Berlin Aging Study

Initiators and Goals of the Berlin Aging Study

The multidisciplinary Berlin Aging Study (BASE) was conducted by the Research Group on Aging and Societal Development of the Berlin-Brandenburg Academy of Sciences in collaboration with institutes and research centers belonging to the Freie Universität Berlin, Humboldt-Universität zu Berlin, and the Max Planck Institute for Human Development. The study was initially directed by a steering committee consisting of the late Paul B. Baltes (chair), head of the BASE Psychology Unit, Karl Ulrich Mayer (vice-chair), head of the BASE Sociology and Social Policy Unit, Hanfried Helmchen, head of the BASE Psychiatry Unit, and Elisabeth Steinhagen-Thiessen, head of the BASE Internal Medicine and Geriatrics Unit. The project was supported financially by several organizations including the German Federal Ministry for Research and Technology (1989–1991, 13 TA 011 + 13 TA 011/A) and the German Federal Ministry for Family Affairs, Senior Citizens, Women, and Youth (1992–1998, 314-1722-102/9 + 314-1722-102/9a) as well as by the institutions involved. From 2008 to 2011, it was co-sponsored by the Federal Ministry of Education and Research (01UW0808). The Center for Lifespan Psychology at the Max Planck Institute for Human Development has led the project since 2008 and Ulman Lindenberger has meanwhile become the speaker of BASE (see also https://www.mpib-berlin.mpg.de/research/research-centers/lip/projects/aging/base).

BASE was established in 1989 to investigate questions about old and very old age from the joint and collaborative perspectives of four disciplines: psychiatry, psychology, sociology, and geriatrics/internal medicine (see Baltes et al., 1993, 2001). The initial goal of BASE (1990–1993) was to obtain extensive baseline information about old and very old age using a 14-session multidisciplinary Intensive Protocol covering older people’s mental and physical health, psychological functioning, and social and economic situation. The study design called for an age-by-sex stratified heterogeneous (locally representative) sample of 70- to over-100-year-olds. The central goal of the longitudinal study was to follow-up the core sample of participants who completed this Intensive Protocol. Extensive information about the baseline sample selectivity, assessment procedures, and cross-sectional findings are reported in Mayer and Baltes (1996; Baltes & Mayer, 1999; Baltes & Smith, 1997; Lindenberger et al., 1999).

Theoretical Orientations and Prototypical Questions of BASE

During the planning of BASE, the research questions, hypotheses, measures, and analyses were based on four theoretical orientations of gerontological research (see also Baltes et al., 1993, 2001): (1) Differential aging, (2) continuity and discontinuity of aging, (3) plasticity and reserve capacity of older people, and (4) aging as a systemic phenomenon.

The initial three prototypical questions of the first in-depth collection of cross-sectional and retrospective data are shown in the figure. The fourth question led to the continuation of BASE as a longitudinal study. 

Figure 1. Schema of the prototypical questions of the Berlin Aging Study (cf. Baltes et al., 2001)

1. Are individual age differences predictable on the basis of life history data?
2. How large are age differences within domains, and which direction do they take?
3. How are the transdisciplinary relationships across different domains?
4. How do characteristics, processes, and links between them change in the course of aging?

BASE Core Study & Longitudinal Follow-ups: Design and Sample Descriptions

In total, N = 516 individuals from the districts of former West-Berlin participated fully in the Intensive Protocol at the first occasion of measurement (T1, representing 27% of the initial verified parent sample of N = 1908 obtained from the city registry). This baseline or core sample consisted of 258 men and 258 women aged between 70 and 103 years at intake. For age-related comparisons, the sample was divided into six age groups/cohorts (each n = 86): 70–74 years (born 1922–1915), 75–79 years (born 1917–1910), 80-84 years (1913–1905), 85–89 years (born 1908–1900), 90–94 years (1902–1896), and 95–103 years (born 1897–1883). Because of the three-year time period required for testing, there is some overlap in birth cohort membership across adjacent age groups. For most participants, assessment extended over a period of 3 to 5 months (mean: 4.4 months; range: 1.2–15.4 months) and therefore required major commitment albeit most of the sessions were carried out at their place of residence. Table 1 shows the demographic characteristics of the core sample.

An additional N = 412 individuals (22% of the parent sample) only completed a 90-minute multidisciplinary Intake Assessment (IA; [Ersterhebung  (EE)]), which was collected as the first session of the Intensive Protocol (IP).

For ethical reasons, potential participants were informed at the outset about the general format and content of the design and their expected involvement. Those who were incapable of understanding the purpose and conditions of the study or who were evaluated by a psychiatrist or medical doctor as too ill to participate were excluded from the study (see also Nuthmann & Wahl, 2007). Such exclusions concerned about 8% of the verified parent sample.

Table 1. Demographic characteristics of the BASE core sample at the first occasion of measurement (T1).

Longitudinal Follow-ups of the BASE Core Sample

The longitudinal follow-up of BASE participants took two main routes (cf. Figure 2). First, survivors of the N = 516 core sample were recontacted and asked to participate again in BASE. At T2, a single-session repeat of the multidisciplinary Intake Assessment took place. Further, there were repeats of the multidisciplinary Intensive Protocol on six occasions, T3 (1995–1996, reduced to 6 sessions), T4 (1997–1998, 6 sessions), T5 (2000, 3 sessions), and T6 (2004, 3 sessions), T7 (2005–2006, 7 sessions), and T8 (2008–2009, 6 sessions). On each occasion, the multidisciplinary first-session Intake Assessment (IA) was repeated. The second longitudinal strategy involved following the entire initial verified parent sample (N = 1908) for mortality. The last member of the core sample passed away aged 107 in 2019. She had taken part at all eight measurement occasions.

Figure 2. Longitudinal research design of the Berlin Aging Study (BASE).

Selectivity Analyses

Because of the breadth and depth of the Intensive Protocol and the examined age range, the analysis of selectivity effects had high priority. Data collected during the recruitment phase, microcensus material, and mortality information from the city registry were used to assess various selectivity aspects (e.g., sample selectivity, experimental and mortality-related selectivity; see Baltes et al., 2001; Lindenberger et al., 2001). As explained above, the study began with a verified random initial sample of 1908 persons from the city register (Nuthmann & Wahl, 2007). 417 persons refused to provide any information beyond the basic information of the city registry so that the baseline participation rate at the first occasion of measurement was 78% (N = 1491). After direct contact 227 persons (12%) were excluded from BASE out of medical or ethical grounds. Apart from the core sample of 516 persons who completed the whole Intensive Protocol (27% of the initial sample), a further 412 persons (21%) took part in the 90-minute multidisciplinary Intake Assessment and another 336 (18%) participated in a 20-minute short interview. With respect to average functioning (with small effect sizes) and mortality, the core sample (N = 516) was positively selected. However, there were no strong indications of selectivity effects regarding individual heterogeneity or covariation between variables.

Lindenberger et al. (2002) also examined the change of the sample from T1 to T3 and found that except for age and intelligence, experimental selectivity effects regarding means and prevalence rates were generally small. Such selectivity effects are hardly avoidable in longitudinal studies with a sample of old and very old people, and particularly when taking into account the data collection intensity that is typical for BASE. Further selectivity analyses have been reported in many of the BASE publications presenting longitudinal findings (e.g., Ghisletta et al., 2006; see https://www.base-berlin.mpg.de/publications for a complete list of BASE publications).

BASE-II

In 2009, researchers from Berlin and Tübingen initiated the Berlin Aging Study II (BASE-II; Bertram et al., 2014; Delius et al., 2015; Demuth et al., 2019; Gerstorf et al., 2016; cf. https://www.base2.mpg.de). Like BASE, BASE-II was set up as a multidisciplinary and multi-institutional longitudinal study that captures a wide range of different functional domains. Geriatrics and internal medicine as well as immunology, psychology, genetics, sociology, and economics are among the disciplines involved. The recruitment of the BASE-II cohort, which was completed in 2014, resulted in a consolidated baseline sample of 1600 older adults aged 60 to 80 years and of 600 younger adults aged 20 to 35 years (Bertram et al., 2014). Many of the measurement instruments used in BASE were re-applied in BASE-II, allowing cohort comparisons (e.g., Gerstorf et al., 2023). The data are not included here, but can be requested from BASE-II.