Page 34 - IJPS-11-4

International Journal of
Population Studies Stress in early adulthood

reported a high-stress score in one of the three rounds of conflicts between work and private life, for example, if
questionnaires; and (iii) individuals who reported high- work takes too much energy and time from private life
stress levels in two or all three rounds of questionnaires. (NFA, 2021). Based on the five response categories (scores
Missing data were handled using multiple imputations. of 0, 25, 50, 75, or 100), a mean sum score between 0 and
th
The accumulated stress scale was based on the imputation 100 was calculated and dichotomized at the 75 percentile
of stress scale scores at ages 15, 18, and 21. of the scale, with scores >50 indicating a high degree of
work-life conflict.
The parental socioeconomic background was measured
using the following two perspectives: equivalized Work justice was measured by four items from the
household income and the highest attained level of DPQ (NFA, 2021) asking about justice in relation to, for
education of the parents, according to Statistics Denmark, example, how conflicts are solved, decisions are made,
using information from the Central Office of Civil or how you are treated at the workplace. Based on the
Registration (Pedersen, 2011). The equivalized household five response categories (scores of 0, 25, 50, 75, or 100),
income was determined by calculating the average income a mean sum score between 0 and 100 was calculated and
th
over a 4-year period when the children were between 7 and dichotomized at the 25 percentile of the scale, with scores
10 years old. This period was chosen as it reflects a time <56 indicating low work justice.
when the children were still living at home and financially Worries about losing one’s job were measured by three
dependent on their parents. Equivalized income takes items from the DPQ (NFA, 2021), asking about worries
into account the number of household members and is about getting fired, troubles in finding a new job, or worries
calculated as a weighted average. The weight assigned is 1.0 about being moved to a new job against your will. Based on
for the first adult, 0.5 for the second adult and individuals the five response categories (scores of 0, 25, 50, 75, or 100),
above age 14, and 0.3 for children below age 14 (Baadsgaard a mean sum score between 0 and 100 was calculated and
& Quitzau, 2011). Household income was split into tertiles: dichotomized at the 75 percentile of the scale, with scores
th
low (<11,072 €), medium (11,072 – 13,489 €), and high >33 indicating many worries about losing one’s job.
(>13,489 €) yearly equivalized average income.
2.3. Analysis
The highest level of education achieved by the parents,
based on information provided by both parents, was The final set of data had a small number of missing values
collected in 2004. In addition, the participants’ own highest in both outcome variables (perceived stress at age 28,
educational attainment was obtained at age 28 in 2017. n = 17 [<0.01%]; work stress at age 28, n = 9 [<0.01%]).
In both cases, the education levels were categorized into Among exposures and covariates, the number of missing
primary level (<10 years), secondary level (10 – 13 years), values ranged from 8 (<0.01%) for psychosocial work
and tertiary level (>13 years) of education (Jensen & climate to 365 (29%) for perceived stress at age 21.
Rasmussen, 2011). Information about gender also came Initially, a multiple imputation chained model was
from registers (Pedersen, 2011). Information about the time performed. This model utilized 100 imputations to impute
since the last educational completion was likewise derived missing values for various variables. These variables
from the educational registers (Jensen & Rasmussen, 2011). included five stress scale measures spanning from age 15
Temporary employment at age 28 was assessed by an item – 28, two socioeconomic position measures, one measure
asking, “In your main occupation, you are,” with the two of temporary employment, and four psychosocial work
response options “permanently employed” or “temporarily environment measures. Chains were constructed, utilizing
employed” from the national work environment cohort either regress logit or ologit, depending on the nature of
(NFA, 2020). Questionnaire information about the the data.
following four psychosocial working conditions at age 28 Certain variables had complete information, including
was used: gender, highest completed education, and time since
educational completion. In addition to the aforementioned
Psychosocial work climate was collected by the item:
“how is the atmosphere and the mental work climate usually variables, other variables were included in the imputation
process: (i) self-reported information about depressive
in your workplace?” from the national work environment symptoms at ages 15, 18, 21, and 28 was included when
cohort (NFA, 2020). The six response categories were imputing the stress measures and (ii) additional measures
dichotomized into good (very good, good, reasonably of work satisfaction, influence, and workload were
good) and bad (not so good, bad, very bad).
included when imputing measures of the psychosocial
Work-life balance was collected by four items from the work environment at age 28. The final estimates were
Danish Psychosocial Questionnaire (DPQ) asking about derived by averaging the sets of estimates obtained from

Volume 11 Issue 4 (2025) 28 https://doi.org/10.36922/ijps.0864