Kouvonen A, Mänty M, Lallukka T, et al. Changes in psychosocial and physical working conditions and psychotropic medication in ageing public sector employees: a record-linkage follow-up study. BMJ Open 2017;7:e015573. doi: 10.1136/bmjopen-2016-015573.
This article was previously published with some errors.
In review of their article the authors found an error in the equation of their measurement of physical work load. In factor analysis, the items indicating physical work load were loaded on Factor one and not on Factor two, as they had erroneously interpreted earlier (Factor two was hazardous exposures).
This error is corrected by re-running all analyses using the correct factor (physical work load). The authors re-conducted the analyses and found only minor differences compared with the original figures given, and it was also found that the error did not alter the conclusions drawn.
The following corrections should be incorporated into any future analysis of the original article:
In Results section, under Abstract, Age and sex-adjusted HR varied from 1.18 to 1.54, instead of 1.66. In addition, the last sentence of the Results in Abstract (Decreased physical work load was associated with lower risk of antidepressant and anxiolytic medications) should be deleted.
In Results section, under the main text, the following should be corrected:
Paragraph 2: :"The association between repeated high physical work load and any psychotropic medication was also statistically significant (HR=1.27, 95% CI: 1.08 to 1.49)." rather than "The association between repeated high physical work load and any psychotropic medication was marginally statistically significant (HR=1.17, 95% CI: 0.98 to 1.39)."
Paragraph 3: "repeated high physical work load HR=1.54 (95% CI: 1.15 to 2.06)" instead of "HR=1.66 (95% CI: 1.24 to 2.23)." In addition, "Increased job control and physical work load were associated with subsequent sedative and hypnotic medication." should read "Increased job control was associated with subsequent sedative and hypnotic medication."
Paragraph 4: The last sentence "Compared with repeatedly high physical work load, decreased physical work load was associated with a lower risk of subsequent antidepressant and anxiolytic medication" should be deleted.
In Discussion, paragraph 6 should have HR=1.54 instead of HR=1.66.
The review led also the three tables, figure 3 and online appendix 2 to be slightly revised. The revised tables are available with this article as tables 1, 2 and 3, figure 3, and appendix 2. The revised figures in tables have been highlighted in red.
Table 1 Distribution of demographics (Phase 1; 2000–2002), working conditions (Phase 1 – Phase 2; 2007) and any psychotropic medication* between Phase 2 and 2013, the Helsinki Health Study, Finland (%) (n=3587)
| No medication N (%) | Medication N (%) | Mean DDDs (SD)† | |
| Sex | |||
| Women | 2034 (71) | 847 (29) | 496.0 (773.3) |
| Men | 545 (77) | 161 (23) | 487.1 (793.0) |
| Age | |||
| 40 | 632 (72) | 247 (28) | 587.5 (949.9) |
| 45 | 636 (68) | 298 (32) | 537.6 (788.6) |
| 50 | 664 (72) | 259 (28) | 429.5 (660.0) |
| 55 | 627 (76) | 196 (24) | 391.2 (627.2) |
| 60 | 20 (71) | 8 (29) | 662.1 (818.6) |
| Job control | |||
| High-High | 1075 (73) | 402 (27) | 453.2 (749.3) |
| High-Low | 358 (71) | 148 (29) | 458.3 (761.1) |
| Low-High | 300 (72) | 117 (28) | 541.5 (793.9) |
| Low-Low | 846 (71) | 341 (29) | 543.1 (806.6) |
| Job demands | |||
| Low-Low | 884 (74) | 304 (26) | 411.6 (647.4) |
| Low-High | 419 (72) | 165 (28) | 476.1 (716.0) |
| High-Low | 432 (72) | 169 (28) | 463.7 (694.2) |
| High-High | 844 (70) | 370 (30) | 585.1 (915.3) |
| Physical work load | |||
| Low-Low | 1763 (73) | 644 (27) | 465.4 (728.0) |
| Low-High | 222 (70) | 93 (30) | 539.6 (798.4) |
| High-Low | 222 (71) | 86 (29) | 588.1 (997.3) |
| High-High | 382 (67) | 185 (33) | 530.0 (810.5) |
*Participants with psychotropic medication purchases in 3 months preceding Phase 2 were excluded.
†Mean of defined daily doses (DDDs) and their standard deviations (SDs) in those who had psychotropic medication purchases during follow-up.
Table 2 Age and sex adjusted hazard ratios (HRs) and their 95% confidence intervals (95% CIs) for associations of changes in working conditions between Phase 1 (2000–2002) and Phase 2 (2007) and subsequent psychotropic medication between Phase 2 and the end of 2013, the Helsinki Health Study, Finland
| Any psychotropic (n=3587) |
Antidepressants (N06A) (n=3660) |
Anxiolytics (N05B) (n=3867) |
Sedatives and hypnotics (N05C) (n=3808) | |||||||||
| N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | |
| Job control | ||||||||||||
| High-High | 1477 (402) | 1.00 | Reference | 1499 (264) | 1.00 | Reference | 1573 (102) | 1.00 | Reference | 1559 (245) | 1.00 | Reference |
| High-Low | 506 (148) | 1.08 | 0.90 to 1.31 | 518 (95) | 1.04 | 0.82 to 1.32 | 542 (43) | 1.22 | 0.86 to 1.74 | 532 (94) | 1.13 | 0.89 to 1.43 |
| Low-High | 417 (117) | 1.03 | 0.84 to 1.27 | 425 (80) | 1.06 | 0.83 to 1.36 | 457 (27) | 0.90 | 0.59 to 1.38 | 449 (91) | 1.33 | 1.04 to 1.69 |
| Low-Low | 1187 (341) | 1.08 | 0.94 to 1.25 | 1218 (244) | 1.18 | 0.99 to 1.41 | 1295 (113) | 1.37 | 1.05 to 1.79 | 1268 (205) | 1.03 | 0.86 to 1.24 |
| Job demands | ||||||||||||
| Low-Low | 1188 (304) | 1.00 | Reference | 1208 (205) | 1.00 | Reference | 1267 (83) | 1.00 | Reference | 1254 (185) | 1.00 | Reference |
| Low-High | 584 (165) | 1.09 | 0.90 to 1.32 | 598 (117) | 1.10 | 0.88 to 1.38 | 632 (45) | 1.05 | 0.73 to 1.52 | 619 (108) | 1.20 | 0.94 to 1.52 |
| High-Low | 601 (169) | 1.09 | 0.91 to 1.32 | 610 (106) | 1.01 | 0.80 to 1.27 | 649 (41) | 0.96 | 0.66 to 1.39 | 639 (110) | 1.16 | 0.92 to 1.47 |
| High-High | 1214 (370) | 1.22 | 1.04 to 1.42 | 1244 (255) | 1.20 | 1.00 to 1.45 | 1319 (116) | 1.33 | 1.00 to 1.76 | 1296 (232) | 1.21 | 0.99 to 1.47 |
| Physical work load | ||||||||||||
| Low-Low | 2407 (644) | 1.00 | Reference | 2451 (427) | 1.00 | Reference | 2575 (170) | 1.00 | Reference | 2546 (412) | 1.00 | Reference |
| Low-High | 315 (93) | 1.13 | 0.91 to 1.41 | 322 (66) | 1.22 | 0.94 to 1.58 | 341 (23) | 1.04 | 0.67 to 1.60 | 337 (58) | 1.07 | 0.82 to 1.41 |
| High-Low | 298 (86) | 1.09 | 0.87 to 1.37 | 305 (62) | 1.17 | 0.90 to 1.52 | 333 (30) | 1.39 | 0.94 to 2.05 | 323 (50) | 0.95 | 0.71 to 1.28 |
| High-High | 567 (185) | 1.27 | 1.08 to 1.49 | 582 (128) | 1.30 | 1.07 to 1.59 | 618 (62) | 1.54 | 1.15 to 2.06 | 602 (115) | 1.20 | 0.97 to 1.47 |
Participants with psychotropic medication purchases in question (the medication groups were not mutually exclusive) in 3 months preceding Phase 2 were excluded.
Table 3 Age and sex adjusted hazard ratios (HRs) and their 95% confidence intervals (95% CIs) for associations of changes in working conditions between Phase 1 (2000–2002) and Phase 2 (2007) and subsequent psychotropic medication between Phase 2 and the end of 2013, the Helsinki Health Study, Finland
| Any psychotropic (n=3587) |
Antidepressants (N06A) (n=3660) |
Anxiolytics (N05B) (n=3867) |
Sedatives and hypnotics (N05C) (n=3808) | |||||||||
| N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | N (cases) | HR | 95% CI | |
| Job control | ||||||||||||
| High-High | 1477 (402) | 0.92 | 0.80 to 1.07 | 1499 (264) | 085 | 0.71 to 1.01 | 1573 (102) | 0.73 | 0.56 to 0.96 | 1559 (245) | 0.97 | 0.81 to 1.17 |
| High-Low | 506 (148) | 1.00 | 0.82 to 1.21 | 518 (95) | 0.88 | 0.70 to 1.12 | 542 (43) | 0.89 | 0.63 to 1.27 | 532 (94) | 1.09 | 0.86 to 1.40 |
| Low-High | 417 (117) | 0.95 | 0.77 to 1.18 | 425 (80) | 0.90 | 0.70 to 1.16 | 457 (27) | 0.66 | 0.43 to 1.00 | 449 (91) | 1.29 | 1.00 to 1.65 |
| Low-Low | 1187 (341) | 1.00 | Reference | 1218 (244) | 1.00 | Reference | 1295 (113) | 1.00 | Reference | 1268 (205) | 1.00 | Reference |
| Job demands | ||||||||||||
| Low-Low | 1188 (304) | 0.82 | 0.71 to 0.96 | 1208 (205) | 0.83 | 0.69 to 1.00 | 1267 (83) | 0.75 | 0.57 to 1.00 | 1254 (185) | 0.83 | 0.68 to 1.01 |
| Low-High | 584 (165) | 0.90 | 0.75 to 1.08 | 598 (117) | 0.91 | 0.73 to 1.14 | 632 (45) | 0.79 | 0.56 to 1.12 | 619 (108) | 0.99 | 0.79 to 1.25 |
| High-Low | 601 (169) | 0.90 | 0.75 to 1.08 | 610 (106) | 0.84 | 0.67 to 1.05 | 649 (41) | 0.72 | 0.50 to 1.03 | 639 (110) | 0.96 | 0.77 to 1.21 |
| High-High | 1214 (370) | 1.00 | Reference | 1244 (255) | 1.00 | Reference | 1319 (116) | 1.00 | Reference | 1296 (232) | 1.00 | Reference |
| Physical work load | ||||||||||||
| Low-Low | 2407 (644) | 0.79 | 0.67 to 0.93 | 2451 (427) | 0.77 | 0.63 to 0.94 | 2575 (170) | 0.65 | 0.49 to 0.87 | 2546 (412) | 0.84 | 0.68 to 1.03 |
| Low-High | 315 (93) | 0.89 | 0.70 to 1.15 | 322 (66) | 0.94 | 0.70 to 1.26 | 341 (23) | 0.67 | 0.42 to 1.08 | 337 (58) | 0.90 | 0.65 to 1.23 |
| High-Low | 298 (86) | 0.86 | 0.67 to 1.12 | 305 (62) | 0.90 | 0.66 to 1.21 | 333 (30) | 0.90 | 0.58 to 1.40 | 323 (50) | 0.80 | 0.57 to 1.11 |
| High-High | 567 (185) | 1.00 | Reference | 582 (128) | 1.00 | Reference | 618 (62) | 1.00 | Reference | 602 (115) | 1.00 | Reference |
Participants with psychotropic medication purchases in question (the medication groups were not mutually exclusive) in 3 months preceding Phase 2 were excluded.
Figure 3 Survival curves for any psychotropic medication by changes in physical working conditions.
Appendix 2 Physical working conditions - age
Appendix 2 Physical working conditions - high-high
Appendix 2 Physical working conditions - high-low
Appendix 2 Physical working conditions - low-high
Appendix 2 Physical working conditions - sex
