Occupational Medicine 2010;60:2128 Published online 10 August 2009 doi:10.

1093/occmed/kqp113

Predictors of the discrepancy between calendar and biological age

Gabriele Freude1, Olga Jakob2, Peter Martus2, Uwe Rose1 and Reingart Seibt3
...................................................................................................................................................................................

Background The rate of ageing can differ considerably between individuals. This might result in major differences between calendar age (CA) and biological age (BA).
...................................................................................................................................................................................

Aims Methods

To identify work- and health-related predictors of the discrepancy between CA and BA. The sample analysed in this study consisted of 371 subjects of different occupational groups (teachers, ofce workers, nursery school teachers and managers). BA was measured with the vitality measuring station, which recorded 45 vitality indicators of physical, mental and social functions. Work ability index, effortreward imbalance and relaxation inability were measured to determine work- and health-related predictors. The greatest discrepancy between CA and BA (9 years) was found for the subgroup of managers, followed by female teachers (5 years). Managers showed also the best results in work ability, the effort reward balance and relaxation ability. By means of multiple regression analysis, particularly mental attitudes and resources towards work, occupational reward and the body fat percentage were identied as relevant predictors for the discrepancy between CA and BA.

...................................................................................................................................................................................

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

...................................................................................................................................................................................

Results

...................................................................................................................................................................................

Conclusions Our study indicates that not only health- but also work-related factors are associated with vitality and BA of employees. We assume that measures focused on promoting of health (healthy diet and physical activities) and improving working conditions (e.g. job satisfaction and social support and stress prevention) may also affect the ageing process positively.
...................................................................................................................................................................................

Key words

Biological age; calendar age; health- and work-related predictors; vitality; work ability.

...................................................................................................................................................................................

Introduction
There is no doubt that the demographic development will affect the age structure of workers in the workplace. All forecasts assume a rise in the average age of employees [1]. Since ageing is associated with changes in functional capacity, suitable measures for maintaining and promoting health and work ability of ageing workers will be of increasing importance for occupational safety and health. Biological age (BA) that refers to the human functional capacity can differ considerably between individuals [24] of the same calendar age (CA). Individuals might be healthier, more vital and biologically younger than might be expected from their CA and vice versa. On
1

Department of Mental Health and Cognitive Capacity, Federal Institute for Occupational Safety and Health, Berlin, Germany. Institute for Biometry and Clinical Epidemiology, Charite, Berlin, Germany.

2 3

Institute and Clinic of Occupational and Social Medicine, Technical University of Dresden, Dresden, Germany. Correspondence to: G. Freude, Department of Mental Health and Cognitive Capacity, Federal Institute for Occupational Safety and Health, Berlin, Germany. e-mail: freude.gabriele@baua.bund.de

the one hand, genetic intrinsic factors that can hardly be inuenced contribute to the interindividual differences; on the other hand, extrinsic factors act on the ageing process, which can affect a persons functional capacity and its changes with age both positively and negatively. Within the conceptual framework of successful ageing [5,6], the effectiveness of extrinsic factors that counteract the intrinsic age-accelerating factors is assumed. Losses in function with advancing age can be avoided or at least restricted [7] by protective extrinsic factors. The inuence of work-related factors on the ageing process has hardly been studied to date. In general, there is no doubt that a high level of engagement with life including productive engagement in occupation [8,9] is a strong predictor of successful ageing. Barnes-Farrell and Pietrowski [10] studied the relation between work stress and workers self-perceptions of the ageing process. The reported presence of stressors and the extent of strain were positively correlated with feeling older than ones CA. Hacker [11] postulated that poor working conditions might induce work-induced ageing, e.g. working conditions that impair physical and mental functionality can lead to premature ageing.

The Author 2009. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

22 OCCUPATIONAL MEDICINE

There are different approaches to evaluate the rate of ageing on basis of BA measurements [4,12]. BA cannot be measured directly but merely approximated using certain vitality and age indicators [4]. Data of BA analysed in this study were obtained using the so-called vitality measuring station (VMS) [13]. This methodological approach is based on the concept of vitality [13], which denes the construct of vitality as an individuals age-typical whole-organism functionality in their unity of physical, mental and social efciency. Efciency in the different functional domains was determined on basis of 45 vitality indicators implemented in the VMS approach. The aim of this study was to examine the discrepancy between CA and BA in employees of different occupational groups, which were assumed to differ in work ability and other work-related factors. We were interested in analysing the BACA discrepancy in relation to work- and health-related factors. Furthermore, we tested which of these factors are relevant predictors for the discrepancy between CA and BA. We hypothesized that protective work- and health-related indicators are related to BA that is lower than CA, i.e. to positive discrepancies between CA and BA.

Methods
Data for this analysis were pooled from three studies [1418] conducted with identical questionnaires [relaxation inability (RI), Work ability index (WAI) and effort reward imbalance (ERI)] and measurements (VMS). The sample consisted of female teachers [17,18], female ofce workers [15], female nursery school teachers [14], male teachers [15,18] and male managers [16]. Except for the nursery teachers, all employees worked .35 h/week. Working 35 h/week is equivalent to working full time in Germany. They work an average of 40.3 h/week with only marginal differences between different groups [19]. The higher proportion of nursery teachers working ,35 h/week was explained by the fact that a relatively high percentage of this professional group (.50%) work part time [20]. In our study, discrepancy between CA and BA as dependent variable is dened by taking CA minus BA. BA was measured with the VMS. This inventory comprises 45 vitality indicators to assess functionality in the cardiopulmonary system (blood pressure, pulse at rest and after workload and tness index after submaximal workload), the musculoskeletal system (muscular strength, speed and coordination capabilities) and of the sense organs (visual/auditory acuity and visual/auditory reaction time). Furthermore, psychological and mental functions (verbal/cognitive response, concentration, exibility, strategy building and memory capacity), physical and psychological complaints and social factors (obligations and leisure

time, behaviour and stress and social competence) were estimated by means of VMS. Values of each indicator were standardized to values for a dened reference population and summarized in a value for BA [12,13]. Health-and work-related determinants were analysed by means of the RI questionnaire [21], the WAI [2224] and the ERI [25,26] questionnaire. Furthermore, body mass index (BMI) and body composition (body fat and active cell mass) were measured as relevant health-related factors. RI was examined using the corresponding subscale from the standardized RI questionnaire [21]. The RI index consists of six items referring to inability to relax from work. The score for RI ranges from 6 to 24 (high RI values indicate a high RI and low values a high relaxation ability). The general cut-off for noticeable values is 19 and for very noticeable values is 21. The WAI describes how well an employee is capable to do his job [24]. It is based on the concept of work ability [2224], which considers that work ability is determined by the interaction between specic work demands, individual health conditions and mental resources. The WAI questionnaire consists of seven subscales referring to different aspects of work ability (WAI 1current work ability compared with the life time best, WAI 2work ability in relation to the physical and mental demands at work, WAI 3current number of diseases or injuries diagnosed by a doctor, WAI 4subjective estimation of work ability impairment due to diseases, WAI 5sick leave during the past 12 months, WAI 6personal prognosis of work ability in the next 2 years and WAI 7mental attitudes and resources towards work). The cumulative index of WAI ranges from 7 to 49 points. It is divided into the following categories: poor (727 points), moderate (2836 points), good (3743 points) and excellent (4449 points) work ability. The ERI at work was measured by the short version of ERI questionnaire with 23 items [25,26]. The subscale effort (range: 630, high values indicate a high effort at work) refers to the perceived time pressure, work interruptions and disturbances, responsibilities at work, overtime, etc. The subscale reward (range: 1155, high values indicate a high level of occupational reward) measures satisfaction with nancial- and status-related aspects, esteem rewards and job security. If the ratio between effort and reward at work is higher than 1 (high effort and low occupational reward), health risks can be expected. Descriptive analyses encompass means and standard deviations for quantitative measurements and percentages for categorical variables. For ordinal and nonnormally distributed variables, medians are presented. Normal distribution was assessed by inspecting skewness and kurtosis of the respective residuals. Skewness and kurtosis should lie between 21 and 1 to accept normal distribution. Comparisons between subgroups were

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

G. FREUDE ET AL.: DISCREPANCY BETWEEN CALENDAR AND BIOLOGICAL AGE 23

conducted using analysis of variance (ANOVA) with subsequent pairwise comparisons (Tukeys B) or chi-square tests according to the scaling of the target variable. To determine potential work- and health-related predictors and to weigh positive and negative inuences, bivariate correlation and regression analyses were calculated as a rst step and these were then extended to cover multiple regression models. Only such predictors were involved in the models, which are not in the VMS inventory for BA measurement; in total, 23 potential predictors were analysed. The level of signicance was 0.05 (two sided) for all statistical tests. The subgroup comparisons of Model 1 were corrected for multiplicity using Tukeys B method. However, no adjustment for multiplicity was used in Model 2 and Model 3. Thus, P values ,5%/23 0.002 lead to statistical signicance after adjustment for multiplicity on the 5% level using the Bonferroni correction. All analyses were done using commercially available software (SPSS for Windows, release 15.0).

Results
The sample consisted of 371 subjects belonging to the following ve subgroups: female teachers (n 5 100) [17,18], female ofce workers (n 5 60) [15], female nursery school teachers (n 5 65) [14], male teachers (n 5 99) [15,18] and male managers (n 5 47) [16]. Participation rates differed among the ve subgroups (female teachers, 58%; ofce workers, 57%; nursery school teachers, 86%; male teachers, 28% and managers, 70%). The subjects were aged between 20 and 64 years (mean age: 45.6 years). The majority of subjects were married (80%). Education and vocational training of the groups differed in the following ways: all female teachers completed high school, followed by 81% of the group of managers and 78% of male teachers. The majority of ofce workers (67%) and nursery school teachers (91%) completed secondary school. A university education was completed by 100% of female teachers, 77% of male teachers and 49% of managers. A vocational training was completed by 75% of ofce workers and 100% of nursery school teachers. As seen in Figure 1, the highest discrepancy between CA and BA (9 years) was established for the subgroup of management personnel, followed by the subgroup of female teachers (5 years), while for the subgroups of male teachers, nursery school teachers and ofce workers, no signicant discrepancies between CA and BA were found. The scatter plots shown in Figure 1 (lower graph) emphasize the high degree of interindividual differences with respect to BA for persons of the same CA. Dots below the reference line (BA 5 CA) represent subjects who are biologically younger. Dots above the reference line represent biologically older subjects. It is obvious that

almost all dots representing the group of managers are located below the reference line supporting the high level of vitality in this subgroup. Furthermore, they showed the lowest number of diseases and the lowest number of sick leave days (Table 1). Relaxation ability (measured on basis of RI questionnaire) was also best in the group of managers (mean RI 5 10.2). The female teachers had the worst (mean RI 5 17.1). Nevertheless, a high proportion of managers showed health risks as indicated by a high percentage of overweight persons (72%), and 9% were obese (BMI $ 30 kg/m2), which is a predictor for cardiovascular diseases. The proportion of smokers was highest in the group of nursery school teachers and lowest in the group of female teachers. Work ability was also highest in the group of managers (highest mean WAI score and highest values in the WAI items; Figure 2) with the greatest percentage of persons with excellent work ability (94%) compared to teachers (female: 15% and male: 22%), ofce workers (30%) and nursery school teachers (26%). The dimensions effort, reward and the ERI ratio (Table 2) were most favourable for the group of managers (lowest values in the effort subscale and highest values in the reward subscales). The worst values in the dimension reward were found in the group of male teachers. The multivariate analysis includes three general linear models (Table 3). Model 1 is equivalent to a one-factorial ANOVA with group as the only factor. The constant in this model equals the average difference between CA and BA for the group of manager personnel. The other parameters represent the difference between the four remaining groups and this group. Managers had the most favourable results, followed by female teachers. Model 2 explores potential characteristics of subjects, which predict the difference between CA and BA. WAI 7 (mental attitudes and resources towards work), fat mass and reward are signicantly associated with the outcome variable. From column 2 of Table 3, we can conclude that an increase of one point of the WAI 7 (mental resources) increases the difference between CA and BA by 1.6 years with a condence interval between 0.4 and 2.8 years. Model 3 demonstrates whether the group differences of Model 1 may be explained by the predictors identied in Model 2. In general, the differences between the groups become lower. There are only slight changes for nursery teachers and ofce workers but relevant changes for both groups of school teachers. Thus, the covariates explain at least in part why teachers have worse results compared to management personnel. The covariates explain to a much lesser degree why ofce workers and nursery school teachers also showed worse results compared to the management personnel.

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

24 OCCUPATIONAL MEDICINE

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

Figure 1. Upper graph: differences between CA and BA for different subgroups (means and SEs). Lower graph: scatter plots for BA versus CA for the groups of male teachers and management personnel. The black line represents a reference line (BA 5 CA). Dots under these lines represent individuals whose BA is lower than CA. It is obvious that almost all managers are biologically younger compared to their CA.

Discussion
In our study, the highest discrepancy between CA and BA was found in the group of managers compared to the other subgroups analysed in this study. BA was considerably below CA with an average gain of 9 years, i.e. the managers examined were biologically substantially younger and more vital than would be expected from their CA.

The best vitality status is consistent with the result that this subgroup also exhibited good results in relevant work- and health-related factors. Managers rated their work ability as the best compared to other subgroups. According to the concept of work ability [2224], we assumed a good consistency between individual resources and the work demands. This assumption was also supported by the result that managers

G. FREUDE ET AL.: DISCREPANCY BETWEEN CALENDAR AND BIOLOGICAL AGE 25

Table 1. Selected health-related factors for different subgroups Teachers (female), n 5 100 Ofce workers (female), n 5 60 Nursery school teachers (female), n 5 65 43.9 44.9 2.0 #9 14.1 25.0 23 43 23 6 9.2 6 8.4 6 1.8 6 3.1 6 3.8 Teachers (male), n 5 99 Management personnel (male), n 5 47 49.2 40.2 1.1 0 10.2 26.4 23 72 17 6 7.5 6 8.2 6 1.2 6 3.4 6 2.7 Total, n 5 371

CA (years) BA (years) Illnesses (quantity) Sick leave days RI BMI (kg/m2) Body fat mass (%) Overweight (BMI $ 25 kg/m2; %) Smoker (%)

44.6 40.6 2.2 #9 17.1 23.7 20 28 10

6 7.7 6 8.0 6 1.8 6 3.7 6 3.4

42.6 44.5 2.0 #9 13.5 24.3 23 38 15

6 8.7 6 5.3 6 1.7 6 3.8 6 3.8

47.9 46.0 1.6 #9 14.3 26.5 21 69 7

6 6.4 6 7.0 6 1.5 6 3.5 6 2.8

45.6 43.4 1.8 #9 14.4 25.1 22 49 13

6 8.1 6 7.8 6 1.7 6 4.1 6 3.5

Entries are means and SDs, percentages, according to the scaling of the variable. Sick leave days were coded as an ordinal scale as no sick leave days versus one to nine sick leave days versus more than nine sick leave days. Entries are medians of this scale, not on a scale coding exact number of days.

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

Figure 2. WAI for different subgroups with WAI categories (means and SEs).

Table 2. Effort, reward and ERI ratio for different subgroups Teachers (female) Effort Reward Effortreward ratio 17.1 6 4.3 48.6 6 5.1 0.7 6 0.2 Ofce workers (female) 14.7 6 3.6 47.5 6 6.3 0.6 6 0.2 Nursery school teachers (female) 14.2 6 3.0 51.2 6 3.9 0.5 6 0.1 Teachers (male) Management personnel (male) 13.3 6 3.2 52.5 6 2.9 0.5 6 0.1 Total

14.9 6 3.6 42.7 6 7.6 0.7 6 0.3

15.1 6 3.9 47.8 6 6.6 0.6 6 0.2

showed a good balance between effort and reward, which indicates a relatively low level of work-related stress. Furthermore, managers reported not only few illnesses, few sick leave days and high relaxation ability but also fewer health risk factors, such as being overweight and smoking.

By means of multiple regression analysis, mental attitudes and resources towards work, occupational reward and the body fat mass were identied as relevant predictors for the discrepancy between CA and BA. As already mentioned, the inuence of work in combination with health-related factors on the ageing process

26 OCCUPATIONAL MEDICINE

Table 3. Predictors of the CABA discrepancy Predictor Constant Teachers (female)management personnel (male) Ofce workers (female)management personnel (male) Nursery school teachers (female)management personnel (male) Teachers (male)management personnel (male) Mental resources Fat mass Reward R2/R2adj. 0.1/0.1*** Model 1 9.0 24.05** (20.3) CI: (26.4 to 27.8*** (20.4) CI: (20.4 to 27.0*** (20.4) CI: (29.6 to 27.2*** (20.5) CI: (29.4 to Model 2 25.1 21.7) 25.1) 24.5) 24.9) 1.6** (0.2) CI: (0.4 to 2.8) 20.2*** (20.2) CI: (20.3 to 20.1) 0.2** (0.2) CI: (0.1 to 0.3) 0.2/0.1*** Model 3 0.9 21.6n.s. (20.1) CI: (24.0 to 0.8) 26.0*** (20.3) CI: (28.6 to 23.5) 26.6*** (20.3) CI: (29.0 to 24.1) 23.8** (20.3) CI: (26.4 to 21.2) 1.7** (0.2) CI: (0.5 to 2.9) 20.2*** (20.2) CI: (20.3 to 20.1) 0.1* (0.1) CI: (0.01 to 0.2) 0.2/0.2***

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

*P , 0.05, **P , 0.01, ***P , 0.001; n.s., not signicant. Entries are unstandardized regression coefcients (within brackets, standardized regression coefcients are given). Additionally, condence intervals are presented for the unstandardized coefcients. In the last row of the table, the percentage of explained variance is described by adjusted and unadjusted R2 values. The P values in the last row refer to overall tests of the entire model. Model 1: P values refer to the comparison versus male management personnel; additional signicances: teachers (female) are signicantly different from ofce workers, nursery school teachers and teachers (male), P , 0.05 each; ofce workers, nursery school teachers and teachers (male) are not different from each other, corrected subgroup comparisons (Tukeys B). Model 2: P values refer to the test for beta 5 0 (no inuence of the covariate on the discrepancy of BA and CA). Model 3: P values for subgroups refer to the comparison versus male management personnel adjusted for the covariates below; additional signicances: teachers (female) versus ofce workers (female) and nursery school teachers (female), P , 0.005.

has hardly been studied to date. The application of highly standardized instruments in this study to measure BA as well as RI, WAI and ERI can be regarded as a comprehensive and innovative approach to analyse predictors for the discrepancy between CA and BA. These instruments were applied in all subgroups analysed in this study. Individual employees who participated in the study beneted from this complex approach because they got valuable and comprehensive information about their level of vitality and related risk factors as well as resources. However, there are signicant limitations on the interpretation of our results in relation to the cross-sectional design and a non-random sampling of different occupational groups. From our experiences, individuals with a high level of vitality are highly motivated to pursue opportunities for diagnostic procedures, resulting in a high participation rate such as in the group of managers. This contrasts with the lower observed participation rate of the less vital teachers. Therefore, a correlation between vitality and (self-) selection cannot be excluded. Sampling different groups with different participation rates might introduce signicant bias in the analysis of associations if factors relevant for (self-) selection are correlated both with determinants and with dependent variables in the current investigation and hence are classical confounders in the analysis of potentially causal relationships. However, there is no systematic correlation between participation rate and vitality of participants. For example, both the occupational group with the highest participation rate (nursery school teachers) and that with the

lowest participation rate (male teachers) show no differences between CA and BA. In contrast, managers with a lower participation rate than that of nursery school teachers revealed the highest CABA discrepancy. Another weakness is that cross-sectional studies do not control for selection effects over time introduced by a process of migration of selected subjects into an occupation and by a selection of a subgroup of healthy workers remaining in this group. With such cross-sectional data, it is difcult to determine the direction of the effects between occupation and the other variables. We concede that self-selection of subjects may also bias the strength of predictors in our regression analyses. This may hold especially in Model 3 where we can explain discrepancies between the groups in part by relevant predictors. On the other hand, we did not nd signicant interactions between the study groups and the predictors; thus, the latter bias may not be crucial. Because of these possible biases, we have to interpret the results with caution, and we cannot draw nal conclusions about the effects of work- and health-related factors on the CABA discrepancy. We found interesting associations between different predictors and discrepancy and between CA and BA discrepancy, which need further clarication on causal relationships.One mighthypothesizeonthebasis ofourdata that managers in our sample have good resources, which compensate for high workload. For example, they reported high levels of job satisfaction, esteem rewards and social support, a high job security (92% of managers believed

G. FREUDE ET AL.: DISCREPANCY BETWEEN CALENDAR AND BIOLOGICAL AGE 27

that their job was not at risk) and high educational qualications that may act positively during the ageing process. Causal relationships between work- and health-related predictors and CABA discrepancy have to be investigated in longitudinal studies. Our study provides rst valuable information about possible association between predictors and CABA discrepancy, which can be regarded as a predictor of the rate of ageing. We assume that measures focused on promoting health (healthy diet and physical activities) and improving working conditions (e.g. job satisfaction and social support and stress prevention) may also affect the ageing process positively.

Key points
This study shows that discrepancy between calendar and biological age is associated with different work- and health-related determinants. Protective factors (high work ability, appropriate relation between effort and reward at work and good relaxation ability) are associated with a lower biological age than actual calendar age. We assume that measures focused on health promotion and improving working conditions (e.g. job satisfaction and social support and stress prevention) may also affect the ageing process positively.

Funding
Federal Institute for Occupational Safety and Health, Berlin, Germany.

Conicts of interest
None declared.

References
1. Eisenmenger M, Potzsch O, Sommer B. 11. Koordinierte Bevolkerungs-vorausberechnung Annahmen und Ergebnisse. 11. Co-ordinated Population forecasts assumptions and results. Wiesbaden, Germany: Statistisches Bundesamt, 2006. 2. Birren JE, Schaie W. Handbook of the Psychology of Aging. 6th edn. San Diego, CA: Academic Press, 2006. 3. Ilmarinen J. Functional capacity and ageing. In: Ilmarinen J, ed. Towards a longer work life Ageing and the quality of work life in the European Union. Jyvaskyla, Finland: Gummerus Kirjapaino Oy, 2006; 171202. 4. Arking R. How do we age. In: Arking R, ed. The Biology of Aging: Observations and Principles. 3rd edn. New York, NY: Oxford University Press, 2006; 135319. 5. Baltes MM, Carstensen LL. The process of successful ageing. Ageing Soc 1996;16:397422. 6. Bowling A, Dieppe P. What is successful ageing and who should dene it? BMJ 2005;331:15481551.

7. Rowe JW, Kahn RL. Human aging: usual and successful. Science 1987;237:143149. 8. Depp CA, Jeste DV. Denitions and predictors of successful aging: a comprehensive review of larger quantitative studies. Am J Geriatr Psychiatry 2006;14:620. 9. Robson SM, Hansson RO, Abalos A, Booth M. Successful aging: criteria for aging well in the workplace. J Career Dev 2006;33:156177. 10. Barnes JL, Pietrowski MJ. Discrepancies between chronological age and personal age as reection of unrelieved worker stress. Work and Stress 1991;5:177187. 11. Hacker W. Leistungsforderliches und gesundheitsgerechtes r Gestalten wissensintensiver geistig-schopferischer Arbeit fu die Arbeitslebensspanne. Performance-promoting and health-adjusted design of skill-intensive work lifespan. Proceedings of the conference Generation 60plus tauglich fur die Arbeitswelt 2020?. Bad Arolsen, Germany: Marie Luise und Ernst Becker Stiftung, 2006;109116. 12. Ries W, Sauer J. Biologisches Alter. Biological age. Berlin, Germany: Akademie, 1991. 13. Meiner-Po thig D. Vitalitatsdiagnostik nach POTHIG. In: Meiner-Pothig D, Michalak U, eds. Vitalitat und arztliche Intervention: Vitalitatsdiagnostik: Grundlagen Angebote Konsequenzen. Stuttgart, Germany: Hippokrates, 1997; 6472. 14. Seibt R, Khan A, Thinschmidt M, Dutschke J, Weidhaas J. Gesundheitsforderung und Arbeitsfahigkeit in Kindertagesstatten. Healthpromotionandworkabilityinday-carecenters. Bremerhaven, Germany: Wirtschaftsverlag NW, 2005. (Schriftenreihe der Bundesanstalt fur Arbeitsschutz und Arbeitsmedizin: Forschungsbericht, Fb 1049). 15. Seibt R, Thinschmidt M, Lutzkendorf L, Hansch S. rofachkraften Arbeitsfahigkeit und Vitalitat von Lehrern und Bu Ein Vergleich. Work ability and vitality of teachers and ofce workers a comparison. Bremerhaven, Germany: Wirtschaftsverlag NW, 2006(Schriften-reihe der Bundesanstalt fur Arbeitsschutz und Arbeitsmedizin: Forschungsbericht, Fb 1087). 16. Freude G. VitalitatundArbeitsfahigkeitvonFuhrungskraften: Altersassoziierte Veranderungen und salutogenetische Fakto ren. Vitality and work ability of managers: age-associated changes and salutogenetic factors. Arbeitsmed Sozialmed Umweltmed 2007;42:544553. 17. Seibt R, Rehm U, Hubler A. Pravalenz ausgewahlter Diag nosen sachsischer Lehrerinnen und Lehrer in Grundschulen, Mittelschulen und Gymnasien. Prevalence of selected diagnosis from saxonien teachers in primary and grammar schools and in gymnasiums. Arbeitsmed Sozialmed Umweltmed 2007;42:476482. 18. Seibt R, Spitzer S, Blank M, Scheuch K. Predictors of work ability in occupations with psychological stress. Journal of Public Health 2008;6:110. 19. IAQ/HBS Arbeitszeitmonitor. Working Time Monitor. http:// www.iaq.uni-due.de/iaq-report/2009/report2009-01.pdf. Internet offers of the Institute for Work, Skills and Training, within the Faculty of Social Sciences at the University of Duisburg-Essen 2009. 20. IAB: Berufe im Spiegel der Statistik. Vocations Mirrowed in the Statistics 1999-2007. http://www.pallas.iab.de/bisds/ berufe.htm. Internet offers of the Institute for Employment Research (IAB) of the Federal Employment Agency (BA).

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011

28 OCCUPATIONAL MEDICINE

21. Richter P, Rudolf M, Schmidt CF. FABA: Fragebogen zur Analyse belastungsrelevanter Anforderungsbewaltigungen. Questionnaire for analysing exposure-related coping of demands. Manual. Frankfurt (Main), Germany: Swets, 1996. 22. Ilmarinen J, Rantanen J. Promotion of work ability during aging. Am J Ind Med 1999;1:2123. 23. Ilmarinen J. Aging and work coping with strengths and weaknesses. In: Tuomi K, ed. Eleven-year follow-up of ageing workers. Scand J Work Environ Health. 1997; 1:35.

24. Hasselhorn HM, Freude G. Der Work Ability Index ein Leitfaden. Work Ability Index a guideline. Bremerhaven, Germany: Wirtschaftsverlag NW, 2007(Schriftenreihe der Bundes-anstalt fur Arbeitsschutz und Arbeitsmedizin: Sonderschrift 87). 25. Siegrist J. Adverse health effects of high-effort/low-reward conditions at work. J Occupat Health Psychol 1996;1:2741. 26. Siegrist J. Der Fragebogen zur Messung beruicher Gratikationskrisen. Questionnaire for measuring occupational gratication crisis. Ergo-Med 2003;27:134136.

Downloaded from http://occmed.oxfordjournals.org/ by guest on December 13, 2011