Socioeconomic Inequalities in Staff Physical Activity: Evidence from the SHAHWAR Cohort Study in Iran

: 24 Background: This study was aimed to investigate socioeconomic-related inequality in physical 25 activity (PA) among Staff of Medical Sciences University in Shahroud, Iran. 26 Methods: Data were extracted from the first phase of the SHAHWAR cohort study. The 27 Concentration index (CI) and Wagstaff decomposition method were applied to determine 28 socioeconomic-related inequality in PPA and its contributors, respectively. 29 Results: CIs of poor total PA (PTPA) and poor work-related PA (PWRPA) were 0.092 and 0.141, 30 indicating their more concentration among staff of higher socioeconomic groups. Furthermore, the 31 negative CI of poor leisure-time PA (PLTPA) (-0.191) suggests that it disfavors staff from lower 32 socioeconomic groups. While PWRPA, Subjective social status (SSS), socioeconomic status 33 (SES), and gender were positively contributed to the measured inequality in PTPA (65.3%, 37.9%, 34 18.6%, and 16.6%, respectively), residence in urban areas and PLTPA have negative contributions 35 (64% and 27%, respectively). Similarly, SES, marital status, and residence in urban areas 36 positively contributed to the inequality of PLTPA by 53.4%, 36.2%, and 23.5% respectively. 37 Whereas, gender had the most negative contribution by 28.7%. Residence in urban areas, SES, 38 SSS, and shift work were all positively contributed to the measured inequality in PWRPA (its more 39 concentration among staff from high SES groups) by 28.4%, 20%, 15.2%, and 12.7%, respectively, 40 while the opposite is true for gender by the contribution of 13.3%. 41 Conclusions: Different patterns of PPA inequality revealed that health promotion programs should 42 aim to educate and support higher SES staff to increase their PA in workplace, leisure time, and 43 transportation, and lower SES staff to increase their leisure-time PA.

further prevalent among the lower socioeconomic groups, [7,10,11] leisure time PA was more 58 prevalent among people with higher SES. [10,12] 59 Governments often implement health promotion programs to increase public health. In this regard, 60 the workplace of people is one of the appropriate settings for the implementation of these 61 programs. [13,14] Because employed people spend a significant proportion of their time at work.

62
[10] Moreover, the return on investment on employees' PA is high [15] because of its favorable 63 effects on productivity, absenteeism, and healthcare costs. Designing cost-effective interventions 64 requires analyzing the current situation of PA distribution among different socioeconomic groups 65 and understanding its contributing factors. In Iran, some studies have been conducted on the impact 66 of socioeconomic status on PA in the general population. However, the employees' PA has not 67 been the focus of these studies. [16][17][18] This study was designed to provide further evidence for Source of data and variables 72 In the present study, data were extracted from the first phase of the SHAHWAR (SHAhroud 73 Healthcare Workers Associated Research) cohort study which was started from 2, OCT, 2019 in 74 Shahroud, located in the Northeast of Iran. SHAHWAR is a subset of PERSIAN Cohort study [19] 75 and focuses on the health of employees. In this cohort study, data from 1178 personnel of Shahroud 76 University of Medical Sciences was prospectively collected after obtaining informed consent.

77
After cleaning data and excluding subjects with missing observations, a total of 1157 staff were 78 included in the analysis. The SHAHWAR cohort study was approved by the Ethics Committee of 79 Shahroud University of Medical Sciences (IR.SHMU. REC.1397.033). 80 In the SHAHWAR cohort study, data for participants' PA were measured using the Persian version were categorized as having poor physical activity in total and in each domain of PA if their activity 88 was less than its median level. The medians (interquartile ranges) for leisure time PA, domestic 89 and gardening activities, work-related PA, transport-related PA, and total PA were 3.3 (0, 12.6), 6 90 (2.3, 14), 20.8 (5, 63.4), 30.9 (20, 47.4), and 89.5 (53.6, 133.1) METs-hour/week, respectively. The principal component analysis (PCA) was used to construct an index of socioeconomic status 92 (SES) for participants.
[20] SES index is derived from a factor analysis of preliminary variables 93 that included: household assets and properties, entertainment and travel related variables, 94 education and access to information and occupational variables. SES scores obtained from PCA 95 were used to classify participants into the five SES quintiles from the lowest (1st quintile) to the 96 highest (5th quintile).

97
One another variable which was measured in this study was subjective social status (SSS); the 98 social class that staff consider to belong it (lowest, low, medium, high, highest percentage of PPA on the Y-axis against the cumulative percentage of SES score on the X-axis.

106
The CI is calculated as twice the area between the CC and the line of equality. The value of CI 107 ranges from -1 to +1. If CC lies above (below) the line of equality, the CI takes a negative 108 (positive) value, indicating that PPA was more concentrated among staff with lower (higher) SES.

109
When the CC crosses the line of equality, the CI equals zero which implies that PPA was equally 110 distributed among staff from different socioeconomic strata. The conventional CIs of the PPA as 111 a whole and in its four domains were calculated as follows:  were more likely to be from higher SES groups, constitute 78.6% of the total university staff. who live in urban areas were less physically active in their work place than their rural counterparts.

234
As it is explained before, difference in PA environment and social and cultural factors could be 235 the main reasons for the less PLTPA among urban resident staff compared to their rural 236 counterparts.

237
Our study indicated that SES was accounted for about one-fifth (19.9%) and more than half 238 (53.4%) of the measured socioeconomic inequality in WRPPA and PLTPA, respectively. Staff 239 from the higher socioeconomic groups usually work in jobs that require less PA due to their higher 240 education, more work experience or a more stable job position. Similar results was also seen in 241 other studies. [7,11,25]

319
The datasets used and/or analyzed during the current study are available from the corresponding 320 author on reasonable request.

321
Competing interests 322 The authors declare that they have no competing interests.