Neighborhood age composition

Data on the distribution of age categories in neighborhoods in the Netherlands were obtained from Statistics Netherlands, which was available on the 1st of January in 2011 [22]. In the Netherlands municipalities are divided into neighborhoods (in Dutch: ‘buurten’) and districts (in Dutch: ‘wijken’). Neighborhoods form the lowest aggregation level and are defined from a building point of view or socio-economic structure whereas districts are a sum of consecutive neighborhoods. Assessing social pathways, such as social cohesion, feeling at home and social participation in a neighborhood, we considered the lowest neighborhood level as the most relevant aggregation level to answer our research question. This aggregation level seems to fit best with the relevant activity space around the homes of older adults. We linked the data on age categories with individual level data from the GLOBE study on ‘neighborhood-codes’ (in Dutch: ‘buurtcodes’). On average, a neighborhood includes approximately 1800–2000 inhabitants. The number of inhabitants of neighborhoods in our study population varied between 80 to 9500 in 2011 [22].

For each neighborhood, the percentage of people in the age categories 0 to 15 years old, 15 to 25 years old, 25 to 45 years old, 45 to 65 years old, and 65 years and older were available. To describe the neighborhood age composition for each respondent, we constructed the Herfindahl-Hirschman Index (HHI); a concentration index indicating the homogeneity of the age composition in a neighborhood. The HHI is a measure of market concentration in economics [23], but can also be used to determine other concentrations such as age concentration. To calculate the HHI, we used the formula below (1) where Si is the proportion of the specific neighborhood age group i in the total population from the neighborhood and N the number of age categories (N = 5). To improve interpretation, we multiplied the total by 100 [24].

The theoretical range from the index runs from 0 to 100, with 0 (minimal homogeneity) representing a neighborhood where everyone has a different age category and 100 (maximal homogeneity) representing a neighborhood where everyone is in the same age category [25]. A disadvantage of the HHI is that no distinction can be made in the composition of age groups (e.g. assuming a high HHI, no distinction can be made between a high percentage of older adults or a high percentage of for example young adults). We therefore also assessed the percentages of specific age groups in a neighborhood to unravel the association between the neighborhood age composition and mental health.