Age, burnout and physical and psychological work ability among nurses

Age, burnout and physical and psychological work ability among nurses Abstract Background The ageing of the US labour force highlights the need to examine older adults’ physical and psychological ability to work, under varying levels of occupational burnout. Aims To examine how age and burnout interact in predicting physical and psychological work ability. Methods Using a cohort of actively working nurses, we assessed factors on the Work Ability Index at 12-month follow-up and determined how these were related to age and exhaustion-related burnout at baseline. Results The study group consisted of 402 nurses aged 25–67 (mean = 41.7). Results indicated age by burnout interactions in which decrements in physical work ability with greater age were observed at all but the lowest level of burnout (1.5 SD below mean: β = −0.14, 95% CI −0.36, 0.07; 1 SD below: β = −0.23, 95% CI −0.39, −0.06; mean: β = −0.39, 95% CI −0.50, −0.29; 1 SD above: β = −0.56, 95% CI −0.70, −0.42; 1.5 SD above: β = −0.64, 95% CI −0.83, −0.46). In contrast, we observed decrements in psychological work ability with age at higher levels of burnout only (1 SD above: β = −0.20, 95% CI −0.35, −0.05; 1.5 SD above: β = −0.30, 95% CI −0.49, −0.11); at lower levels of burnout, older age was associated with improvements in this (1 SD below: β = 0.19, 95% CI 0.03, 0.35; 1.5 SD below: β = 0.29, 95% CI 0.08, 0.50). Conclusions Findings indicated physical and psychological dimensions of work ability that differed by age and occupational burnout. This emphasizes the need for interventions to reduce burnout and to address age-related strengths and vulnerabilities relating to physical and psychological work ability. Burnout, older workers, work ability Introduction A range of demographic and social factors are leading to a ‘greying’ labour force in many countries. Factors including better health, increased life expectancy and lower birth rates over time have led to higher age-dependency ratios, and resulting increases in the age of entitlement to retirement benefits [1]. Consequently, labour force participation after age 65 continues to rise in developed economies including Europe and the USA [2]. Given the prospective need for older adults in many labour markets to work longer, it is important to understand how to help individuals maintain work ability as they age. Much of the research on promoting work ability in ageing uses the Work Ability Index (WAI). This questionnaire assesses workers’ perceived performance of their job relative to its mental and physical demands [3]. Studies have found the WAI to predict multiple psychosocial and health outcomes in ageing, including increased health care utilization [4], retirement due to disability [5] and function and well-being in retirement [6]. Research across countries and occupations has generally found work ability to decline with age [7] due to greater physical decrements and chronic medical conditions [8]. However, evidence also indicates age-associated gains in cognitive function, including indicators of crystallized intelligence such as vocabulary, skills and work-related knowledge, and the ability to maintain psychological and emotional health [9,10]. This suggests the need to better understand the effect of age by type of work ability. Clinicians generally assess the WAI as a total score and researchers have generally considered this measure to be a uni-dimensional scale; however, research suggests that the WAI consists of more than one dimension. In one such study [11], researchers examined a sample of German workers from different occupational groups and found marked improvement in a bi-dimensional solution consisting of items assessing health-related work ability, including number of injuries or diseases diagnosed by a physician, sick leave taken and an estimate of work impairment due to disease, and items relating to subjective work ability, including psychological resources, subjective estimates of current work ability and perceived prognosis of work ability. Similar dimensionality was reported in the European Nurses’ Early Exit (NEXT) Study [12], in 8 out of 10 European countries studied. These studies suggest a psychological dimension to work ability that includes a subjective appraisal of work ability and mental resources, and a physical dimension of work ability that includes an objective characterization of health based on accumulated medical conditions and injuries. One condition that may adversely influence both physical and psychological dimensions of work ability is burnout. In the current study, we defined the core feature of burnout as the presence of emotional, physical and cognitive exhaustion due to job stressors [13]. Burnout is important to the work ability of older adults, given that they appear to need more job controls to buffer the effects of job-related stressors [14], recover from stress less quickly [15] and are more sensitive to the effects of job-related stress and burnout on self-efficacy [16] and age-related cognitive decline [17]. This suggests that older workers may be sensitive to burnout symptoms, and the effects of these on work ability. This research on the interaction between work ability, age and burnout suggests that an individual’s capacity to maintain work ability with age may differ across physical and psychological dimensions, and that it may depend on the severity of burnout symptoms. In this study, we examined the association of physical and psychological work ability with age, and moderation of these associations by burnout. In doing so, we sought to confirm a bi-dimensional WAI model [11] in an independent, single-occupation sample of working nurses. This sample fits our study objectives well, as nurses are particularly vulnerable to burnout [18], based on exposure to a high level of both physical and psychological stressors [19]. Methods We recruited individuals actively working in the nursing field from a health care system in the southeast USA. Health system administration provided e-mail addresses of nurses employed across all departments. We targeted recruitment for 400 persons to achieve a sample size adequate to detect statistically significant differences. Eligible individuals were required to be actively working in the nursing field (not on medical leave, disability leave or family leave), to have at least 2 years’ experience in the nursing field and be at least 25 years old. In addition, because we conducted cognitive testing as part of our broader project, we excluded individuals with possible confounding neurological conditions (seizures, severe brain trauma and stroke). Race and sex ratios approximated the US nursing population as a whole [20], with the exception that it over-represented Black/African Americans [14%, compared to 5% in US registered nurses (RN), see Table 1]. Participants gave informed consent. This research was approved by the Institutional Review Board of Duke University. This study was based on a prospective cohort design, including assessment of burnout at baseline and of work ability at 12-month follow-up. The current study used data collected from a broader project on work stress and mental health among nurses. We collected baseline data by self-reported questionnaires, including demographic information and assessment of burnout. A trained research technician was present during completion of baseline questionnaires to provide assistance but was positioned so as not to be able to view participant responses. After baseline, we sent follow-up questionnaires monthly via e-mail with a link to a confidential survey, which concluded at month 12 in February of 2016. Although data on burnout were collected monthly, in this study, we used burnout at baseline only, to predict WAI at 12-month follow-up. We used factors derived from the 12-month follow-up assessment of the WAI [3] as the dependent measures in this study. The WAI consists of items related to physical health, including number of injuries or physician- diagnosed diseases endorsed by the participant, out of a list of 49 injuries and diseases (item 3), number of sick leave days (item 5) and an estimate of work impairment due to illness (item 4). Other items assess psychological resources (enjoyment of daily activities, alertness and hopefulness for the future; item 7), subjective estimates of current ability to work compared with lifetime best (item 1), ability to work in relation to job demands (item 2) and perceived prognosis of work ability in 2 years (item 6). We assessed burnout at baseline with the exhaustion subscale of the Oldenburg Burnout Inventory (OLBI; [13]). This subscale is composed of eight items, including, ‘I can tolerate the pressure of my work very well’ and ‘during my work, I often feel emotionally drained’. Items are assessed on a four point scale, ranging from ‘strongly agree’ to ‘strongly disagree’. We reverse-coded the four negatively worded items before calculating the mean of all eight items, yielding total scores ranging from 1 to 4. We found the internal consistency of these items to be acceptable in our sample (α = 0.81). In a recent study [21], researchers found that 40–45% of the variance in overall OLBI burnout at three waves across 3 years can be accounted for by a stable trait component, suggesting stability as well as change in burnout over time. We selected baseline age, sex, educational level and nursing certification status as covariates based on their associations with work ability in previous studies. We categorized education as: high school equivalent, associate’s degree, bachelors’ degree and master’s degree/PhD, and nursing certification as RNs versus other certification types. To confirm the factor structure of the WAI and to test the association of age and burnout to work ability at the 12-month follow-up, we used confirmatory factor analysis (CFA) with covariates models [22]. This method consists of two steps. In the first, the adequacy of a measurement model is tested. In the second, factors from this model are linearly regressed with covariates, to identify differences in factor means by covariates. In accordance with Martus et al. [11], we tested the adequacy of four CFA measurement models. In model A, we tested a one-factor solution, with all seven WAI items included. For models B and C, we entered items 1, 2 and 7, along with items 4 and 6, to reflect psychological work ability, and items 3 and 5 to reflect physical work ability. To test the assumption of factor orthogonality, we constrained the covariance between factors in model B to zero. However, this made the model under-justified. To address this, we constrained the factor loadings in this model to equality, so as to have fewer parameters to estimate. In model C, we tested the same item structure as in model B, but freely estimated the covariance between factors. We expanded this in model D, in which we cross-loaded items 4 and 6. This was done because these items assess both psychological and physical work ability, in that they assess the participants’ subjective estimation of ability in regards to the participants’ health. In step 2, we linearly regressed factors from the best-fitting model on covariates, including centred burnout and centred age at baseline (model 1); demographic covariates (sex, education and nurse type; model 2); and the interaction between burnout and age (model 3), controlling for any covariates found to be significant in model 2. For all models, we reported standardized factor loadings, factor covariances and predictor coefficients where applicable, along with 95% confidence intervals (CIs) and fit statistics [23]. We conducted descriptive statistics with SAS version 9.3 [24], and CFA models with Mplus version 7 [25]. Results The study population was 402 individuals, which included 372 RNs and 28 other nursing workers (licensed practical nurses, certified nursing assistants and medical assistants). Mean age (SD) at baseline among the 352 participants who completed follow-up at month 12 was 42.1 (11.4), with 20% being age 55 or older (n = 70). Most were female (n = 325, 92%) and had an RN certification (n = 326, 93%) with either an associate or bachelor’s degree (n = 91, 26% and n = 205, 59%). The completion rate at follow-up month 12 was 88% (n = 352). We found that loss to follow-up was not significantly related to age [t(397) = −1.52], burnout [t(400) = 1.18], sex [χ2(1) = 0.88], education [χ2(1) = 0.98], nurse type [χ2(1) = 0.02] or total WAI at baseline [t(400) = −0.40], suggesting that loss to follow-up was not biased on these variables. Table 1 contains zero-order correlations between age and other demographic characteristics, total work ability at month 12 and burnout. These indicated lower work ability based on WAI total score at month 12 to be associated with higher severity of burnout symptoms at baseline, r(350) = −0.40, P < 0.001, and greater age and female sex to be associated with lower burnout [r(347) = −0.13, P < 0.05 and t(350) = 3.32, P < 0.01]. Table 1. Bivariate relationships between age and other demographic characteristics, total work ability at month 12 and OLBI exhaustion Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Sample restricted to those who completed follow-up at month 12. 95% CI, 95% confidence intervals for Pearson correlations based on Fisher’s z transformation; non-RN, non-RN nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RN, registered nurse; total WAI, Work Ability Index total score at follow-up month 12. View Large Table 1. Bivariate relationships between age and other demographic characteristics, total work ability at month 12 and OLBI exhaustion Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Sample restricted to those who completed follow-up at month 12. 95% CI, 95% confidence intervals for Pearson correlations based on Fisher’s z transformation; non-RN, non-RN nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RN, registered nurse; total WAI, Work Ability Index total score at follow-up month 12. View Large Tables 2 and 3 report correlation coefficients of study variables, and results of CFA models. Findings for CFA model A indicated that although all item loadings were significant, a one-factor solution did not adequately fit observed data [χ2(14) = 121.9, P < 0.001; standardized root mean square residual (SRMR) = 0.09; root mean square error of approximation (RMSEA) = 0.15; comparative fit index (CFI) = 0.74; Akaike information criterion (AIC) = 6866]. Findings for model B also exhibited poor fit [χ2(18) = 140.2, P < 0.001; SRMR = 0.16; RMSEA = 0.14; AIC = 6877]. Findings for model C appeared to exhibit better absolute fit [χ2(13) = 96, P < 0.001; SRMR = 0.08; AIC = 6842], though fit adjusted for model parsimony was poor (RMSEA = 0.14), as was fit compared to that of the null model (CFI = 0.80). In contrast, model D, the two-factor non-orthogonal solution with items 4 and 6 cross-loaded onto both factors, demonstrated good fit on all indices [χ2(11) = 15.4, P = 0.17; SRMR = 0.03; RMSEA = 0.03; CFI = 0.99], and comparatively greater fit than other models (AIC = 6766), including model C [likelihood ratio test: χ2(2) = 80.4, P < 0.001]. Because this measurement structure demonstrated good fit in this study and in Martus et al. [11], we used this measurement model for subsequent models of age and moderation by burnout. Table 2. Pearson correlations between variables entered into confirmatory factor analyses Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex: Oldenburg Burnout Inventory, Exhaustion subscale; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 2. Pearson correlations between variables entered into confirmatory factor analyses Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex: Oldenburg Burnout Inventory, Exhaustion subscale; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 3. Confirmatory factor analysis of WAI Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Parameter estimate (est.) and 95% CI reported for each factor loading. AIC, Akaike information criterion; CFI, comparative fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aModel A: one-factor solution. bModel B: two-factor solution, 4 and 6 on WAI psychological, factor covariance constrained to 0. cModel C: two-factor solution, 4 and 6 on WAI psychological, factor covariance freely estimated. dModel D: two-factor solution, 4 and 6 on both factors, factor covariance freely estimated. View Large Table 3. Confirmatory factor analysis of WAI Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Parameter estimate (est.) and 95% CI reported for each factor loading. AIC, Akaike information criterion; CFI, comparative fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aModel A: one-factor solution. bModel B: two-factor solution, 4 and 6 on WAI psychological, factor covariance constrained to 0. cModel C: two-factor solution, 4 and 6 on WAI psychological, factor covariance freely estimated. dModel D: two-factor solution, 4 and 6 on both factors, factor covariance freely estimated. View Large Table 4 reports findings from CFA with covariates models regressing WAI physical and WAI psychological on age, burnout, demographic covariates and the interaction between age and baseline burnout. Greater burnout was associated with diminished physical and psychological work ability (β = −0.25, 95% CI −0.36, −0.13 and β = −0.55, 95% CI −0.65, −0.46), while older age was associated with diminished physical work ability only (β = −0.41, 95% CI −0.51, −0.31). Fit statistics for this model indicated acceptable fit [χ2(21) = 52.1, P < 0.001; SRMR = 0.04; RMSEA = 0.07; CFI = 0.94]. Findings did not change markedly with inclusion of covariates (model 2), though the effect for nurse certification was significant, in that nurses with certifications other than RN exhibited diminished psychological work ability (β = −0.17, 95% CI −0.32, −0.03). Because nurse certification was significant, it was retained in model 3, which tested the interaction between age and burnout. Table 4. Confirmatory factor analysis of WAI psychological and WAI physical on OLBI exhaustion, age, demographics and OLBI exhaustion by age interaction Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Loadings for WAI psychological and WAI physical factors available upon request. Parameter estimate (Est.) and 95% CI reported for each coefficient or estimate of factor covariance. AIC, Akaike information criterion; CFI, comparative fit index; non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 4. Confirmatory factor analysis of WAI psychological and WAI physical on OLBI exhaustion, age, demographics and OLBI exhaustion by age interaction Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Loadings for WAI psychological and WAI physical factors available upon request. Parameter estimate (Est.) and 95% CI reported for each coefficient or estimate of factor covariance. AIC, Akaike information criterion; CFI, comparative fit index; non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large In model 3, we found significant interactions between age and baseline burnout for both physical work ability (β = −0.16, 95% CI −0.27, −0.05) and psychological work ability (β = −0.19, 95% CI −0.29, −0.08), with fit statistics indicating acceptable fit [χ2(31) = 66.0, P < 0.001; SRMR = 0.04; RMSEA = 0.057, 95% CI 0.04–0.08; CFI = 0.94]. To interpret the interactions, we tested and plotted simple slopes for age in predicting work ability outcomes at different conditional values of burnout (the mean, and 1.0 and 1.5 SDs below and above the mean). As shown in Figure 1, findings revealed that at low levels of burnout (1 SD below mean), and at mean and high (1.5 SDs above mean) levels of burnout, older age predicted diminished physical work ability (1 SD below: β = −0.23, 95% CI −0.39, −0.06; mean: β = −0.39, 95% CI −0.50, −0.29; 1 SD above: β = −0.56, 95% CI −0.70, −0.42; 1.5 SD above: β = −0.64, 95% CI −0.83, −0.46), while at very low levels of burnout (1.5 SD below mean), age was unrelated to physical work ability (β = −0.14, 95% CI −0.36, 0.07). With respect to psychological work ability (see Figure 2), results indicated that at mean levels of burnout, age did not predict psychological work ability (β = −0.01, 95% CI −0.11, 0.10); however, at higher levels of burnout (1 and 1.5 SDs above the mean of OLBI Ex), older age predicted diminished psychological work ability (1 SD above: β = −0.20, 95% CI −0.35, −0.05; 1.5 SD above: β = −0.30, 95% CI −0.49, −0.11), while at lower levels of burnout (1 and 1.5 SDs below the mean), older age predicted increased psychological work ability (1 SD below: β = 0.19, 95% CI 0.03, 0.35; 1.5 SD below: β = 0.29, 95% CI 0.08, 0.50). Figure 1. View largeDownload slide Physical work ability on age at conditional values of exhaustion burnout. Figure 1. View largeDownload slide Physical work ability on age at conditional values of exhaustion burnout. Figure 2. View largeDownload slide Psychological work ability on age at conditional values of exhaustion burnout. Figure 2. View largeDownload slide Psychological work ability on age at conditional values of exhaustion burnout. Discussion The current study confirmed a bi-dimensional factor structure for the WAI representing physical and psychological dimensions and extended this with two novel findings. Firstly, we found differences in physical and psychological dimensions for age and burnout, with older age being associated with lower physical work ability only, and greater burnout being associated with lower physical and psychological work ability. Secondly, we found that age and burnout interact. In the psychological dimension, older age predicted lower work ability at high levels of burnout, whereas at low levels of burnout, older age predicted higher work ability. In the physical dimension, older age generally predicted lower work ability, with the exception of no association at very low levels of burnout symptoms. The study has both strengths and limitations. We conducted assessments on a large cohort of actively working health professionals at baseline and 12-month follow-up. With a mean age of 42 years, many of the older workers in our sample are closer to midlife than later life; however, this is consistent with evidence linking reduced work ability and burnout in mid-career to higher risk for early retirement [7], and with approaches emphasizing interventions in midlife to maintain health and ability in late life [26]. We were limited in our use of convenience sampling of individuals who responded promptly to recruiting e-mails. This may have introduced unintended bias toward healthy, non-burned-out workers, but our descriptive data indicate an adequate range of burnout scores comparable to a sample of Norwegian nurses [27], and a demographic composition consistent with that of the nursing workforce in general. We note this study uses only the exhaustion component of burnout. Although exhaustion is theorized to be the core component of burnout, in the sense that it precedes other aspects of burnout [28], future studies would benefit from examining other components of burnout such as disengagement or personal accomplishment. Finally, our findings also highlight the need to examine how such interventions generalize to other human service professions. The current study supports previous research suggesting the principal metric of the WAI—the total score—may be better represented as a bi-dimensional measure [11,12], particularly with respect to age and severity of burnout. This is consistent with the broader literature suggesting overall work ability declines with age [7,29], but it highlights important exceptions. In line with the broader literature, physical work ability was lower with increasing age, and this relationship worsened with greater burnout, and, at higher levels of burnout, psychological work ability was also lower with increasing age. This suggests that at high burnout, older workers may not be as able to direct adequate physical and psychological resources toward their job duties. This is consistent with evidence finding older individuals need more job resources as a buffer to work stress [14], and recover less quickly from stress [15]. In contrast to the broader literature, at low levels of burnout, older workers exhibited greater psychological work ability. This suggests that under optimal conditions, older workers may be more able to engage cognitive, psychological and occupational strengths that improve with age [9,30]. Our study findings suggest that the WAI total score provides an incomplete clinical impression of the work ability of older individuals, who appear to have more resilience on psychological aspects of work ability under low-to-medium levels of burnout, but less resilience on physical aspects overall. This suggests that deficits in physical health are the most consistent contributors to decline in work ability with age, and that interventions targeting disease prevention and health maintenance from midlife and earlier may be effective ways to maintain work ability with age. In addition, support for mental/emotional resources in the workplace may provide an important buffer against burnout and its effect on the work ability of older individuals. Key points This study confirmed a bi-dimensional structure in the Work Ability Index in the form of physical and psychological work ability. It also found age relationships with these dimensions that were moderated by exhaustion-related burnout. Findings emphasize the need to reduce burnout and to address age-related strengths and vulnerabilities relating to physical and psychological work ability. Funding This work was supported by the German Federal Institute for Occupational Safety and Health (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, BAuA: Project F2318). Support for Daniel Hatch was provided by the National Institutes of Health (NIH Grant No. T32-AG000029). Competing interests None declared. References 1. Social Security Administration . Retirement Planner: Benefits by Year of Birth . 2016 . Woodlawn, MD : United States Social Security Administration. https://www.ssa.gov/planners/retire/agereduction.html 2. He W , Goodkind D , Kowal P. International Population Reports, P95/16-1, An Aging World: 2015 . Washington, DC : US Census Bureau , 2016 . 3. Tuomi K , Ilmarinen J , Jahkola A , Katjarinne L , Tulkki A. Work Ability Index . Helsinki, Finland : Finnish Institute of Occupational Health , 1998 . 4. Bethge M , Spanier K , Neugebauer T , Mohnberg I , Radoschewski FM . Self-reported poor work ability—an indicator of need for rehabilitation? A cross-sectional study of a sample of German employees . Am J Phys Med Rehabil 2015 ; 94 : 958 – 966 . Google Scholar CrossRef Search ADS PubMed 5. Roelen CA , Heymans MW , Twisk JW , van der Klink JJ , Groothoff JW , van Rhenen W . Work Ability Index as tool to identify workers at risk of premature work exit . J Occup Rehabil 2014 ; 24 : 747 – 754 . Google Scholar CrossRef Search ADS PubMed 6. Tuomi K , Huuhtanen P , Nykyri E , Ilmarinen J . Promotion of work ability, the quality of work and retirement . Occup Med (Lond) 2001 ; 51 : 318 – 324 . Google Scholar CrossRef Search ADS PubMed 7. Camerino D , Conway PM , Van der Heijden BI et al. ; NEXT-Study Group . Low-perceived work ability, ageing and intention to leave nursing: a comparison among 10 European countries . J Adv Nurs 2006 ; 56 : 542 – 552 . Google Scholar CrossRef Search ADS PubMed 8. Kaiser RM . Physiological and clinical considerations of geriatric patient care . In: Steffens DC , Blazer DG , Thakur ME , eds. The American Psychiatric Publishing Textbook of Geriatric Psychiatry . 5th edn . Arlington, VA : American Psychiatric Publishing , 2015 ; 33 – 59 . Google Scholar CrossRef Search ADS 9. Carstensen LL , Fung HH , Charles ST . Socioemotional selectivity theory and the regulation of emotion in the second half of life . Motiv Emot 2003 ; 27 : 103 – 123 . Google Scholar CrossRef Search ADS 10. Umanath S , Marsh EJ . Understanding how prior knowledge influences memory in older adults . Perspect Psychol Sci 2014 ; 9 : 408 – 426 . Google Scholar CrossRef Search ADS PubMed 11. Martus P , Jakob O , Rose U , Seibt R , Freude G . A comparative analysis of the Work Ability Index . Occup Med (Lond) 2010 ; 60 : 517 – 524 . Google Scholar CrossRef Search ADS PubMed 12. Radkiewicz P , Widerszal-Bazyl M . Psychometric properties of Work Ability Index in the light of comparative survey study . Int Congr Ser 2005 ; 1280 : 304 – 309 . Google Scholar CrossRef Search ADS 13. Demerouti E , Bakker AB , Nachreiner F , Schaufeli WB . The job demands-resources model of burnout . J Appl Psychol 2001 ; 86 : 499 – 512 . Google Scholar CrossRef Search ADS PubMed 14. Shultz KS , Wang M , Crimmins EM , Fisher GG . Age differences in the demand-control model of work stress: an examination of data from 15 European countries . J Appl Gerontol 2010 ; 29 : 21 – 47 . Google Scholar CrossRef Search ADS PubMed 15. Kiss P , De Meester M , Braeckman L . Differences between younger and older workers in the need for recovery after work . Int Arch Occup Environ Health 2008 ; 81 : 311 – 320 . Google Scholar CrossRef Search ADS PubMed 16. Shoji K , Cieslak R , Smoktunowicz E , Rogala A , Benight CC , Luszczynska A . Associations between job burnout and self-efficacy: a meta-analysis . Anxiety Stress Coping 2016 ; 29 : 367 – 386 . Google Scholar CrossRef Search ADS PubMed 17. Andel R , Infurna FJ , Hahn Rickenbach EA , Crowe M , Marchiondo L , Fisher GG . Job strain and trajectories of change in episodic memory before and after retirement: results from the health and retirement study . J Epidemiol Community Health 2015 ; 69 : 442 – 446 . Google Scholar CrossRef Search ADS PubMed 18. Demerouti E , Bakker AB , Nachreiner F , Schaufeli WB . A model of burnout and life satisfaction amongst nurses . J Adv Nurs 2000 ; 32 : 454 – 464 . Google Scholar CrossRef Search ADS PubMed 19. Aiken LH , Sloane DM , Bruyneel L , Van den Heede K , Sermeus W ; RN4CAST Consortium . Nurses’ reports of working conditions and hospital quality of care in 12 countries in Europe . Int J Nurs Stud 2013 ; 50 : 143 – 153 . Google Scholar CrossRef Search ADS PubMed 20. US Department of Health and Human Services . The Registered Nurse Population: Findings From the 2008 National Sample Survey of Registered Nurses . Rockville, MD : Health Resources and Services Administration , Ch. 7; 2010 ; 2 – 7 . PubMed PubMed 21. Brauchli R , Schaufeli WB , Jenny GJ , Füllemann D , Bauer GF . Disentangling stability and change in job resources, job demands, and employee well-being—a three-wave study on the job-demands resources model . J Vocational Behav 2013 ; 83 : 117 – 129 . Google Scholar CrossRef Search ADS 22. Jöreskog KG , Goldberger AS . Estimation of a model with multiple indicators and multiple causes of a single latent variable . J Am Stat Assoc 1975 ; 70 : 631 – 639 . 23. Hu L , Bentler PM . Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives . Struct Eq Model: Multidisciplin J 1999 ; 6 : 1 – 55 . Google Scholar CrossRef Search ADS 24. SAS Institute Inc . SAS Release 9.3 . Cary, NC : SAS Institute Inc . 2012 . 25. Muthén LK , Muthén BO. Mplus User’s Guide . 7th edn . Los Angeles, CA : Muthén & Muthén , 1998–2015 . 26. Szoeke C , Lehert P , Henderson VW , Dennerstein L , Desmond P , Campbell S . Predictive factors for verbal memory performance over decades of aging: data from the women’s healthy ageing project . Am J Geriatr Psychiatry 2016 ; 24 : 857 – 867 . Google Scholar CrossRef Search ADS PubMed 27. Innstrand ST , Langballe EM , Falkum E , Aasland OG . Exploring within- and between-gender differences in burnout: 8 different occupational groups . Int Arch Occup Environ Health 2011 ; 84 : 813 – 824 . Google Scholar CrossRef Search ADS PubMed 28. Maslach C , Schaufeli WB , Leiter MP . Job burnout . Annu Rev Psychol 2001 ; 52 : 397 – 422 . Google Scholar CrossRef Search ADS PubMed 29. Ilmarinen J , Tuomi K . Work ability of aging workers . Scand J Work Environ Health 1992 ; 18 ( Suppl. 2 ): 8 – 10 . Google Scholar PubMed 30. Ng TW , Feldman DC . The relationship of age to ten dimensions of job performance . J Appl Psychol 2008 ; 93 : 392 – 423 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Occupational Medicine Oxford University Press

Age, burnout and physical and psychological work ability among nurses

Loading next page...
 
/lp/ou_press/age-burnout-and-physical-and-psychological-work-ability-among-nurses-TMQFUc0ter
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0962-7480
eISSN
1471-8405
D.O.I.
10.1093/occmed/kqy033
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The ageing of the US labour force highlights the need to examine older adults’ physical and psychological ability to work, under varying levels of occupational burnout. Aims To examine how age and burnout interact in predicting physical and psychological work ability. Methods Using a cohort of actively working nurses, we assessed factors on the Work Ability Index at 12-month follow-up and determined how these were related to age and exhaustion-related burnout at baseline. Results The study group consisted of 402 nurses aged 25–67 (mean = 41.7). Results indicated age by burnout interactions in which decrements in physical work ability with greater age were observed at all but the lowest level of burnout (1.5 SD below mean: β = −0.14, 95% CI −0.36, 0.07; 1 SD below: β = −0.23, 95% CI −0.39, −0.06; mean: β = −0.39, 95% CI −0.50, −0.29; 1 SD above: β = −0.56, 95% CI −0.70, −0.42; 1.5 SD above: β = −0.64, 95% CI −0.83, −0.46). In contrast, we observed decrements in psychological work ability with age at higher levels of burnout only (1 SD above: β = −0.20, 95% CI −0.35, −0.05; 1.5 SD above: β = −0.30, 95% CI −0.49, −0.11); at lower levels of burnout, older age was associated with improvements in this (1 SD below: β = 0.19, 95% CI 0.03, 0.35; 1.5 SD below: β = 0.29, 95% CI 0.08, 0.50). Conclusions Findings indicated physical and psychological dimensions of work ability that differed by age and occupational burnout. This emphasizes the need for interventions to reduce burnout and to address age-related strengths and vulnerabilities relating to physical and psychological work ability. Burnout, older workers, work ability Introduction A range of demographic and social factors are leading to a ‘greying’ labour force in many countries. Factors including better health, increased life expectancy and lower birth rates over time have led to higher age-dependency ratios, and resulting increases in the age of entitlement to retirement benefits [1]. Consequently, labour force participation after age 65 continues to rise in developed economies including Europe and the USA [2]. Given the prospective need for older adults in many labour markets to work longer, it is important to understand how to help individuals maintain work ability as they age. Much of the research on promoting work ability in ageing uses the Work Ability Index (WAI). This questionnaire assesses workers’ perceived performance of their job relative to its mental and physical demands [3]. Studies have found the WAI to predict multiple psychosocial and health outcomes in ageing, including increased health care utilization [4], retirement due to disability [5] and function and well-being in retirement [6]. Research across countries and occupations has generally found work ability to decline with age [7] due to greater physical decrements and chronic medical conditions [8]. However, evidence also indicates age-associated gains in cognitive function, including indicators of crystallized intelligence such as vocabulary, skills and work-related knowledge, and the ability to maintain psychological and emotional health [9,10]. This suggests the need to better understand the effect of age by type of work ability. Clinicians generally assess the WAI as a total score and researchers have generally considered this measure to be a uni-dimensional scale; however, research suggests that the WAI consists of more than one dimension. In one such study [11], researchers examined a sample of German workers from different occupational groups and found marked improvement in a bi-dimensional solution consisting of items assessing health-related work ability, including number of injuries or diseases diagnosed by a physician, sick leave taken and an estimate of work impairment due to disease, and items relating to subjective work ability, including psychological resources, subjective estimates of current work ability and perceived prognosis of work ability. Similar dimensionality was reported in the European Nurses’ Early Exit (NEXT) Study [12], in 8 out of 10 European countries studied. These studies suggest a psychological dimension to work ability that includes a subjective appraisal of work ability and mental resources, and a physical dimension of work ability that includes an objective characterization of health based on accumulated medical conditions and injuries. One condition that may adversely influence both physical and psychological dimensions of work ability is burnout. In the current study, we defined the core feature of burnout as the presence of emotional, physical and cognitive exhaustion due to job stressors [13]. Burnout is important to the work ability of older adults, given that they appear to need more job controls to buffer the effects of job-related stressors [14], recover from stress less quickly [15] and are more sensitive to the effects of job-related stress and burnout on self-efficacy [16] and age-related cognitive decline [17]. This suggests that older workers may be sensitive to burnout symptoms, and the effects of these on work ability. This research on the interaction between work ability, age and burnout suggests that an individual’s capacity to maintain work ability with age may differ across physical and psychological dimensions, and that it may depend on the severity of burnout symptoms. In this study, we examined the association of physical and psychological work ability with age, and moderation of these associations by burnout. In doing so, we sought to confirm a bi-dimensional WAI model [11] in an independent, single-occupation sample of working nurses. This sample fits our study objectives well, as nurses are particularly vulnerable to burnout [18], based on exposure to a high level of both physical and psychological stressors [19]. Methods We recruited individuals actively working in the nursing field from a health care system in the southeast USA. Health system administration provided e-mail addresses of nurses employed across all departments. We targeted recruitment for 400 persons to achieve a sample size adequate to detect statistically significant differences. Eligible individuals were required to be actively working in the nursing field (not on medical leave, disability leave or family leave), to have at least 2 years’ experience in the nursing field and be at least 25 years old. In addition, because we conducted cognitive testing as part of our broader project, we excluded individuals with possible confounding neurological conditions (seizures, severe brain trauma and stroke). Race and sex ratios approximated the US nursing population as a whole [20], with the exception that it over-represented Black/African Americans [14%, compared to 5% in US registered nurses (RN), see Table 1]. Participants gave informed consent. This research was approved by the Institutional Review Board of Duke University. This study was based on a prospective cohort design, including assessment of burnout at baseline and of work ability at 12-month follow-up. The current study used data collected from a broader project on work stress and mental health among nurses. We collected baseline data by self-reported questionnaires, including demographic information and assessment of burnout. A trained research technician was present during completion of baseline questionnaires to provide assistance but was positioned so as not to be able to view participant responses. After baseline, we sent follow-up questionnaires monthly via e-mail with a link to a confidential survey, which concluded at month 12 in February of 2016. Although data on burnout were collected monthly, in this study, we used burnout at baseline only, to predict WAI at 12-month follow-up. We used factors derived from the 12-month follow-up assessment of the WAI [3] as the dependent measures in this study. The WAI consists of items related to physical health, including number of injuries or physician- diagnosed diseases endorsed by the participant, out of a list of 49 injuries and diseases (item 3), number of sick leave days (item 5) and an estimate of work impairment due to illness (item 4). Other items assess psychological resources (enjoyment of daily activities, alertness and hopefulness for the future; item 7), subjective estimates of current ability to work compared with lifetime best (item 1), ability to work in relation to job demands (item 2) and perceived prognosis of work ability in 2 years (item 6). We assessed burnout at baseline with the exhaustion subscale of the Oldenburg Burnout Inventory (OLBI; [13]). This subscale is composed of eight items, including, ‘I can tolerate the pressure of my work very well’ and ‘during my work, I often feel emotionally drained’. Items are assessed on a four point scale, ranging from ‘strongly agree’ to ‘strongly disagree’. We reverse-coded the four negatively worded items before calculating the mean of all eight items, yielding total scores ranging from 1 to 4. We found the internal consistency of these items to be acceptable in our sample (α = 0.81). In a recent study [21], researchers found that 40–45% of the variance in overall OLBI burnout at three waves across 3 years can be accounted for by a stable trait component, suggesting stability as well as change in burnout over time. We selected baseline age, sex, educational level and nursing certification status as covariates based on their associations with work ability in previous studies. We categorized education as: high school equivalent, associate’s degree, bachelors’ degree and master’s degree/PhD, and nursing certification as RNs versus other certification types. To confirm the factor structure of the WAI and to test the association of age and burnout to work ability at the 12-month follow-up, we used confirmatory factor analysis (CFA) with covariates models [22]. This method consists of two steps. In the first, the adequacy of a measurement model is tested. In the second, factors from this model are linearly regressed with covariates, to identify differences in factor means by covariates. In accordance with Martus et al. [11], we tested the adequacy of four CFA measurement models. In model A, we tested a one-factor solution, with all seven WAI items included. For models B and C, we entered items 1, 2 and 7, along with items 4 and 6, to reflect psychological work ability, and items 3 and 5 to reflect physical work ability. To test the assumption of factor orthogonality, we constrained the covariance between factors in model B to zero. However, this made the model under-justified. To address this, we constrained the factor loadings in this model to equality, so as to have fewer parameters to estimate. In model C, we tested the same item structure as in model B, but freely estimated the covariance between factors. We expanded this in model D, in which we cross-loaded items 4 and 6. This was done because these items assess both psychological and physical work ability, in that they assess the participants’ subjective estimation of ability in regards to the participants’ health. In step 2, we linearly regressed factors from the best-fitting model on covariates, including centred burnout and centred age at baseline (model 1); demographic covariates (sex, education and nurse type; model 2); and the interaction between burnout and age (model 3), controlling for any covariates found to be significant in model 2. For all models, we reported standardized factor loadings, factor covariances and predictor coefficients where applicable, along with 95% confidence intervals (CIs) and fit statistics [23]. We conducted descriptive statistics with SAS version 9.3 [24], and CFA models with Mplus version 7 [25]. Results The study population was 402 individuals, which included 372 RNs and 28 other nursing workers (licensed practical nurses, certified nursing assistants and medical assistants). Mean age (SD) at baseline among the 352 participants who completed follow-up at month 12 was 42.1 (11.4), with 20% being age 55 or older (n = 70). Most were female (n = 325, 92%) and had an RN certification (n = 326, 93%) with either an associate or bachelor’s degree (n = 91, 26% and n = 205, 59%). The completion rate at follow-up month 12 was 88% (n = 352). We found that loss to follow-up was not significantly related to age [t(397) = −1.52], burnout [t(400) = 1.18], sex [χ2(1) = 0.88], education [χ2(1) = 0.98], nurse type [χ2(1) = 0.02] or total WAI at baseline [t(400) = −0.40], suggesting that loss to follow-up was not biased on these variables. Table 1 contains zero-order correlations between age and other demographic characteristics, total work ability at month 12 and burnout. These indicated lower work ability based on WAI total score at month 12 to be associated with higher severity of burnout symptoms at baseline, r(350) = −0.40, P < 0.001, and greater age and female sex to be associated with lower burnout [r(347) = −0.13, P < 0.05 and t(350) = 3.32, P < 0.01]. Table 1. Bivariate relationships between age and other demographic characteristics, total work ability at month 12 and OLBI exhaustion Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Sample restricted to those who completed follow-up at month 12. 95% CI, 95% confidence intervals for Pearson correlations based on Fisher’s z transformation; non-RN, non-RN nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RN, registered nurse; total WAI, Work Ability Index total score at follow-up month 12. View Large Table 1. Bivariate relationships between age and other demographic characteristics, total work ability at month 12 and OLBI exhaustion Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Study variables Total WAI OLBI Ex Test value 95% CI P value Test value 95% CI P value Age r(347) = −0.17 −0.27, −0.06 <0.01 r(347) = −0.13 −0.23, −0.03 <0.01 Sex t(350) = −1.27 NS t(350) = 3.32 <0.001  Female 38.49, 39.66 2.57, 2.67  Male 38.43, 42.46 2.13, 2.51 Education F(3,341) = 1.64 NS F(3,341) = 2.24 NS  High school 35.59, 41.87 2.29, 2.84  Associates 37.48, 39.66 2.47, 2.66  Bachelors 38.98, 40.44 2.52, 2.65  Masters or doctorate 36.42, 39.80 2.64, 2.93 Nurse type t(350) = 0.55 NS t(350) = 0.50 NS  RN 38.64, 39.81 2.55, 2.66  Non-RN 36.40, 40.83 2.37, 2.75 OLBI Ex r(350) = −0.39 −0.48, −0.30 <0.001 – – – Sample restricted to those who completed follow-up at month 12. 95% CI, 95% confidence intervals for Pearson correlations based on Fisher’s z transformation; non-RN, non-RN nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RN, registered nurse; total WAI, Work Ability Index total score at follow-up month 12. View Large Tables 2 and 3 report correlation coefficients of study variables, and results of CFA models. Findings for CFA model A indicated that although all item loadings were significant, a one-factor solution did not adequately fit observed data [χ2(14) = 121.9, P < 0.001; standardized root mean square residual (SRMR) = 0.09; root mean square error of approximation (RMSEA) = 0.15; comparative fit index (CFI) = 0.74; Akaike information criterion (AIC) = 6866]. Findings for model B also exhibited poor fit [χ2(18) = 140.2, P < 0.001; SRMR = 0.16; RMSEA = 0.14; AIC = 6877]. Findings for model C appeared to exhibit better absolute fit [χ2(13) = 96, P < 0.001; SRMR = 0.08; AIC = 6842], though fit adjusted for model parsimony was poor (RMSEA = 0.14), as was fit compared to that of the null model (CFI = 0.80). In contrast, model D, the two-factor non-orthogonal solution with items 4 and 6 cross-loaded onto both factors, demonstrated good fit on all indices [χ2(11) = 15.4, P = 0.17; SRMR = 0.03; RMSEA = 0.03; CFI = 0.99], and comparatively greater fit than other models (AIC = 6766), including model C [likelihood ratio test: χ2(2) = 80.4, P < 0.001]. Because this measurement structure demonstrated good fit in this study and in Martus et al. [11], we used this measurement model for subsequent models of age and moderation by burnout. Table 2. Pearson correlations between variables entered into confirmatory factor analyses Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex: Oldenburg Burnout Inventory, Exhaustion subscale; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 2. Pearson correlations between variables entered into confirmatory factor analyses Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Study variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 WAI item 1: Current work ability – – – – – – – – – – – – – – –  2: Work ability–demands 0.53 – – – – – – – – – – – – – –  3: Diseases 0.09 0.16 – – – – – – – – – – – – –  4: Disease impairment 0.21 0.27 0.46 – – – – - – – – – – – –  5: Sick leave 0.14 0.12 0.32 0.28 – – – – – – – – – – –  6: Work ability prognosis 0.21 0.28 0.14 0.31 0.10 – – – – – – – – – –  7: Mental resources 0.42 0.41 0.07 0.21 0.09 0.27 – – – – – – – – – 8. Age 0.06 −0.01 −0.35 −0.07 −0.08 −0.10 0.03 – – – – – – – – 9. Male sex 0.06 0.13 0.03 0.09 0.10 0.00 0.06 0.01 – – – – – – – Educationa 10: High school 0.00 −0.05 0.00 0.07 0.05 0.00 0.03 0.03 0.26 – – – – – –  11: Associates 0.00 −0.02 −0.13 −0.06 0.00 −0.11 0.00 0.16 0.00 −0.11 – – – – –  12: Bachelors 0.01 0.08 0.18 0.06 0.03 0.10 0.04 −0.26 −0.04 −0.21 −0.70 – – – –  13: Masters/PhD −0.01 −0.04 −0.08 −0.04 −0.01 0.02 −0.06 0.11 −0.07 −0.06 −0.21 −0.42 – – – 14. Nurse type: non-RN −0.03 −0.11 −0.01 0.09 0.06 −0.07 −0.03 −0.01 0.18 0.66 −0.03 −0.18 −0.02 – – 15. OLBI Ex −0.32 −0.35 −0.15 −0.31 −0.09 −0.20 −0.43 −0.13 −0.17 −0.01 −0.04 −0.05 0.14 −0.05 – Non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex: Oldenburg Burnout Inventory, Exhaustion subscale; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 3. Confirmatory factor analysis of WAI Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Parameter estimate (est.) and 95% CI reported for each factor loading. AIC, Akaike information criterion; CFI, comparative fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aModel A: one-factor solution. bModel B: two-factor solution, 4 and 6 on WAI psychological, factor covariance constrained to 0. cModel C: two-factor solution, 4 and 6 on WAI psychological, factor covariance freely estimated. dModel D: two-factor solution, 4 and 6 on both factors, factor covariance freely estimated. View Large Table 3. Confirmatory factor analysis of WAI Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Model Aa Model Bb Model Cc Model Dd Est. 95% CI Est. 95% CI Est. 95% CI Est. 95% CI WAI item factor loadings  1: Current work ability 0.67 0.58, 0.75 0.69 0.60, 0.78 0.68 0.59, 0.76 0.70 0.62, 0.79  2: Work ability–demands 0.71 0.63, 0.80 0.62 0.56, 0.69 0.73 0.64, 0.81 0.74 0.65, 0.82  7: Mental resources 0.57 0.48, 0.66 0.60 0.54, 0.66 0.58 0.48, 0.67 0.58 0.49, 0.68  4: Disease impairment 0.45 0.34, 0.56 0.54 0.48, 0.59 0.42 0.31, 0.53 0.18 0.03, 0.32  6: Work ability prognosis 0.42 0.31, 0.53 0.34 0.30, 0.39 0.41 0.30, 0.52 0.32 0.20, 0.45  3: Diseases 0.30 0.18, 0.42 0.44 0.36, 0.52 0.62 0.42, 0.83 0.68 0.55, 0.82  5: Sick leave 0.25 0.13, 0.37 0.72 0.58, 0.85 0.51 0.33, 0.68 0.43 0.31, 0.54  4: Disease impairment – – – 0.64 0.49, 0.80  6: Work ability prognosis – – – 0.20 0.06, 0.33 Factor covariance – 0 – 0.42 0.25, 0.59 0.27 0.10, 0.45 Fit statistics  χ2, df, P value 121.9, 14, <0.001 140.2, 18, <0.001 95.8, 13, <0.001 15.4, 11, 0.17  SRMR 0.09 0.16 0.08 0.03  RMSEA, 95% CI 0.15, 0.12–0.17 0.14, 0.12–0.16 0.14, 0.11–0.16 0.03, 0.00–0.07  CFI 0.74 0.70 0.80 0.99  AIC 6866 6877 6842 6766 Parameter estimate (est.) and 95% CI reported for each factor loading. AIC, Akaike information criterion; CFI, comparative fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aModel A: one-factor solution. bModel B: two-factor solution, 4 and 6 on WAI psychological, factor covariance constrained to 0. cModel C: two-factor solution, 4 and 6 on WAI psychological, factor covariance freely estimated. dModel D: two-factor solution, 4 and 6 on both factors, factor covariance freely estimated. View Large Table 4 reports findings from CFA with covariates models regressing WAI physical and WAI psychological on age, burnout, demographic covariates and the interaction between age and baseline burnout. Greater burnout was associated with diminished physical and psychological work ability (β = −0.25, 95% CI −0.36, −0.13 and β = −0.55, 95% CI −0.65, −0.46), while older age was associated with diminished physical work ability only (β = −0.41, 95% CI −0.51, −0.31). Fit statistics for this model indicated acceptable fit [χ2(21) = 52.1, P < 0.001; SRMR = 0.04; RMSEA = 0.07; CFI = 0.94]. Findings did not change markedly with inclusion of covariates (model 2), though the effect for nurse certification was significant, in that nurses with certifications other than RN exhibited diminished psychological work ability (β = −0.17, 95% CI −0.32, −0.03). Because nurse certification was significant, it was retained in model 3, which tested the interaction between age and burnout. Table 4. Confirmatory factor analysis of WAI psychological and WAI physical on OLBI exhaustion, age, demographics and OLBI exhaustion by age interaction Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Loadings for WAI psychological and WAI physical factors available upon request. Parameter estimate (Est.) and 95% CI reported for each coefficient or estimate of factor covariance. AIC, Akaike information criterion; CFI, comparative fit index; non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large Table 4. Confirmatory factor analysis of WAI psychological and WAI physical on OLBI exhaustion, age, demographics and OLBI exhaustion by age interaction Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Model 1 Model 2 Model 3 Est. 95% CI Est. 95% CI Est. 95% CI Coefficients  WAI psychological on   OLBI Ex −0.55 −0.65, −0.46 −0.56 −0.66, −0.46 −0.56 −0.65, −0.46   Age −0.02 −0.14, 0.09 −0.02 −0.13, 0.10 −0.01 −0.11, 0.10   OLBI Ex × age − − − −0.19 −0.29, −0.08   Sex − − 0.03 −0.08, 0.14 – –   Educationa    High school − − 0.17 −0.04, 0.37 − −    Associates − − 0.14 −0.23, 0.52 − −    Bachelors − − 0.21 −0.21, 0.62 − −    Masters/PhD − − 0.16 −0.12, 0.44 − −   Nurse type (non-RN) − − −0.17 −0.32, −0.03 −0.10 −0.21, 0.01 R2 0.30 0.32 0.35  WAI physical on   OLBI Ex −0.25 −0.36, −0.13 −0.24 −0.36, −0.12 −0.25 −0.37, −0.13   Age −0.41 −0.51, −0.31 −0.39 −0.49, −0.28 −0.39 −0.50, −0.29   OLBI Ex × age − − − −0.16 −0.27, −0.05   Sex − − 0.01 −0.11, 0.12 − −   Educationa    High school − − 0.08 −0.13, 0.29 − −    Associates − − 0.03 −0.36, 0.41 − −    Bachelors − − 0.14 −0.29, 0.56 − −    Masters/PhD − − 0.06 −0.23, 0.34 − −   Nurse type (non-RN) − − −0.03 −0.18, 0.12 0.004 −0.11, 0.12 R2 0.20 0.21 0.23 Factor covariance 0.10 −0.05, 0.26 0.09 −0.06, 0.25 0.07 −0.10, 0.23 Fit statistics  χ2, df, P value 52.15, 21, <0.001 84.52, 51, <0.01 65.98, 31, <0.001  SRMR 0.04 0.03 0.04  RMSEA (95% CI) 0.07 (0.04, 0.09) 0.04 (0.03, 0.06) 0.057 (0.04, 0.08)  CFI 0.94 0.94 0.94  AIC 6589.0 6602.4 6574.3 Loadings for WAI psychological and WAI physical factors available upon request. Parameter estimate (Est.) and 95% CI reported for each coefficient or estimate of factor covariance. AIC, Akaike information criterion; CFI, comparative fit index; non-RN, non-registered nursing professional: certified nursing assistant, licensed practical nurse and medical assistants; OLBI Ex, Oldenburg Burnout Inventory, Exhaustion subscale; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; WAI, Work Ability Index at follow-up month 12. aEach category of education tested as a dummy-coded variable. View Large In model 3, we found significant interactions between age and baseline burnout for both physical work ability (β = −0.16, 95% CI −0.27, −0.05) and psychological work ability (β = −0.19, 95% CI −0.29, −0.08), with fit statistics indicating acceptable fit [χ2(31) = 66.0, P < 0.001; SRMR = 0.04; RMSEA = 0.057, 95% CI 0.04–0.08; CFI = 0.94]. To interpret the interactions, we tested and plotted simple slopes for age in predicting work ability outcomes at different conditional values of burnout (the mean, and 1.0 and 1.5 SDs below and above the mean). As shown in Figure 1, findings revealed that at low levels of burnout (1 SD below mean), and at mean and high (1.5 SDs above mean) levels of burnout, older age predicted diminished physical work ability (1 SD below: β = −0.23, 95% CI −0.39, −0.06; mean: β = −0.39, 95% CI −0.50, −0.29; 1 SD above: β = −0.56, 95% CI −0.70, −0.42; 1.5 SD above: β = −0.64, 95% CI −0.83, −0.46), while at very low levels of burnout (1.5 SD below mean), age was unrelated to physical work ability (β = −0.14, 95% CI −0.36, 0.07). With respect to psychological work ability (see Figure 2), results indicated that at mean levels of burnout, age did not predict psychological work ability (β = −0.01, 95% CI −0.11, 0.10); however, at higher levels of burnout (1 and 1.5 SDs above the mean of OLBI Ex), older age predicted diminished psychological work ability (1 SD above: β = −0.20, 95% CI −0.35, −0.05; 1.5 SD above: β = −0.30, 95% CI −0.49, −0.11), while at lower levels of burnout (1 and 1.5 SDs below the mean), older age predicted increased psychological work ability (1 SD below: β = 0.19, 95% CI 0.03, 0.35; 1.5 SD below: β = 0.29, 95% CI 0.08, 0.50). Figure 1. View largeDownload slide Physical work ability on age at conditional values of exhaustion burnout. Figure 1. View largeDownload slide Physical work ability on age at conditional values of exhaustion burnout. Figure 2. View largeDownload slide Psychological work ability on age at conditional values of exhaustion burnout. Figure 2. View largeDownload slide Psychological work ability on age at conditional values of exhaustion burnout. Discussion The current study confirmed a bi-dimensional factor structure for the WAI representing physical and psychological dimensions and extended this with two novel findings. Firstly, we found differences in physical and psychological dimensions for age and burnout, with older age being associated with lower physical work ability only, and greater burnout being associated with lower physical and psychological work ability. Secondly, we found that age and burnout interact. In the psychological dimension, older age predicted lower work ability at high levels of burnout, whereas at low levels of burnout, older age predicted higher work ability. In the physical dimension, older age generally predicted lower work ability, with the exception of no association at very low levels of burnout symptoms. The study has both strengths and limitations. We conducted assessments on a large cohort of actively working health professionals at baseline and 12-month follow-up. With a mean age of 42 years, many of the older workers in our sample are closer to midlife than later life; however, this is consistent with evidence linking reduced work ability and burnout in mid-career to higher risk for early retirement [7], and with approaches emphasizing interventions in midlife to maintain health and ability in late life [26]. We were limited in our use of convenience sampling of individuals who responded promptly to recruiting e-mails. This may have introduced unintended bias toward healthy, non-burned-out workers, but our descriptive data indicate an adequate range of burnout scores comparable to a sample of Norwegian nurses [27], and a demographic composition consistent with that of the nursing workforce in general. We note this study uses only the exhaustion component of burnout. Although exhaustion is theorized to be the core component of burnout, in the sense that it precedes other aspects of burnout [28], future studies would benefit from examining other components of burnout such as disengagement or personal accomplishment. Finally, our findings also highlight the need to examine how such interventions generalize to other human service professions. The current study supports previous research suggesting the principal metric of the WAI—the total score—may be better represented as a bi-dimensional measure [11,12], particularly with respect to age and severity of burnout. This is consistent with the broader literature suggesting overall work ability declines with age [7,29], but it highlights important exceptions. In line with the broader literature, physical work ability was lower with increasing age, and this relationship worsened with greater burnout, and, at higher levels of burnout, psychological work ability was also lower with increasing age. This suggests that at high burnout, older workers may not be as able to direct adequate physical and psychological resources toward their job duties. This is consistent with evidence finding older individuals need more job resources as a buffer to work stress [14], and recover less quickly from stress [15]. In contrast to the broader literature, at low levels of burnout, older workers exhibited greater psychological work ability. This suggests that under optimal conditions, older workers may be more able to engage cognitive, psychological and occupational strengths that improve with age [9,30]. Our study findings suggest that the WAI total score provides an incomplete clinical impression of the work ability of older individuals, who appear to have more resilience on psychological aspects of work ability under low-to-medium levels of burnout, but less resilience on physical aspects overall. This suggests that deficits in physical health are the most consistent contributors to decline in work ability with age, and that interventions targeting disease prevention and health maintenance from midlife and earlier may be effective ways to maintain work ability with age. In addition, support for mental/emotional resources in the workplace may provide an important buffer against burnout and its effect on the work ability of older individuals. Key points This study confirmed a bi-dimensional structure in the Work Ability Index in the form of physical and psychological work ability. It also found age relationships with these dimensions that were moderated by exhaustion-related burnout. Findings emphasize the need to reduce burnout and to address age-related strengths and vulnerabilities relating to physical and psychological work ability. Funding This work was supported by the German Federal Institute for Occupational Safety and Health (Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, BAuA: Project F2318). Support for Daniel Hatch was provided by the National Institutes of Health (NIH Grant No. T32-AG000029). Competing interests None declared. References 1. Social Security Administration . Retirement Planner: Benefits by Year of Birth . 2016 . Woodlawn, MD : United States Social Security Administration. https://www.ssa.gov/planners/retire/agereduction.html 2. He W , Goodkind D , Kowal P. International Population Reports, P95/16-1, An Aging World: 2015 . Washington, DC : US Census Bureau , 2016 . 3. Tuomi K , Ilmarinen J , Jahkola A , Katjarinne L , Tulkki A. Work Ability Index . Helsinki, Finland : Finnish Institute of Occupational Health , 1998 . 4. Bethge M , Spanier K , Neugebauer T , Mohnberg I , Radoschewski FM . Self-reported poor work ability—an indicator of need for rehabilitation? A cross-sectional study of a sample of German employees . Am J Phys Med Rehabil 2015 ; 94 : 958 – 966 . Google Scholar CrossRef Search ADS PubMed 5. Roelen CA , Heymans MW , Twisk JW , van der Klink JJ , Groothoff JW , van Rhenen W . Work Ability Index as tool to identify workers at risk of premature work exit . J Occup Rehabil 2014 ; 24 : 747 – 754 . Google Scholar CrossRef Search ADS PubMed 6. Tuomi K , Huuhtanen P , Nykyri E , Ilmarinen J . Promotion of work ability, the quality of work and retirement . Occup Med (Lond) 2001 ; 51 : 318 – 324 . Google Scholar CrossRef Search ADS PubMed 7. Camerino D , Conway PM , Van der Heijden BI et al. ; NEXT-Study Group . Low-perceived work ability, ageing and intention to leave nursing: a comparison among 10 European countries . J Adv Nurs 2006 ; 56 : 542 – 552 . Google Scholar CrossRef Search ADS PubMed 8. Kaiser RM . Physiological and clinical considerations of geriatric patient care . In: Steffens DC , Blazer DG , Thakur ME , eds. The American Psychiatric Publishing Textbook of Geriatric Psychiatry . 5th edn . Arlington, VA : American Psychiatric Publishing , 2015 ; 33 – 59 . Google Scholar CrossRef Search ADS 9. Carstensen LL , Fung HH , Charles ST . Socioemotional selectivity theory and the regulation of emotion in the second half of life . Motiv Emot 2003 ; 27 : 103 – 123 . Google Scholar CrossRef Search ADS 10. Umanath S , Marsh EJ . Understanding how prior knowledge influences memory in older adults . Perspect Psychol Sci 2014 ; 9 : 408 – 426 . Google Scholar CrossRef Search ADS PubMed 11. Martus P , Jakob O , Rose U , Seibt R , Freude G . A comparative analysis of the Work Ability Index . Occup Med (Lond) 2010 ; 60 : 517 – 524 . Google Scholar CrossRef Search ADS PubMed 12. Radkiewicz P , Widerszal-Bazyl M . Psychometric properties of Work Ability Index in the light of comparative survey study . Int Congr Ser 2005 ; 1280 : 304 – 309 . Google Scholar CrossRef Search ADS 13. Demerouti E , Bakker AB , Nachreiner F , Schaufeli WB . The job demands-resources model of burnout . J Appl Psychol 2001 ; 86 : 499 – 512 . Google Scholar CrossRef Search ADS PubMed 14. Shultz KS , Wang M , Crimmins EM , Fisher GG . Age differences in the demand-control model of work stress: an examination of data from 15 European countries . J Appl Gerontol 2010 ; 29 : 21 – 47 . Google Scholar CrossRef Search ADS PubMed 15. Kiss P , De Meester M , Braeckman L . Differences between younger and older workers in the need for recovery after work . Int Arch Occup Environ Health 2008 ; 81 : 311 – 320 . Google Scholar CrossRef Search ADS PubMed 16. Shoji K , Cieslak R , Smoktunowicz E , Rogala A , Benight CC , Luszczynska A . Associations between job burnout and self-efficacy: a meta-analysis . Anxiety Stress Coping 2016 ; 29 : 367 – 386 . Google Scholar CrossRef Search ADS PubMed 17. Andel R , Infurna FJ , Hahn Rickenbach EA , Crowe M , Marchiondo L , Fisher GG . Job strain and trajectories of change in episodic memory before and after retirement: results from the health and retirement study . J Epidemiol Community Health 2015 ; 69 : 442 – 446 . Google Scholar CrossRef Search ADS PubMed 18. Demerouti E , Bakker AB , Nachreiner F , Schaufeli WB . A model of burnout and life satisfaction amongst nurses . J Adv Nurs 2000 ; 32 : 454 – 464 . Google Scholar CrossRef Search ADS PubMed 19. Aiken LH , Sloane DM , Bruyneel L , Van den Heede K , Sermeus W ; RN4CAST Consortium . Nurses’ reports of working conditions and hospital quality of care in 12 countries in Europe . Int J Nurs Stud 2013 ; 50 : 143 – 153 . Google Scholar CrossRef Search ADS PubMed 20. US Department of Health and Human Services . The Registered Nurse Population: Findings From the 2008 National Sample Survey of Registered Nurses . Rockville, MD : Health Resources and Services Administration , Ch. 7; 2010 ; 2 – 7 . PubMed PubMed 21. Brauchli R , Schaufeli WB , Jenny GJ , Füllemann D , Bauer GF . Disentangling stability and change in job resources, job demands, and employee well-being—a three-wave study on the job-demands resources model . J Vocational Behav 2013 ; 83 : 117 – 129 . Google Scholar CrossRef Search ADS 22. Jöreskog KG , Goldberger AS . Estimation of a model with multiple indicators and multiple causes of a single latent variable . J Am Stat Assoc 1975 ; 70 : 631 – 639 . 23. Hu L , Bentler PM . Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives . Struct Eq Model: Multidisciplin J 1999 ; 6 : 1 – 55 . Google Scholar CrossRef Search ADS 24. SAS Institute Inc . SAS Release 9.3 . Cary, NC : SAS Institute Inc . 2012 . 25. Muthén LK , Muthén BO. Mplus User’s Guide . 7th edn . Los Angeles, CA : Muthén & Muthén , 1998–2015 . 26. Szoeke C , Lehert P , Henderson VW , Dennerstein L , Desmond P , Campbell S . Predictive factors for verbal memory performance over decades of aging: data from the women’s healthy ageing project . Am J Geriatr Psychiatry 2016 ; 24 : 857 – 867 . Google Scholar CrossRef Search ADS PubMed 27. Innstrand ST , Langballe EM , Falkum E , Aasland OG . Exploring within- and between-gender differences in burnout: 8 different occupational groups . Int Arch Occup Environ Health 2011 ; 84 : 813 – 824 . Google Scholar CrossRef Search ADS PubMed 28. Maslach C , Schaufeli WB , Leiter MP . Job burnout . Annu Rev Psychol 2001 ; 52 : 397 – 422 . Google Scholar CrossRef Search ADS PubMed 29. Ilmarinen J , Tuomi K . Work ability of aging workers . Scand J Work Environ Health 1992 ; 18 ( Suppl. 2 ): 8 – 10 . Google Scholar PubMed 30. Ng TW , Feldman DC . The relationship of age to ten dimensions of job performance . J Appl Psychol 2008 ; 93 : 392 – 423 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Occupational MedicineOxford University Press

Published: Mar 26, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off