions (North: n = 253; East: n = 252, South: n = 260) with different production conditions were visited. Data on CM incidence and measures of mastitis monitoring were collected via structured questionnaire-based interviews. Additionally, dairy herd improvement (DHI) test day data from the 365 days preceding the interview were obtained. The median (Q0.1, Q0.9) farmer reported incidence of mild CM was 14.8% (3.5, 30.8%) in North, 16.2% (1.9, 50.4%) in East, and 11.8% (0.0, 30.7%) in South. For severe CM the reported incidence was 4.0% (0.0, 12.2%), 2.0% (0.0, 10.8%), and 2.6% (0.0, 11.0%) for North, East, and South, respectively. The median aWIM was 60.7% (53.4, 68.1%), 59.0% (49.7, 65.4%), and 60.2% (51.5, 67.8%), whereas the median aNIR was 17.1% (13.6, 21.6%), 19.9% (16.2, 24.9%), and 18.3% (14.4, 22.0%) in North, East, and South, respectively with large seasonal variations. Median aLCC was ≤1.1% (≤ 0.7%, ≤ 1.8%) in all regions and HM was 28.4% (19.7, 37.2%), 35.7% (26.7, 44.2%), and 23.5% (13.1, 35.9%), in North, East and South, respectively. Participation in a DHI testing program (N: 95.7%, E: 98.8%, S: 89.2%) and premilking (N: 91.1%, E: 93.7%, S: 90.2%) were widely used. Several aspects of udder health monitoring, including exact documentation of CM cases, regular microbiological analysis of milk samples and the use of a veterinary herd health consultancy service were not applied on many farms. The results of this study can be used by dairy farmers and their advisors as benchmarks for the assessment of the udder health situation in their herds.

A Video Segmentation Pipeline for Assessing changes in Pupil Response to Light After Cannabis Consumption

Due to long-standing federal restrictions on cannabis-related research, the implications of cannabis legalization on traffic and occupational safety are understudied. Accordingly, there is a need for objective and validated measures of acute cannabis impairment that may be applied in public safety and occupational settings. Pupillary response to light may offer an avenue for detection that outperforms typical sobriety tests and THC concentrations. We developed a video processing and analysis pipeline that extracts pupil sizes during a light stimulus test administered with goggles utilizing infrared videography. The analysis compared pupil size trajectories in response to a light for those with occasional, daily, and no cannabis use before and after smoking. Pupils were segmented using a combination of image pre-processing techniques and segmentation algorithms which were validated using manually segmented data and found to achieve 99% precision and 94% F-score. Features extracted from the pupil size trajectories captured pupil constriction and rebound dilation and were analyzed using generalized estimating equations. We find that acute cannabis use results in less pupil constriction and slower pupil rebound dilation in the light stimulus test.

Using generalized additive models to decompose time series and waveforms, and dissect heart-lung interaction physiology

Common physiological time series and waveforms are composed of repeating cardiac and respiratory cycles. Often, the cardiac effect is the primary interest, but for, e.g., fluid responsiveness prediction, the respiratory effect on arterial blood pressure also convey important information. In either case, it is relevant to disentangle the two effects. Generalized additive models (GAMs) allow estimating the effect of predictors as nonlinear, smooth functions. These smooth functions can represent the cardiac and respiratory cycles' effects on a physiological signal. We demonstrate how GAMs allow a decomposition of physiological signals from mechanically ventilated subjects into separate effects of the cardiac and respiratory cycles. Two examples are presented. The first is a model of the respiratory variation in pulse pressure. The second demonstrates how a central venous pressure waveform can be decomposed into a cardiac effect, a respiratory effect and the interaction between the two cycles. Generalized additive models provide an intuitive and flexible approach to modelling the repeating, smooth, patterns common in medical monitoring data.

How Has Climate Change Driven the Evolution of Rice Distribution in China?

Estimating the impact of climate change risks on rice distribution is one of the most important elements of climate risk management. This paper is based on the GEE (Google Earth Engine) platform and multi-source remote sensing data; the authors quantitatively extracted rice production distribution data in China from 1990 to 2019, analysed the evolution pattern of rice distribution and clusters and explored the driving effects between climatic and environmental conditions on the evolution of rice production distribution using the non-parametric quantile regression model. The results show that: The spatial variation of rice distribution is significant, mainly concentrated in the northeast, south and southwest regions of China; the distribution of rice in the northeast is expanding, while the distribution of rice in the south is extending northward, showing a spatial evolution trend of "north rising and south retreating". The positive effect of precipitation on the spatial distribution of rice has a significant threshold. This shows that when precipitation is greater than 800 mm, there is a significant positive effect on the spatial distribution of rice production, and this effect will increase with precipitation increases. Climate change may lead to a continuous northward shift in the extent of rice production, especially extending to the northwest of China. This paper's results will help implement more spatially targeted climate change adaptation measures for rice to cope with the changes in food production distribution caused by climate change.

Effect of environmental regulation policy synergy on carbon emissions in China under consideration of the mediating role of industrial structure

A comprehensive understanding of the effect of environmental regulation policy synergy on carbon emissions is essential for the design and decision-making of the other policy combinations. In this paper, the panel data from 30 provinces in China during 2000-2019 are used as samples to investigate the heterogeneity of different policy combinations and compare the differences using a Bayesian non-parametric generalized additive quantile method. The research results imply that the environmental regulation policy synergy has a significant heterogeneous effect on carbon emissions on different levels. Specifically, the effect of environmental regulation policy synergy on carbon emissions has an inverted U-shape trend at the low quintiles, i.e., before the turning point, it shows a "green paradox" effect, and after the turning point, it shows an "emission reduction" effect. However, the effect of environmental regulation policy synergy on carbon emissions shows a significant N-shape trend at the high quintiles, i.e., before the second turning point, it shows an "emission reduction" effect, and after the second turning point, it shows a "green paradox" effect. Moreover, the effect of environmental regulation policy combinations on carbon emissions reduction has been found to vary significantly across regions.

Behavioral Signatures of Memory Resources for Language: Looking beyond the Lexicon/Grammar Divide

Although there is a broad consensus that both the procedural and declarative memory systems play a crucial role in language learning, use, and knowledge, the mapping between linguistic types and memory structures remains underspecified: by default, a dual-route mapping of language systems to memory systems is assumed, with declarative memory handling idiosyncratic lexical knowledge and procedural memory handling rule-governed knowledge of grammar. We experimentally contrast the processing of morphology (case and aspect), syntax (subordination), and lexical semantics (collocations) in a healthy L1 population of Polish, a language rich in form distinctions. We study the processing of these four types under two conditions: a single task condition in which the grammaticality of stimuli was judged and a concurrent task condition in which grammaticality judgments were combined with a digit span task. Dividing attention impedes access to declarative memory while leaving procedural memory unaffected and hence constitutes a test that dissociates which types of linguistic information each long-term memory construct subserves. Our findings confirm the existence of a distinction between lexicon and grammar as a generative, dual-route model would predict, but the distinction is graded, as usage-based models assume: the hypothesized grammar-lexicon opposition appears as a continuum on which grammatical phenomena can be placed as being more or less "ruly" or "idiosyncratic." However, usage-based models, too, need adjusting as not all types of linguistic knowledge are proceduralized to the same extent. This move away from a simple dichotomy fundamentally changes how we think about memory for language, and hence how we design and interpret behavioral and neuroimaging studies that probe into the nature of language cognition.

Benchmarking calf health: Assessment tools for dairy herd health consultancy based on reference values from 730 German dairies with respect to seasonal, farm type, and herd size effects

Good calf health is crucial for a successfully operating farm business and animal welfare on dairy farms. To evaluate calf health on farms and to identify potential problem areas, benchmarking tools can be used by farmers, herd managers, veterinarians, and other advisory persons in the field. However, for calves, benchmarking tools are not yet widely established in practice. This study provides hands-on application for on-farm benchmarking of calf health. Reference values were generated from a large dataset of the “PraeRi” study, including 730 dairy farms with a total of 13,658 examined preweaned dairy calves. At herd level, omphalitis (O, median 15.9%) was the most common disorder, followed by diarrhea (D, 15.4%) and respiratory disease (RD, 2.9%). Abnormal weight bearing (AWB) was rarely detected (median, 0.0%). Calves with symptoms of more than one disorder at the same time (multimorbidity, M) were observed with a prevalence of 2.3%. The enrolled farms varied in herd size, farm operating systems, and management practices and thus represented a wide diversity in dairy farming, enabling a comparison with similar managed farms in Germany and beyond. To ensure comparability of the data in practice, the reference values were calculated for the whole data set, clustered according to farm size (1–40 dairy cows (n = 130), 41–60 dairy cows (n = 99), 61–120 dairy cows (n = 180), 121–240 dairy cows (n = 119) and farms with more than 240 dairy cows (n = 138), farm operating systems (conventional (n = 666), organic (n = 64)) and month of the year of the farm visit. There was a slight tendency for smaller farms to have a lower prevalence of disorders. A statistically significant herd-size effect was detected for RD (p = 0.008) and D (p < 0.001). For practical application of these reference values, tables, diagrams, and an Excel^® (Microsoft^®) based calf health calculator were developed as tools for on-farm benchmarking (https://doi.org/10.6084/m9.figshare.c.6172753). In addition, this study provides a detailed description of the colostrum, feeding and housing management of preweaned calves in German dairy farms of different herd sizes and farm type (e.g., conventional and organic).

Changes in limiting factors for forager population dynamics in Europe across the last glacial-interglacial transition

Population dynamics set the framework for human genetic and cultural evolution. For foragers, demographic and environmental changes correlate strongly, although the causal relations between different environmental variables and human responses through time and space likely varied. Building on the notion of limiting factors, namely that at any one time, the scarcest resource caps population size, we present a statistical approach to identify the dominant climatic constraints for hunter-gatherer population densities and then hindcast their changing dynamics in Europe for the period between 21,000 to 8000 years ago. Limiting factors shifted from temperature-related variables (effective temperature) during the Pleistocene to a regional mosaic of limiting factors in the Holocene dominated by temperature seasonality and annual precipitation. This spatiotemporal variation suggests that hunter-gatherers needed to overcome very different adaptive challenges in different parts of Europe and that these challenges varied over time. The signatures of these changing adaptations may be visible archaeologically. In addition, the spatial disaggregation of limiting factors from the Pleistocene to the Holocene coincided with and may partly explain the diversification of the cultural geography at this time.

How do mobility restrictions and social distancing during COVID-19 affect oil price?

This paper provides an analysis of the effect of the COVID-19 outbreak on the crude oil price. Using a newly developed air mobility index and Apple's driving trends index, we assess the effect of human mobility restrictions and social distancing during the COVID-19 pandemic on the crude oil price. We apply a quantile regression model, which evaluates different quantiles of the crude oil price. We also conduct an extreme value modeling, which examines the lower tail of the crude oil price distribution. We find that both the air mobility index and driving trends index significantly influence lower and upper quantiles of the WTI crude oil price. The extreme value models suggest that there is a potential risk of a negative crude oil price for a sudden extreme fall of air mobility.

Who Is at Risk of Poor Mental Health Following Coronavirus Disease-19 Outpatient Management?

Background

Coronavirus Disease-19 (COVID-19) convalescents are at risk of developing a de novo mental health disorder or worsening of a pre-existing one. COVID-19 outpatients have been less well characterized than their hospitalized counterparts. The objectives of our study were to identify indicators for poor mental health following COVID-19 outpatient management and to identify high-risk individuals.

Methods

We conducted a binational online survey study with adult non-hospitalized COVID-19 convalescents (Austria/AT: n = 1,157, Italy/IT: n = 893). Primary endpoints were positive screening for depression and anxiety (Patient Health Questionnaire; PHQ-4) and self-perceived overall mental health (OMH) and quality of life (QoL) rated with 4 point Likert scales. Psychosocial stress was surveyed with a modified PHQ stress module. Associations of the mental health and QoL with socio-demographic, COVID-19 course, and recovery variables were assessed by multi-parameter Random Forest and Poisson modeling. Mental health risk subsets were defined by self-organizing maps (SOMs) and hierarchical clustering algorithms. The survey analyses are publicly available (https://im2-ibk.shinyapps.io/mental_health_dashboard/).

Results

Depression and/or anxiety before infection was reported by 4.6% (IT)/6% (AT) of participants. At a median of 79 days (AT)/96 days (IT) post-COVID-19 onset, 12.4% (AT)/19.3% (IT) of subjects were screened positive for anxiety and 17.3% (AT)/23.2% (IT) for depression. Over one-fifth of the respondents rated their OMH (AT: 21.8%, IT: 24.1%) or QoL (AT: 20.3%, IT: 25.9%) as fair or poor. Psychosocial stress, physical performance loss, high numbers of acute and sub-acute COVID-19 complaints, and the presence of acute and sub-acute neurocognitive symptoms (impaired concentration, confusion, and forgetfulness) were the strongest correlates of deteriorating mental health and poor QoL. In clustering analysis, these variables defined subsets with a particularly high propensity of post-COVID-19 mental health impairment and decreased QoL. Pre-existing depression or anxiety (DA) was associated with an increased symptom burden during acute COVID-19 and recovery.

Conclusion

Our study revealed a bidirectional relationship between COVID-19 symptoms and mental health. We put forward specific acute symptoms of the disease as "red flags" of mental health deterioration, which should prompt general practitioners to identify non-hospitalized COVID-19 patients who may benefit from early psychological and psychiatric intervention.

Clinical Trial Registration

[ClinicalTrials.gov], identifier [NCT04661462].

Application of the Athlete Biological Passport Approach to the Detection of Growth Hormone Doping

CONTEXT

Because of its anabolic and lipolytic properties, growth hormone (GH) use is prohibited in sport. Two methods based on population-derived decision limits are currently used to detect human GH (hGH) abuse: the hGH Biomarkers Test and the Isoforms Differential Immunoassay.

OBJECTIVE

We tested the hypothesis that longitudinal profiling of hGH biomarkers through application of the Athlete Biological Passport (ABP) has the potential to flag hGH abuse.

METHODS

Insulin-like growth factor 1 (IGF-1) and procollagen III peptide (P-III-NP) distributions were obtained from 7 years of anti-doping data in elite athletes (n = 11 455) and applied as priors to analyze individual profiles from an hGH administration study in recreational athletes (n = 35). An open-label, randomized, single-site, placebo-controlled administration study was carried out with individuals randomly assigned to 4 arms: placebo, or 3 different doses of recombinant hGH. Serum samples were analyzed for IGF-1, P-III-NP, and hGH isoforms and the performance of a longitudinal, ABP-based approach was evaluated.

RESULTS

An ABP-based approach set at a 99% specificity level flagged 20/27 individuals receiving hGH treatment, including 17/27 individuals after cessation of the treatment. ABP sensitivity ranged from 12.5% to 71.4% across the hGH concentrations tested following 7 days of treatment, peaking at 57.1% to 100% after 21 days of treatment, and was maintained between 37.5% and 71.4% for the low and high dose groups 1 week after cessation of treatment.

CONCLUSION

These findings demonstrate that longitudinal profiling of hGH biomarkers can provide suitable performance characteristics for use in anti-doping programs.

Assessing vegetation recovery from energy development using a dynamic reference approach

Ecologically relevant references are useful for evaluating ecosystem recovery, but references that are temporally static may be less useful when environmental conditions and disturbances are spatially and temporally heterogeneous. This challenge is particularly acute for ecosystems dominated by sagebrush (Artemisia spp.), where communities may require decades to recover from disturbance. We demonstrated application of a dynamic reference approach to studying sagebrush recovery using three decades of sagebrush cover estimates from remote sensing (1985-2018). We modelled recovery on former oil and gas well pads (n = 1200) across southwestern Wyoming, USA, relative to paired references identified by the Disturbance Automated Reference Toolset. We also used quantile regression to account for unmodelled heterogeneity in recovery, and projected recovery from similar disturbance across the landscape. Responses to weather and site-level factors often differed among quantiles, and sagebrush recovery on former well pads increased more when paired reference sites had greater sagebrush cover. Little (<5%) of the landscape was projected to recover within 100 years for low to mid quantiles, and recovery often occurred at higher elevations with cool and moist annual conditions. Conversely, 48%-78% of the landscape recovered quickly (within 25 years) for high quantiles of sagebrush cover. Our study demonstrates advantages of using dynamic reference sites when studying vegetation recovery, as well as how additional inferences obtained from quantile regression can inform management.

Pharmacological and non-pharmacological treatments and outcomes for new-onset atrial fibrillation in ICU patients: the CAFE scoping review and database analyses

BACKGROUND

New-onset atrial fibrillation occurs in around 10% of adults treated in an intensive care unit. New-onset atrial fibrillation may lead to cardiovascular instability and thromboembolism, and has been independently associated with increased length of hospital stay and mortality. The long-term consequences are unclear. Current practice guidance is based on patients outside the intensive care unit; however, new-onset atrial fibrillation that develops while in an intensive care unit differs in its causes and the risks and clinical effectiveness of treatments. The lack of evidence on new-onset atrial fibrillation treatment or long-term outcomes in intensive care units means that practice varies. Identifying optimal treatment strategies and defining long-term outcomes are critical to improving care.

OBJECTIVES

In patients treated in an intensive care unit, the objectives were to (1) evaluate existing evidence for the clinical effectiveness and safety of pharmacological and non-pharmacological new-onset atrial fibrillation treatments, (2) compare the use and clinical effectiveness of pharmacological and non-pharmacological new-onset atrial fibrillation treatments, and (3) determine outcomes associated with new-onset atrial fibrillation.

METHODS

We undertook a scoping review that included studies of interventions for treatment or prevention of new-onset atrial fibrillation involving adults in general intensive care units. To investigate the long-term outcomes associated with new-onset atrial fibrillation, we carried out a retrospective cohort study using English national intensive care audit data linked to national hospital episode and outcome data. To analyse the clinical effectiveness of different new-onset atrial fibrillation treatments, we undertook a retrospective cohort study of two large intensive care unit databases in the USA and the UK.

RESULTS

Existing evidence was generally of low quality, with limited data suggesting that beta-blockers might be more effective than amiodarone for converting new-onset atrial fibrillation to sinus rhythm and for reducing mortality. Using linked audit data, we showed that patients developing new-onset atrial fibrillation have more comorbidities than those who do not. After controlling for these differences, patients with new-onset atrial fibrillation had substantially higher mortality in hospital and during the first 90 days after discharge (adjusted odds ratio 2.32, 95% confidence interval 2.16 to 2.48; adjusted hazard ratio 1.46, 95% confidence interval 1.26 to 1.70, respectively), and higher rates of subsequent hospitalisation with atrial fibrillation, stroke and heart failure (adjusted cause-specific hazard ratio 5.86, 95% confidence interval 5.33 to 6.44; adjusted cause-specific hazard ratio 1.47, 95% confidence interval 1.12 to 1.93; and adjusted cause-specific hazard ratio 1.28, 95% confidence interval 1.14 to 1.44, respectively), than patients who did not have new-onset atrial fibrillation. From intensive care unit data, we found that new-onset atrial fibrillation occurred in 952 out of 8367 (11.4%) UK and 1065 out of 18,559 (5.7%) US intensive care unit patients in our study. The median time to onset of new-onset atrial fibrillation in patients who received treatment was 40 hours, with a median duration of 14.4 hours. The clinical characteristics of patients developing new-onset atrial fibrillation were similar in both databases. New-onset atrial fibrillation was associated with significant average reductions in systolic blood pressure of 5 mmHg, despite significant increases in vasoactive medication (vasoactive-inotropic score increase of 2.3; p < 0.001). After adjustment, intravenous beta-blockers were not more effective than amiodarone in achieving rate control (adjusted hazard ratio 1.14, 95% confidence interval 0.91 to 1.44) or rhythm control (adjusted hazard ratio 0.86, 95% confidence interval 0.67 to 1.11). Digoxin therapy was associated with a lower probability of achieving rate control (adjusted hazard ratio 0.52, 95% confidence interval 0.32 to 0.86) and calcium channel blocker therapy was associated with a lower probability of achieving rhythm control (adjusted hazard ratio 0.56, 95% confidence interval 0.39 to 0.79) than amiodarone. Findings were consistent across both the combined and the individual database analyses.

CONCLUSIONS

Existing evidence for new-onset atrial fibrillation management in intensive care unit patients is limited. New-onset atrial fibrillation in these patients is common and is associated with significant short- and long-term complications. Beta-blockers and amiodarone appear to be similarly effective in achieving cardiovascular control, but digoxin and calcium channel blockers appear to be inferior.

FUTURE WORK

Our findings suggest that a randomised controlled trial of amiodarone and beta-blockers for management of new-onset atrial fibrillation in critically ill patients should be undertaken. Studies should also be undertaken to provide evidence for or against anticoagulation for patients who develop new-onset atrial fibrillation in intensive care units. Finally, given that readmission with heart failure and thromboembolism increases following an episode of new-onset atrial fibrillation while in an intensive care unit, a prospective cohort study to demonstrate the incidence of atrial fibrillation and/or left ventricular dysfunction at hospital discharge and at 3 months following the development of new-onset atrial fibrillation should be undertaken.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN13252515.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 25, No. 71. See the NIHR Journals Library website for further project information.

Adaptive Methods for Short-Term Electricity Load Forecasting During COVID-19 Lockdown in France

The coronavirus disease 2019 (COVID-19) pandemic has urged many governments in the world to enforce a strict lockdown where all nonessential businesses are closed and citizens are ordered to stay at home. One of the consequences of this policy is a significant change in electricity consumption patterns. Since load forecasting models rely on calendar or meteorological information and are trained on historical data, they fail to capture the significant break caused by the lockdown and have exhibited poor performances since the beginning of the pandemic. In this paper we introduce two methods to adapt generalized additive models, alleviating the aforementioned issue. Using Kalman filters and fine-tuning allows to adapt quickly to new electricity consumption patterns without requiring exogenous information. The proposed methods are applied to forecast the electricity demand during the French lockdown period, where they demonstrate their ability to significantly reduce prediction errors compared to traditional models. Finally, expert aggregation is used to leverage the specificities of each predictions and enhance results even further.

Paradigmatic enhancement of stem vowels in regular English inflected verb forms

Many theories of word structure in linguistics and morphological processing in cognitive psychology are grounded in a compositional perspective on the (mental) lexicon in which complex words are built up during speech production from sublexical elements such as morphemes, stems, and exponents. When combined with the hypothesis that storage in the lexicon is restricted to the irregular, the prediction follows that properties specific to regular inflected words cannot co-determine the phonetic realization of these inflected words. This study shows that the stem vowels of regular English inflected verb forms that are more frequent in their paradigm are produced with more enhanced articulatory gestures in the midsaggital plane, challenging compositional models of lexical processing. The effect of paradigmatic probability dovetails well with the Paradigmatic Enhancement Hypothesis and is consistent with a growing body of research indicating that the whole is more than its parts.

Relationships between renal function variations and relative changes in cardiac troponin T concentrations based on quantile generalized additive models (qgam)

OBJECTIVES

The relationship between high-sensitive cardiac troponin T concentration (hs-cTnT) and renal markers levels is known. However, the extent to which their variations are associated remains to be explored. Objective: model the relationship between relative changes in hs-cTnT (Δhs-cTnT) and variations in creatinine (Δcre) or estimated glomerular filtration rate (ΔeGFR), using a quantile generalized additive model (qgam).

METHODS

Concomitant plasma Δhs-cTnT and Δcre from patients aged 18-100 years, selected with a time variation (Δtime) of 3 h-7 days, were collected over a 5.8-year period. Relationships between Δhs-cTnT and covariates Δcre (A) or ΔeGFR (B), including age, Δtime, hour of blood sampling (HSB) and covariates interactions were fitted using qgam.

RESULTS

On the whole (n=106567), Δhs-cTnT was mainly associated with Δcre, in a positive and nonlinear way (-21, -6, +5, +20, +55% for -50, -20, +20, +50, +100%, respectively), but to a lesser extent with age (min -9%, max +2%), Δtime (min -4%, max +8%), and HSB (min -5%, max +7%). Δhs-cTnT was negatively associated with ΔeGFR (+46, +7, -5, -11, -20% for -50, -20, +20, +50, +100%, respectively). Classifying Δhs-cTnT as consistent or not with myocardial injury based on recommendations, an interpretation of Δhs-cTnT adjusted for model A or B led to statistically significant but small diagnostic discrepancies (<2%), as compared to an interpretation based on Δhs-cTnT only.

CONCLUSIONS

From a laboratory and statistical standpoint, considering renal function variations when interpreting relative changes in cardiac troponin T has a minor impact on the diagnosis rate of myocardial injury.

Point and Interval Forecasting of Zonal Electricity Prices and Demand Using Heteroscedastic Models: The IPEX Case

Since the electricity market liberalisation of the mid-1990s, forecasting energy demand and prices in competitive markets has become of primary importance for energy suppliers, market regulators and policy makers. In this paper, we propose a non-parametric model to obtain point and interval predictions of price and demand. It does not require any parametric assumption on the distribution of the error term or on the functional relationships linking the response variable to covariates. The assumed location–scale model provides a non-parametric estimation of the conditional mean and of the conditional variance by means of a Generalised Additive Model. Interval forecasts, at any given confidence level, are then obtained using a further non-parametric estimation of the innovation’s quantile. Since both the conditional mean and the conditional variance of the response variable are non-linear functions of covariates depending on calendar factors, renewable energy productions and other market variables, the resulting model is very flexible. It easily adapts to market conditions as well as to the non-linear characteristics of demand, supply and prices. An application to hourly data for the Italian electricity market, over the period 2015–2019 period, shows the one-day-ahead forecasting performance of the model for zonal electricity prices and level of demand.

GFR in Healthy Aging: an Individual Participant Data Meta-Analysis of Iohexol Clearance in European Population-Based Cohorts

BACKGROUND

Population mean GFR is lower in older age, but it is unknown whether healthy aging is associated with preserved rather than lower GFR in some individuals.

METHODS

We investigated the cross-sectional association between measured GFR, age, and health in persons aged 50-97 years in the general population through a meta-analysis of iohexol clearance measurements in three large European population-based cohorts. We defined a healthy person as having no major chronic disease or risk factors for CKD and all others as unhealthy. We used a generalized additive model to study GFR distribution by age according to health status.

RESULTS

There were 935 (22%) GFR measurements in persons who were healthy and 3274 (78%) in persons who were unhealthy. The mean GFR was lower in older age by -0.72 ml/min per 1.73 m² per year (95% confidence interval [95% CI], -0.96 to -0.48) for men who were healthy versus -1.03 ml/min per 1.73 m² per year (95% CI, -1.25 to -0.80) for men who were unhealthy, and by -0.92 ml/min per 1.73 m² per year (95% CI, -1.14 to -0.70) for women who were healthy versus -1.22 ml/min per 1.73 m² per year (95% CI, -1.43 to -1.02) for women who were unhealthy. For healthy and unhealthy people of both sexes, both the 97.5th and 2.5th GFR percentiles exhibited a negative linear association with age.

CONCLUSIONS

Healthy aging is associated with a higher mean GFR compared with unhealthy aging. However, both the mean and 97.5 percentiles of the GFR distribution are lower in older persons who are healthy than in middle-aged persons who are healthy. This suggests that healthy aging is not associated with preserved GFR in old age.

Short-Term Wind Speed Forecasting Using Statistical and Machine Learning Methods

Wind offers an environmentally sustainable energy resource that has seen increasing global adoption in recent years. However, its intermittent, unstable and stochastic nature hampers its representation among other renewable energy sources. This work addresses the forecasting of wind speed, a primary input needed for wind energy generation, using data obtained from the South African Wind Atlas Project. Forecasting is carried out on a two days ahead time horizon. We investigate the predictive performance of artificial neural networks (ANN) trained with Bayesian regularisation, decision trees based stochastic gradient boosting (SGB) and generalised additive models (GAMs). The results of the comparative analysis suggest that ANN displays superior predictive performance based on root mean square error (RMSE). In contrast, SGB shows outperformance in terms of mean average error (MAE) and the related mean average percentage error (MAPE). A further comparison of two forecast combination methods involving the linear and additive quantile regression averaging show the latter forecast combination method as yielding lower prediction accuracy. The additive quantile regression averaging based prediction intervals also show outperformance in terms of validity, reliability, quality and accuracy. Interval combination methods show the median method as better than its pure average counterpart. Point forecasts combination and interval forecasting methods are found to improve forecast performance.

Forest Inventory with Long Range and High-Speed Personal Laser Scanning (PLS) and Simultaneous Localization and Mapping (SLAM) Technology

The use of new and modern sensors in forest inventory has become increasingly efficient. Nevertheless, the majority of forest inventory data are still collected manually, as part of field surveys. The reason for this is the sometimes time-consuming and incomplete data acquisition with static terrestrial laser scanning (TLS). The use of personal laser scanning (PLS) can reduce these disadvantages. In this study, we assess a new personal laser scanner and compare it with a TLS approach for the estimation of tree position and diameter in a wide range of forest types and structures. Traditionally collected forest inventory data are used as reference. A new density-based algorithm for position finding and diameter estimation is developed. In addition, several methods for diameter fitting are compared. For circular sample plots with a maximum radius of 20 m and lower diameter at breast height (dbh) threshold of 5 cm, tree mapping showed a detection of 96% for PLS and 78.5% for TLS. Using plot radii of 20 m, 15 m, and 10 m, as well as a lower dbh threshold of 10 cm, the respective detection rates for PLS were 98.76%, 98.95%, and 99.48%, while those for TLS were considerably lower (86.32%, 93.81%, and 98.35%, respectively), especially for larger sample plots. The root mean square error (RMSE) of the best dbh measurement was 2.32 cm (12.01%) for PLS and 2.55 cm (13.19%) for TLS. The highest precision of PLS and TLS, in terms of bias, were 0.21 cm (1.09%) and −0.74 cm (−3.83%), respectively. The data acquisition time for PLS took approximately 10.96 min per sample plot, 4.7 times faster than that for TLS. We conclude that the proposed PLS method is capable of efficient data capture and can detect the largest number of trees with a sufficient dbh accuracy.

Exploring and Exploiting Uncertainty: Statistical Learning Ability Affects How We Learn to Process Language Along Multiple Dimensions of Experience

While the effects of pattern learning on language processing are well known, the way in which pattern learning shapes exploratory behavior has long gone unnoticed. We report on the way in which individual differences in statistical pattern learning affect performance in the domain of language along multiple dimensions. Analyzing data from healthy monolingual adults' performance on a serial reaction time task and a self-paced reading task, we show how individual differences in statistical pattern learning are reflected in readers' knowledge of linguistic co-occurrence patterns and in their exploration and exploitation of content-specific and task-general information. First, we investigated the extent to which an individual's pattern learning correlates with his or her sensitivity to systematic morphological and syntactic co-occurrences, as evidenced while reading authentic sentences. We found that the stream of morphological and syntactic information has a more pronounced effect on the reading speed of, as we will label them, content-sensitive learners in that the more probable the co-occurrence pattern, the faster their reading of that pattern will be. Next, we investigated how differences in pattern learning are reflected in the ways in which individuals approach the reading task itself and adapt to it. Casting this relation in terms of exploration/exploitation strategies, known from Reinforcement Learning, we conclude that content-sensitive learners are also more likely to initially probe (explore) a wider range of directly relevant patterns, which they can later use (exploit) to optimize their reading performance further. By affecting exploratory behavior, pattern learning influences the information that is gathered and becomes available for exploitation, thereby increasing the effect pattern learning has on language cognition.