On 9 January 2026, the Medical Statistics and Ageing group at the LUMC is organizing a mini-symposium on the latest developments in the analysis of longitudinal data for ageing research, titled “Large-scale longitudinal data for ageing research: from classic to novel questions.” Please find the programme below, including abstracts. If you would like to attend, please let us know via the registration form below.

Registration is closed

Please share this announcement with other researchers who may be interested, thank you! We look forward to seeing many of you there!

Mini-symposium “Large-scale longitudinal data for ageing research: from classic to novel questions”

Date: Friday, January 9, 2026
️Location: LUMC (Albinusdreef 2, Leiden), Lecture Hall 2

Program

09:45 – 10:00: Walk-in & Welcome

10:00 – 10:30: First, second and third generation biological clocks: a methodological perspective
(Mar Rodriguez Girondo, Medical Statistics, LUMC)

10:30 – 11:00: Regression with Interval-Censored Covariates: Overview and Application to Detection-Limited Circulating Carotenoids
(Guadalupe Gomez Melis, Polytechnic University of Catalonia, Spain)

11:00 – 11:30: Coffee break

11:30 – 12:00: Intergenerational transmission and family clustering of survival, health, and socioeconomic resources
(Niels van den Berg, Molecular Epidemiology, LUMC)

12:00-12:30: Longitudinal studies of human biological aging – opportunities and challenges
(Sara Hägg, Karolinska Institutet, Sweden)

12:30 – 12:45 Concluding remarks

Abstracts

First-, second- and third- generation biological clocks: a methodological perspective
(Mar Rodriguez Girondo, Medical Statistics, LUMC)

Among individuals of the same chronological age, the rate of aging varies widely. The concept of biological age aims to capture this variability and may provide a more accurate indicator of an individual’s true aging status than chronological age. Identifying reliable biomarkers of biological age is a central goal in current aging research. However, the concept lacks a clear operational definition, resulting in multiple biological age prediction approaches that often rest on untested assumptions. In this presentation, I will first critically examine popular cross-sectional aging clocks, showing that they rely on a common, strong, and untestable assumption. Next, I will introduce a novel methodological framework for conceptualizing and analyzing biological age data, integrating advanced survival analysis techniques within omics-based prediction models. We propose that biological age, as a latent and holistic construct, can be operationalized in terms of observable time-to-event outcomes—such as age-at-onset of age-related diseases or mortality. Finally, I will discuss future directions, highlighting how the simultaneous analysis of detailed age-at-disease-onset profiles from electronic health records offers new opportunities for methodological research in aging, alongside the challenges it presents. Illustrations using data from the UK Biobank and Leiden Longevity Study will be provided.

Regression with Interval-Censored Covariates: Overview and Application to Detection-Limited Circulating Carotenoids
(Guadalupe Gomez Melis, Polytechnic University of Catalonia, Barcelona, Spain)

Interval-censored covariates arise in biomedical research when laboratory measurements fall below detection or quantification limits, leading to incomplete information on clinically relevant exposures. Regression methods that appropriately account for interval-censored covariates remain limited. In this talk I will present an overview of interval censoring and introduce a likelihood-based approach, GELc, for estimating generalized linear models when a key covariate is observed exactly for some individuals and interval-censored for others. The approach incorporates an augmented version of Turnbull’s nonparametric estimator, yielding consistent, asymptotically normal estimates with computable standard errors under mild regularity conditions.

The practical relevance of this framework is illustrated through an observational nutrition study of circulating carotenoids, where detection and quantification limits are intrinsic to biomarker quantification and motivate the need for rigorous modeling of covariates subject to mixed exact and interval-censored measurement.

Intergenerational transmission and family clustering of survival, health, and socioeconomic resources
(Niels van den Berg, Molecular Epidemiology, LUMC)

The genetic component underlying longevity represents key mechanisms contributing to a life-long decreased mortality and morbidity risk. Identifying the mechanisms involved is challenging, mainly because of uncertainty in defining long-lived cases with heritable longevity amongst phenocopies and complex gene x environment interactions. Hence, we investigated the longevity trait and its transmission from one generation to the next. In large-scale family-tree data from Utah (UPDB) and the Netherlands (LINKS), we studied 20,360 unselected families containing index persons, their parents, siblings, spouses, and children, comprising 314,819 individuals. We found strong evidence that longevity is transmitted as a quantitative genetic trait among the top 10% survivors of their birth cohort. The survival advantage amounted to 31% for individuals with top 10% surviving first and second-degree relatives in both databases and across two generations, even in the absence of non-long-lived parents. Subsequently, we developed the Longevity Relatives Count (LRC) score as an instrument to quantify the number of long-lived family members and observed that the survival advantage of study participants increased with each additional long-lived family member. Applying the LRC score to the LLS (Netherlands) and SEDD (Sweden; register data) showed that an increasing number of long-lived ancestors associates with an increasing delay in disease incidence (Fig1). As compared to their partners, members of long-lived families have a delayed onset of medication use, multimorbidity and blood-based profiles indicating improved metabolic health and low inflammation in mid-life. Our results indicate that an increasing number of long-lived ancestors marks a decade of healthspan extension, healthier metabolomics profiles, and can be used for more optimized case definitions. Building on these findings, future work will refine and generalize the LRC score into a broader family-based survival metric, enabling wider application across existing studies and supporting the disentanglement of gene x environment interactions in healthy aging and longevity.

Longitudinal studies of human biological aging – opportunities and challenges
(Sara Hagg, Karolinska Institutet, Stockholm, Sweden)

In my research, I use samples from several longitudinal studies of aging from Sweden to study human biological aging. Specifically, we have assessed repeated levels of different biomarkers of aging, such as DNA methylation levels, in old Swedish twins in the SATSA cohort in up to six waves with follow-up in registries for 30 years. Other biomarkers of aging include telomere length, physiological age, glycan age and functional measures with grip strength, cognitive assessments and frailty. In my talk, I will present results from different projects using these data, with specific focus on opportunities and challenges when studying aging in a longitudinal perspective.