This is a population-based longitudinal cohort study of SA and DP among white-collar workers employed in the private trade and retail sector during the years 2010-2016.
Data and study population
We used data from three national Swedish administrative registers linked at the individual level through the use of the personal identity number  (PIN, a unique ten-digit number assigned to all Swedish residents): (1) Longitudinal Integration Database for Health Insurance and Labor Market Studies (LISA by Swedish acronym) held by Statistics Sweden: to identify the source population and to obtain information on age, gender, country of birth, type of living area, marital status, level of education, income, size of workplace, occupational code according to Swedish standard classification of occupations (SSYK by Swedish acronym, the Swedish version of the international classification of occupations, ISCO), sector and branch of activity according to the Swedish standard of industrial classification (SNI by Swedish acronym) in 2012-2016; (2) Microdata for analysis of the Social Insurance Database (MiDAS) held by the Social Insurance Agency: for information on periods of AS > 14 days and DP (dates, duration (25, 50, 75 or 100% full time) for 2010-2016); (3) the register of causes of death maintained by the National Board of Health and Welfare: for information on the date of death.
The study population consisted of all persons aged 18 to 67 and registered as residing in Sweden on December 31, 2011 and December 31, 2012, with an occupational code according to SSYK indicating a white-collar occupation, employed in a private sector. business in the commerce and retail sector according to the SNI, and in 2012 had income from work, parental benefits and/or AS/PD that amounted to at least 7,920 SEK (i.e. 75% of the income level needed to qualify for SA benefits from the Social Insurance Agency). The limit of 75% of the minimum income to qualify for SA benefits has been set since, in many cases, SA benefits represent around 75% of working income; without this adjustment, people with low income and long-term AS might have fallen below the minimum income level to be included in the study . Those who were employed in the public sector, self-employed, or had a full-time PA through 2012 were excluded, and those who died or emigrated before January 1, 2017 were excluded. The final study population was 189,321 individuals.
Public health insurance in Sweden
All persons living in Sweden aged ≥ 16 years with income from work or unemployment benefits, who due to illness or injury have a reduced ability to work, are covered by social insurance national public service providing SA services. After a first qualifying day, the employer pays sickness benefit for the first 14 days of an SA period, then SA benefits are paid by the Social Insurance Agency. Self-employed have more qualifying days. Unemployed persons receive SA benefits from the Social Insurance Agency after the first day of qualification. A medical certificate is required after 7 days of self-certification. In this study, data on SA with benefits from the Social Insurance Agency were used. Periods of SA
SA and DP can be granted on a part-time or full-time basis (25, 50, 75 or 100% of normal working hours). SA benefits cover 80% and DP benefits 64% of loss of earnings, both up to a certain level [28, 29]. We used the SA/PD net days so that the partial SA/PD days were combined, for example, two days of part-time absence for 50% were added to one net day. The first 14 days of the SA periods were counted as being of the same duration as day 15 for purposes of calculating net days. In all analyses, SA and PD days were combined. The average number of SA/PD net days/calendar year was calculated for each individual .
Unless otherwise specified, we used information on the following variables from 2012: Sex: woman or man; Age: 18–24, 25–34, 35–44, 45–54, 55–64, or 65–67; Native country: Sweden, other Nordic country, other EU25 or rest of the world, including missing persons; Level of education: compulsory school (≤ 9 years old or missing), high school (10-12 years old) or college/university (≥ 13 years old); Family situation: married/cohabiting with children Type of living space: large city (Stockholm, Gothenburg or Malmö), medium city > 90,000 inhabitants less than 30 km from the city center), or small city/rural (Work Requirements/Control: based on an occupational exposure matrix derived from Statistics Sweden’s Swedish Work Environment Survey classified according to tertiles into the following nine groups: high demands/high control, high demands/medium control, high demands/low control, medium demands/high control, medium demands/medium control, medium demands/low control, low demands/ high control, low requirements/medium control, low requirements/low control (for more information on the occupational exposure matrix and categorization, see Norberg et al. ); Having changed sector of activity in 2016, based on the SNI, classified into the following six groups: trade and retail (no change), manufacturing, services, transport, construction and installation, care and education, or hospitality; Having changed sectors in 2016: State, region, municipality, private company (no change), or other; Change of occupation (based on SSYK code) 2012-2016: no change or change within sub-major group (according to SSYK classifications), change of sub-major group within major group, move to higher major group (e.g., from major group 2 Professionals to major group 1 Managers) and downgrading Major Group (eg, from Major Group 1 Managers to Major Group 2 Professionals).
We used a procedure known as group-based trajectory modeling (Proc traj in SAS) to identify possible trajectories of mean days SA/DP/year over the seven years 2010-2016. Those who had 0 days during the whole of the seven years studied were classified in a separate group, since their trajectory was directly observable. We then fit a group-based trajectory model using only those who had SA and/or PD for at least one of the seven years.
We determined the best-fitting model related to the number of trajectories using the Bayesian Information Criterion (BIC) , where values close to 0 indicate a better fit. The BIC value of a trajectory model was compared to an n-1 trajectory model, and if the BIC value was smaller, the n-cluster model was considered a better fit than the n-1 trajectory model. In order to keep the number of trajectories within interpretable scope, a requirement of a minimum of 5% of the study population for the smallest trajectory was introduced (of the population used to build the model, i.e. i.e. excluding those that had 0 days each year, due to the size of this group relative to all others) . We also used the method described by Coté et al.  to assess the fit of the model, the average probability of belonging to the trajectory must be greater than 0.70 in all the trajectories. The final model was the one that met all criteria of decreasing BIC value, at least 5% of the study population in each trajectory, and an average probability ≥ 0.70 of belonging to the trajectory in all trajectories.
In the following type of analysis, once the optimal number of trajectories had been calculated, the individual probabilities of belonging to a particular trajectory were estimated using a multinomial logit function. Individuals were assigned to the trajectory to which they had the greatest probability of belonging.
We used multinomial logistic regression to determine the association between each of the variables and membership in each trajectory with crude and mutually adjusted odds ratios (ORs) and 95% confidence intervals (CIs). The reference groups for each respective variable were chosen as the largest groups, or the groups expected to have the lowest risk of belonging to trajectories other than “No SA/DP” (since ORs > 1 are easier to interpret).
Since those who are ≥ 65 years old are not eligible for DP benefits (but may still be eligible for SA benefits), we performed sensitivity analyzes where we excluded all those who were ≥ 61 years old in 2012, and therefore were ≥ 65 years old for at least one of the follow-up years.
All methods were performed in accordance with current guidelines and regulations. The project has been approved by the Regional Ethical Review Board in Stockholm. All analyzes were performed in SAS v 9.4.