Leonard Henckel
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Leonard Henckel has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Marloes H. Maathuis
Seminar for Statistics, ETH Zürich, Zürich, Switzerland
3 shared publications - Benjamin Stucky
Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland
2 shared publications - Hans-Peter Landolt
University of Zurich, Switzerland
2 shared publications - José Haba-Rubio
Centre D’Investigation et de Recherche Sur le Sommeil (CIRS), Centre Hospitalier Universitaire Vaudois (CHUV), 1011 Lausanne, Switzerland
2 shared publications - Pedro Marques-Vidal
Internal Medicine, Lausanne University Hospital (CHUV) and University of Lausanne Department of Medicine Lausanne
2 shared publications - Benjamin Hollering
Max Planck Institute for Mathematics in the Sciences
1 shared publication
Latest Publications
- 1 February 2026
- Published by Bernoulli Society for Mathematical Statistics and Probability in Bernoulli
- Vol. 32 (1) , 638-663
Faithlessness in Gaussian graphical models
The implication problem for conditional independence (CI) asks whether the fact that a probability distribution obeys a given finite set of CI relations implies that a further CI statement also holds in this distribution. This problem has a long and fascinating history, cumulating in positive results about implications now known as the semigraphoid axioms as well as impossibility results about a general finite characterization of CI implications. Motivated by violation of faithfulness assumptions in causal discovery, we study the implication problem in the special setting where the CI relations are obtained from a directed acyclic graphical (DAG) model along with one additional CI statement. Focusing on the Gaussian case, we show that this implication problem reduces to a principal ideal membership problem. Using this we derive necessary and sufficient combinatorial criteria for when the set of faithless distributions decomposes into graphical components. Moreover, prompted by the relevance of strong faithfulness in statistical guarantees for causal discovery algorithms, we give a graphical solution for an approximate CI implication problem, in which we ask whether small values of one additional partial correlation entail small values for yet a further partial correlation.
- 1 February 2026
- Published by Bernoulli Society for Mathematical Statistics and Probability in Bernoulli
- Vol. 32 (1) , 638-663
Faithlessness in Gaussian graphical models
The implication problem for conditional independence (CI) asks whether the fact that a probability distribution obeys a given finite set of CI relations implies that a further CI statement also holds in this distribution. This problem has a long and fascinating history, cumulating in positive results about implications now known as the semigraphoid axioms as well as impossibility results about a general finite characterization of CI implications. Motivated by violation of faithfulness assumptions in causal discovery, we study the implication problem in the special setting where the CI relations are obtained from a directed acyclic graphical (DAG) model along with one additional CI statement. Focusing on the Gaussian case, we show that this implication problem reduces to a principal ideal membership problem. Using this we derive necessary and sufficient combinatorial criteria for when the set of faithless distributions decomposes into graphical components. Moreover, prompted by the relevance of strong faithfulness in statistical guarantees for causal discovery algorithms, we give a graphical solution for an approximate CI implication problem, in which we ask whether small values of one additional partial correlation entail small values for yet a further partial correlation.
- Open Access
- 18 October 2025
- Published by SAGE Publications in Journal of Psychopharmacology
- Vol. 39 (12) , 1437-1448
Community-based causal evidence that high habitual caffeine consumption alters distinct polysomnography-derived sleep variables
Background: Controlled laboratory studies demonstrate that caffeine acutely impairs sleep quality. However, the impact of daily caffeine intake, which is common in society, on community-derived physiological sleep measures is unknown. Aims: Because good quality sleep is important for general health and well-being, we explored causal effects of habitual caffeine consumption on objective and subjective sleep variables collected at home. Methods: We used dedicated, two-sample Mendelian Randomization (MR) and causal matching methods, including MR-Egger, inverse variance weighting, and weighted median, to analyze large community-based datasets taken from the UK Biobank ( n = 485,511) and the HypnoLaus ( n = 1702) cohorts. Results: While self-rated sleep quality and morningness–eveningness did not differ, all statistical models revealed that four or more caffeinated beverages per day shorten total sleep time when compared to fewer caffeine containing drinks per day. The estimated reductions in sleep length varied from 11 to 229 minutes. Intriguingly, consistent with the homeostatic facet of sleep-wake regulation, the shorter sleep in high habitual caffeine consumers was characterized by increased non-rapid-eye movement sleep depth as measured by all-night electrical brain activity. Conclusions: The data show that high habitual caffeine intake alters the characteristics of sleep in the general population, while sparing the major physiological principles of sleep-wake regulation possibly due to adaptation.
- 18 October 2025
- Published by SAGE Publications in Journal of Psychopharmacology
- Vol. 39 (12) , 1437-1448
Open AccessCommunity-based causal evidence that high habitual caffeine consumption alters distinct polysomnography-derived sleep variables
Background: Controlled laboratory studies demonstrate that caffeine acutely impairs sleep quality. However, the impact of daily caffeine intake, which is common in society, on community-derived physiological sleep measures is unknown. Aims: Because good quality sleep is important for general health and well-being, we explored causal effects of habitual caffeine consumption on objective and subjective sleep variables collected at home. Methods: We used dedicated, two-sample Mendelian Randomization (MR) and causal matching methods, including MR-Egger, inverse variance weighting, and weighted median, to analyze large community-based datasets taken from the UK Biobank ( n = 485,511) and the HypnoLaus ( n = 1702) cohorts. Results: While self-rated sleep quality and morningness–eveningness did not differ, all statistical models revealed that four or more caffeinated beverages per day shorten total sleep time when compared to fewer caffeine containing drinks per day. The estimated reductions in sleep length varied from 11 to 229 minutes. Intriguingly, consistent with the homeostatic facet of sleep-wake regulation, the shorter sleep in high habitual caffeine consumers was characterized by increased non-rapid-eye movement sleep depth as measured by all-night electrical brain activity. Conclusions: The data show that high habitual caffeine intake alters the characteristics of sleep in the general population, while sparing the major physiological principles of sleep-wake regulation possibly due to adaptation.
- Open Access
Community-based causal evidence that high habitual caffeine consumption alters distinct polysomnography-derived sleep variables
Caffeine is the most widely consumed psychoactive substance in the World and acutely impairs sleep quality when studied in the laboratory. The repercussions on sleep of chronic caffeine intake, which is common in society, are unknown. Because adequate sleep is important for general health, we explored causal effects of habitual caffeine intake on objective and subjective sleep quality with two-sample Mendelian Randomization and matching methods. We analyzed large community-based datasets taken from the UK Biobank (n=485,511) and HypnoLaus (n=1,702) cohorts. We consistently show that intake of four or more caffeinated beverages daily shortens total sleep time compared to three or fewer caffeinated drinks per day. While we found shorter sleep with all models, the estimated reductions varied from 11-229 minutes, likely due to modest instrument strengths. A concurrent increase in sleep intensity indicates at least partial compensation. The findings suggest that high daily caffeine intake causally alters sleep patterns in the general population.
Open AccessCommunity-based causal evidence that high habitual caffeine consumption alters distinct polysomnography-derived sleep variables
Caffeine is the most widely consumed psychoactive substance in the World and acutely impairs sleep quality when studied in the laboratory. The repercussions on sleep of chronic caffeine intake, which is common in society, are unknown. Because adequate sleep is important for general health, we explored causal effects of habitual caffeine intake on objective and subjective sleep quality with two-sample Mendelian Randomization and matching methods. We analyzed large community-based datasets taken from the UK Biobank (n=485,511) and HypnoLaus (n=1,702) cohorts. We consistently show that intake of four or more caffeinated beverages daily shortens total sleep time compared to three or fewer caffeinated drinks per day. While we found shorter sleep with all models, the estimated reductions varied from 11-229 minutes, likely due to modest instrument strengths. A concurrent increase in sleep intensity indicates at least partial compensation. The findings suggest that high daily caffeine intake causally alters sleep patterns in the general population.
- Open Access
- 3 November 2023
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 111 (3) , 771-788
Graphical tools for selecting conditional instrumental sets
Summary: We consider the efficient estimation of total causal effects in the presence of unmeasured confounding using conditional instrumental sets. Specifically, we consider the two-stage least-squares estimator in the setting of a linear structural equation model with correlated errors that is compatible with a known acyclic directed mixed graph. To set the stage for our results, we characterize the class of linearly valid conditional instrumental sets that yield consistent two-stage least-squares estimators for the target total effect and derive a new asymptotic variance formula for these estimators. Equipped with these results, we provide three graphical tools for selecting more efficient linearly valid conditional instrumental sets: first, a graphical criterion that, for certain pairs of linearly valid conditional instrumental sets, identifies which of the two corresponding estimators has the smaller asymptotic variance second, an algorithm that greedily adds covariates that reduce the asymptotic variance to a given linearly valid conditional instrumental set and, third, a linearly valid conditional instrumental set for which the corresponding estimator has the smallest asymptotic variance that can be ensured with a graphical criterion.
- 3 November 2023
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 111 (3) , 771-788
Open AccessGraphical tools for selecting conditional instrumental sets
Summary: We consider the efficient estimation of total causal effects in the presence of unmeasured confounding using conditional instrumental sets. Specifically, we consider the two-stage least-squares estimator in the setting of a linear structural equation model with correlated errors that is compatible with a known acyclic directed mixed graph. To set the stage for our results, we characterize the class of linearly valid conditional instrumental sets that yield consistent two-stage least-squares estimators for the target total effect and derive a new asymptotic variance formula for these estimators. Equipped with these results, we provide three graphical tools for selecting more efficient linearly valid conditional instrumental sets: first, a graphical criterion that, for certain pairs of linearly valid conditional instrumental sets, identifies which of the two corresponding estimators has the smaller asymptotic variance second, an algorithm that greedily adds covariates that reduce the asymptotic variance to a given linearly valid conditional instrumental set and, third, a linearly valid conditional instrumental set for which the corresponding estimator has the smallest asymptotic variance that can be ensured with a graphical criterion.