Adjustment latitude and attendance requirements as determinants of sickness absence or attendance. Empirical tests of the illness flexibility model

https://doi.org/10.1016/S0277-9536(03)00407-6Get rights and content

Abstract

This study investigates whether the two dimensions of illness flexibility at work, adjustment latitude and attendance requirements are associated to sickness absence and sickness attendance. Adjustment latitude describes the opportunities people have to reduce or in other ways change their work-effort when ill. Such opportunities can be to choose among work tasks or work at a slower pace. Attendance requirements describe negative consequences of being away from work that can affect either the subject, work mates or a third party. In a cross-sectional design data based on self-reports from a questionnaire from 4924 inhabitants in the county of Stockholm were analysed. The results showed that low adjustment latitude, as predicted, increased women's sickness absence. However, it did not show any relation to men's sickness absence and men's and women's sickness attendance. Attendance requirements were strongly associated to both men's and women's sickness absence and sickness attendance in the predicted way. Those more often required to attend were less likely to be absent and more likely to attend work at illness. As this is the first study of how illness flexibility at work affects behaviour at illness, it was concluded that more studies are needed.

Introduction

What determines people's decision to go sick or go to work when they wake up with e.g. a sore throat or an aching back? The choice is influenced by symptoms such as pain location and intensity. Other aspects are, however, likely to impinge on people's decision to be absent or attend work when ill.

A variety of social, psychological and economic aspects have been associated with sickness absence (North et al., 1993; Knutsson & Goine, 1998; Vahtera, Kivimäki, Pentti, & Theorell, 2000; Voss, Floderus, & Diderichen, 2001), while sickness attendance—going to work in spite of illness—has been rather neglected in research. A few studies, though, have linked it with social, psychological and economic aspects (McKevitt, Morgan, Dundas, & Holland, 1997; Aronsson, Gustafsson, & Dallner, 2000), one interpretation being that these non-health aspects do affect the decision.

As sickness attendance (termed sickness presenteeism by Aronsson et al., 2000) is little researched the following will focus on sickness absence.

Nicholson (1977) categorises theories of sickness absence in three explanatory models:

  • 1.

    Pain-avoidance models in which absence is portrayed as a flight from negatively valued aspects of work experience.

  • 2.

    Adjustment and maladjustment models in which absence is viewed as an outcome of organisational socialisation and other adaptive processes to job demands

  • 3.

    Decision models in which absence is described as a rational decision or choice directed toward the attainment of valued goals.

Nicholson discerns two distinct and relatively unconnected schools of thought among predecessors of decision models, the rationalist writers of economics and sociology and the motivation theorists of organisational psychology. Both groups portray absence as a rational behaviour, determined by the absentee's subjective evaluation of the costs and benefits associated with probable outcomes of the alternative behaviour: attendance.

The decision model proposed here differs from the models described by Nicholson by not locating the causes of sickness absence (and sickness attendance) in the subject's evaluation of costs and benefits. Instead our model locates the causes in constraints, both contextual and individual, that may limit people's actions. The perspective proposed here meets one requirement put forward by Kristensen (1991) for an integrated theory: “A theory of sickness absence should consider the individual as a product of his or her environment and, at the same time, as a conscious actor who makes choices within a given social framework”.

Sickness absence (and sickness attendance) are by definition actions attributable to sickness (Marmot, Feeney, Shipley, North, & Syme, 1995). Despite this in previous models of sickness absence ill health is given the same status as other determinants of sickness absence (Ås, 1962; Steers & Rhodes, 1978; Kristensen, 1991). One exception is Nicholson's own model (1977). However, in this model illness and other absence-inducing events are merged with work conditions into an A–B continuum that describes potential avoidability of absence. Individual choice is irrelevant to absences at the A end, while those at the B end are entirely subject to individual choice. Relevant work aspects, determining work ability, are not discerned.

We propose that ill health be treated as a prerequisite cause of sickness absence and sickness attendance and that work conditions affecting people's choice be described separately from the health problem.

Here a model, “the model of illness flexibility” (Fig. 1) is presented. It emphasises the choice people have to make between going sick and going to work when they feel ill. Ill health is thus the starting point for the model. The model assumes that life situations involve different possibilities to embrace ill health by giving different opportunities of remaining at work or being absent. Sickness absence can have other causes than ill health, both “legitimate” and “illegitimate”. Examples of “legitimate” reasons for sickness absence not usually covered by the concept “ill health” are life crises, grief and pregnancy (Nordenfelt, 1986). Examples of “illegitimate” reasons are to be off sick to watch the Olympics on television. The present model will not encompass “illegitimate” sickness absence since, first, the alternative action, to go to work, cannot be considered as sickness attendance. Secondly, such reasons play but a minor part in explaining sickness absence (Vahtera, Kivimäki, & Penti, 2001).

Being absent from work or not when in e.g. grief and life crises is likely to have the same determinants as action prompted by ill health. Further, the alternative action, attendance, may theoretically be considered as sickness attendance. It is therefore desirable that a model should include these “legitimate” reasons for sickness absence. We therefore propose that a loss of function should be considered common to absence-inducing situations, and “loss of function” will be used to describe absence-inducing situations. However, most absence-inducing situations will be due to ill health.

Loss of function will affect work ability but work ability is likely to be determined also by the work conditions people meet. While some people may always have to work fully if they attend work others may be able to choose among work tasks, work at a slower pace or shorten their working day. Adjustment latitude is a central concept in the model. It describes the opportunities people have to reduce or in other ways alter their work effort when e.g. feeling ill. The likelihood of retaining the ability to work should be greater where there is high adjustment latitude compared to where there is low. Work ability is thus seen as both individually and contextually determined.

Work ability is associated with sickness absence. However, whether people actually choose to be absent or not when their work ability is low can also have other causes. A further central concept in the model are the “attendance requirements”. This aims to describe major determinants of the final decision to attend or not.

Attendance requirements describe the negative consequences of absence for e.g. the individual, work-mates or a third party e.g. clients, care recipients, pupils, etc. These requirements may originate both at and outside work. When one is absent, work tasks might accumulate, work-mates might get more to do, or activities are cancelled. Aronsson et al. (2000) found that members of occupational groups whose everyday tasks are to provide care or welfare services, or teach or instruct have a substantially increased risk of being at work when sick. Such increased risk was also found among people who could not be replaced when absent. Both associations can be understood by the concept attendance requirements.

Other aspects that constitute attendance requirements are a weak financial situation and a weak position on the labour market. Internal standards governing what is the right or the wrong way to behave when e.g. ill may also constitute attendance requirements.

The model of illness flexibility has not been studied empirically. The present study aimed to examine whether “adjustment latitude” and “attendance requirements” are associated with sickness absence and sickness attendance. Attendance requirements originating elsewhere than from work were not considered in this study, which accordingly tested the following hypotheses:

  • Low adjustment latitude is associated with higher sickness absence and lower sickness attendance than high adjustment latitude.

  • High attendance requirements are associated with lower sickness absence and higher sickness attendance than low attendance requirements.

As these associations may differ for shorter and longer sickness absence and attendance, days of absence will be taken in account.

Health is strongly related to both sickness absence (Marmot et al., 1995) and sickness attendance (Aronsson et al., 2000), possibly also to adjustment latitude and attendance requirements.

Exhausting family demands can increase the probability of being absent from work. While responsibility for young children and/or many children, has been associated with high absence among women (Blank & Diderichsen, 1995; Chevalier, Luce, Blanc, & Goldberg, 1987), other studies have found no association between parenthood and sickness absence (Mastekaasa, 2000).

Low adjustment latitude may also be associated with small possibilities to adapt family life depending on health and illness. This can be structured by e.g. gender or social class. Women and people of lower socio-economic status may be less able to adjust both work and family life in times of illness. Also, work with high attendance requirements necessitates an undemanding family life.

Financial position is likely to affect an individual's decision whether to be absent when ill. The findings are, however, opposite from what can be expected: financial difficulties in fact increase the tendency to be absent when ill (Blank & Diderichsen, 1995; North et al., 1993). A possible explanation may be that financial difficulties are associated with small adjustment latitude and low attendance requirements. If so this association might originate from financial position and illness flexibility that had joint causes e.g. socio-economic position, gender, and branch of business.

As adjustment latitude entirely, and work attendance requirements partly, are organisationally determined, they may be associated, and either of these aspects of work might affect the relation to sickness absence/attendance of the other. Attendance requirements should therefore be controlled for when analysing adjustment latitude and vice versa.

We assumed that differences in illness flexibility at work, together with health, financial position, labour market position and demands at home mediate the structural aspects previously known to be associated with sickness absence such as socio-economic position and size of workplace (North et al., 1993; Voss et al., 2001). We did not therefore control for the impact of these aspects directly.

Low sickness absence can be given at least two different interpretations. It can reflect good health, but it can also reflect high sickness attendance (McKevitt et al., 1997). This is, however, often neglected in studies of sickness absence. The same goes for sickness attendance. There will be “healthy” individuals that have not experienced an absence-inducing situation during the period studied and who consequently report neither sickness absence nor sickness attendance. Since they have not made the choice, they do not belong to the population for which the model of illness flexibility is valid. Being in good health they may simply have had no reason to consider illness flexibility in their job. We have therefore excluded those reporting neither sickness absence nor sickness attendance. Fig. 2 displays the group excluded.

Section snippets

Material

The present study population includes subjects from two samples. The first, a follow-up study, was conducted in two steps. In 1994 about 13 500 randomly selected inhabitants, aged 22–66, from the county of Stockholm received postal questionnaires, 8539 of which were returned. In 1998 a new version of the questionnaire was sent to those that answered the original in 1994. Exclusion of those that had died, moved abroad or for other reasons were not able to answer the questions gave a net selection

Descriptive analysis

Table 1 shows how sickness absence was distributed on the independent variables among men and women. Generally, a smaller proportion among women than men reported no sickness absence during the year studied. They also had a bigger proportion that reported being sick absent 8 days or more. A bigger proportion among women than men also reported being absent between 1 and 7 days during the studied year. However, the difference between men and women were smaller in this group compared to the other

Discussion

As predicted, low adjustment latitude at work was associated with an increased risk of sickness absence among both sexes. However, the association was rather weak. Adjustment latitude seemed to have little effect on sickness attendance: there was even a slight tendency for those who had to work full-time to go to work more often despite illness than those who could reduce their work effort.

Having high attendance requirements markedly lessened the probability of absence and increased the

Conclusion

Of the four associations studied here, two strongly support the hypothesis, one gives modest support and one none. Illness flexibility at work thus seems to be important in determining whether illness and other reasons for loss of capacity result in sickness absence or sickness attendance. However, the results are sometimes not easily interpreted and more studies on illness flexibility in relation to sickness absence and sickness attendance are needed. One important area for such studies is the

Acknowledgements

Financial support was provided by Swedish Council for Working Life and Social Research, and by Alecta Pension Insurance, Mutual.

References (20)

There are more references available in the full text version of this article.

Cited by (191)

View all citing articles on Scopus
View full text