Article Text
PDF
Abstract
Objectives We undertook a register-based cohort study to evaluate exposure–response relations between cumulative occupational mechanical exposures, and risk of lateral and medial inguinal hernia repair.
Methods Among all men born in Denmark between 1938 and 1988, we established a cohort comprising those aged 18–65 years of age, who had at least 1 year of full-time employment between 1993 and 2007. Using information from a Job Exposure Matrix based on expert judgement and year-by-year information on Danish International Standard Classification of Occupations codes for each individual since 1993, we established time-varying cumulative estimates of exposure to daily lifting activities and standing/walking. Cumulative exposures for lagged 5-year time windows were expressed in a way that corresponds to the pack-year concept of smoking (ton-years, frequent-heavy-lifting years, and standing-years). First-time inguinal hernia repairs in the period 1998–2008 were identified in the Danish Hernia Database. We used a logistic regression technique equivalent to survival analysis, adjusting for age, socioeconomic status, region of residence and calendar year.
Results Within the cohort of 1 545 987 men, we identified 22 926 lateral, 15 877 medial and 1592 pantaloon or unspecified first-time inguinal hernia repairs. The risk of lateral hernia repair increased with ton-years, frequent-heavy-lifting-years, and standing-years, with ORs of up to around 1.4. The exposures correlated, but standing-years remained as the most robust risk factor after adjustment for lifting exposures. In general, the risk of medial hernia repair was unrelated to the exposures.
Conclusions Our findings suggest an increased risk of lateral inguinal hernia repair in relation to occupational mechanical exposures and a preventive potential of around 15% of all cases.
- Inguinal hernia
- Herniorrhaphy
- Occupational exposure
- Lifting
- Risk factor
Statistics from Altmetric.com
What this paper adds
-
Evidence of occupational mechanical exposures as risk factors for inguinal hernias is sparse.
-
The risk of lateral inguinal hernia repair was associated with cumulative exposures to daily lifting activities (total load, frequent heavy lifting) and prolonged standing/walking.
-
The risk of medial inguinal hernia repair was unrelated to these exposures, except that results suggested a relation with cumulative exposure to prolonged standing/walking above a certain threshold.
-
Combination of medial and lateral inguinal hernias in aetiologic research may mask specific relations with risk factors.
Introduction
An inguinal hernia is a protrusion of contents of the abdominal cavity through a defect in the transversalis fascia above the inguinal ligament.1 Medial hernias penetrate through a non-preformed defect, whereas, lateral hernias pass through the inguinal canal. Except in case of serious comorbidities, surgical treatment of symptomatic inguinal hernias is usually recommended, whereas, watchful waiting may be an acceptable option for asymptomatic or minimally symptomatic hernias.1
Repair of inguinal hernia is one of the most common operations in general surgery and is connected with considerable individual and societal costs.2 In Denmark, over the last decade, around 10 000 inguinal hernia repairs have been registered each year, of which 88.1% were performed among men aged 18–65 years.3 For inguinal hernia repair, a lifetime prevalence of 27.2% among males, and 2.6% among females has been reported.4
Considering the high incidence, and the fact that surgical repair is often performed during the patients’ working lives, surprisingly little is known about the aetiological factors, including the impact of occupational mechanical exposures on risk. It is often assumed that heavy physical work, or single strenuous events can cause inguinal hernia,5 and the Danish National Board of Industrial Injuries has provided guidelines for recognising hernias as industrial injuries.6 A recent systematic review concluded that there is insufficient epidemiologic evidence to conclude whether there are causal associations between specific occupational mechanical exposures and the development of inguinal hernia.7 The conclusion was based on eight studies.8–15 The most important limitations of the studies were insufficient exposure assessment and unspecific outcome assessment.7 One cross-sectional study simply compared two groups defined by job titles,13 another focussed on hernias that were assumed to be work-related.11 A third cross-sectional study did not find any significant associations between clinically diagnosed inguinal hernia and physical activity at work, but physical activity at work was not included in multivariable models, and risk estimates were not presented.8 None of the remaining five studies9 ,10 ,12 ,14 ,15 analysed lateral and medial hernias separately, two of the studies also included femoral hernias,9 ,10 and only two studies explicitly restricted the outcome definition to first-time hernias.14 ,15 One case-control study, where a small job exposure matrix (JEM) was used, showed indications of an increasing risk with increasing physical efforts at work, but results were not adjusted for potential confounders.10 Another case-control study relied on self-reported exposures, which means that recall bias might well explain the observed positive associations.9 Two studies simply compared cases and controls with respect to average physical work activity that hardly differed between groups.12 ,14 In the only prospective study using self-reported crude estimates of physical activity during work, 44% of the population was classified as very active, and exposure misclassification seems to be a likely explanation of the negative results.15
The injury mechanism is thought to involve increased abdominal pressure16 ,17 leading to the opening of the inguinal canal, or deterioration of abdominal muscles18 or the transversalis fascia. Having a job that requires standing for long periods has been stated as a risk factor,19 but we have not been able to identify studies that provide evidence for this association. Although there may be little clinical merit in differentiating between medial and lateral hernias,2 the influence of occupational risk factors may differ. This has not been studied.
The aim of the present study was to evaluate the hypothesis that exposure–response relations exist between cumulative measures of specific mechanical exposures in the working environment (total load lifted per day, daily frequency of lifting loads weighing 20 kg or more, and number of hours per day spent standing/walking), and risk of lateral and medial inguinal hernia repair.
Methods
We conducted a population-based male cohort study using data from administrative and medical registers in Denmark, that is, the Danish Civil Registration System (CRS),20 the Employment Classification Module (ECM),21 the Danish National Patient Register (NPR),22 and the Danish Hernia Database (DHDB).23–25 For the purpose of the present study, DHDB has the advantage that a distinction is made between medial and lateral inguinal hernias, which is not the case in NPR. DHDB was established in 1998. The coverage of DHDB has been reported to be 97.1% when compared with NPR.26
Population
The risk cohort included all men born in Denmark, excluding Greenland, between 1 January 1938 and 31 December 1988, alive and living in Denmark as on 1 January 1998 according to CRS, and not diagnosed with inguinal hernia at a hospital between 1 January 1977 and 31 December 1997 according to NPR. Until 31 December 1993, the diagnostic system used was the International Classification of Diseases, 8th revision (ICD-8), and from 1 January 1994, the International Classification of Diseases, 10th revision (ICD-10) was used. A diagnosis of inguinal hernia was identified by ICD-8 codes 550.00–550.09 and 552.09–552.99 and ICD-10 codes K40.0–K40.9.
Outcome assessment
For all cohort members, we identified first-time events of inguinal hernia repair between 1 January 1998 and 31 December 2008 in DHDB. Depending on the localisation of the hernia observed during the operation, the hernias were divided into medial, lateral and other hernias (pantaloon hernias and hernias with missing information on type). Patients who underwent surgery for bilateral hernias counted once in the analyses if both hernias were either medial or lateral (n=778). If the hernias were of different types, the patient was included in the analyses for both hernia types (n=91). We restricted our analyses to repairs of medial and lateral hernias, respectively. The outcomes were further characterised as right- or left-sided.
Exposure assessment
From ECM, we obtained individual information on occupational codes according to the Danish version of the International Standard Classification of Occupations from 1988 (DISCO-88) for each job held since 1 January 1993 (ECM does not hold occupational titles). For each year from 1993 and onwards, we allocated exposure measures to the cohort members by connecting their occupational histories with a JEM.27
The JEM was originally constructed for estimation of occupational mechanical exposures in a study of osteoarthritis leading to total hip replacement.27 The JEM cross-tabulated 121 job groups with expected homogeneous exposure patterns, and five experts’ assessments of number of hours per day spent standing/walking, total load lifted per day, and daily frequency of lifting loads weighing 20 kg or more. The JEM comprised all 2227 different occupational titles in DISCO-88 codes, except 117 titles that were considered obsolete. The five experts included four occupational health physicians and a medical graduate specialising in occupational medicine. They independently rated the exposures, and after settlement of a few major disagreements at a panel meeting, the means of the independent ratings were included in the JEM.
Occupational titles with the same DISCO-88 code could occur in different job groups of the JEM, and to be able to apply the JEM-data using the available DISCO-88 codes, we had to reconfigure the JEM. First, we allocated DISCO-88 codes to all occupational titles in each job group of the JEM. Then we split up the job groups and regrouped the occupational titles according to DISCO-88 codes. Finally, we obtained DISCO-88 code-specific estimates of the three exposure measures by averaging across the original JEM-exposures represented by the occupational titles with the same DISCO-88 code. In a few cases, we disregarded JEM-exposure estimates for rare occupational titles in order to avoid unacceptable misclassification of exposures for more frequent job titles with the same DISCO-88 code.
In the analyses, we a priori chose an exposure time window of 5 years up until 1 year before an event of surgery. For each person and each 5-year exposure time window, we calculated standardised cumulative mechanical exposures based on the exposures in each job (according to the reconfigured JEM), and the number of years with that job within the exposure time window.27
One standing-year was defined as standing/walking 6 h per day for 1 year—for example, standing/walking 3 h per day for 1 year was given a value of 0.5 standing-years. One ton-year was defined as lifting one ton per day for 1 year, and one frequent-heavy-lifting-year as lifting objects weighing 20 kg or more at least 10 times a day for 1 year—for example, lifting objects weighing 20 kg or more 80 times a day for 1 year was given a value of eight frequent-heavy-lifting-years. Standardised cumulative exposures were then obtained by summing up for each year in the exposure time window. This principle corresponds to the calculation of other standardised cumulative exposures such as ‘pack-years’ to express standardised doses of tobacco consumption.27 ,28
In the calculation, we included information from ECM on average number of hours worked per week each year, using the following factors: 1 (≥37 h/week), 0.75 (≥28–<37 h/week), 0.5 (≥18.5–<28 h/week), 0.25 (≥9–<18.5 h/week), and 0 (0–<9 h/week). For years outside the labour market, including years without employment, we used exposure estimates of zero—for example, the number of ton-years would be calculated in the following way for a person who worked full-time (37 h/week) for 1 year in a job lifting 0.5 ton per day, and full-time in another job for 2 years lifting 5 tons per day, followed by a part-time job (30 h/week) for 2 years lifting 8 tons per day: 
For participants who had years of employment with missing DISCO-88 codes, JEM estimates for these years were obtained as the mean of the JEM estimates for all the person's years with information on DISCO-88 codes from 1993 to 2008. If no DISCO-codes were registered at all, the exposure was left missing.
Covariates
In Statistics Denmark, socioeconomic status (SES) is assessed based on the main type of income. For the self-employed, the estimate is based on the number of employees, for persons with no income, the estimate is based on whether they are students or not, and for employees, the estimate is based on DISCO-88 codes from the ECM.21 We classified 22 SES groups into five categories: (1) self-employed, (2) top managers and employees at upper level, (3) employees at intermediate level, (4) employees at basic level, (5) persons outside the labour market including unemployed, and a sixth group with missing information on SES. We chose the information for 2003.
Statistical analyses
In the analyses of risk factors, we applied models with time-varying mechanical exposures using a time-lagged function to link the outcomes with exposures within a time window of 5 years up to 1 year before an event. The risk of inguinal hernia was calculated with a logistic regression technique equivalent to survival analysis.29 The number of follow up intervals (of 1 year each) before developing the outcome was analysed by logistic regression on the total number of observed follow-up intervals. Follow-up time was calculated from 1 January 1998 for persons who were at least 18 years old and whose employment periods added up to at least one full-time year of employment during the previous 5 years as on 31 December 1997. Younger persons and persons with less than 1 year of full-time employment on 1 January 1998 entered follow-up the year after their 18th birthday, or the year after obtaining at least 1 year of full-time employment during a preceding 5-year time window. Follow-up continued until the criterion for either one of the two outcomes became positive, until surgery for a non-specified or pantaloon hernia, or until the individual's 65th birthday, the date of death or emigration (including change of address to Greenland), or 31 December 2008, whichever came first. Exposures were categorised based on increments of one (ton-years and standing-years) or of 10 (frequent-heavy-lifting years). We expected that the exposure variables would be highly correlated, and a priori, we decided to analyse one exposure variable at a time, and to include the following covariates in our full models: age at 1 January each year (continuous, with one year increments), SES (1–5, categorised with group 1 as reference), region of residence (eight geographic regions based on zip codes), calendar year, and number of follow-up intervals (whole years, see below). These analyses were supplemented by tests for trend using exposure category as a continuous variable. We also performed unadjusted analyses and analyses with adjustment for age and number of follow-up intervals alone. Two-by-two correlation analysis was done between ton-years, frequent-heavy-lifting-years, and standing-years.
To evaluate whether standing-years or ton-years/frequent-heavy-lifting-years had the larger impact on risk, we performed analyses that included both standing-years and one of the two lifting variables at a time. We also undertook analyses of risk in relation to standing-years stratified by exposure to ton-years (≤2, >2), and we constructed a variable with six categories of combined exposure to standing-years and ton-years (0–≤2, >2–≤4 and >4 standing-years cross-tabulated with 0–≤2 and >2 ton-years). Furthermore, we conducted adjusted analyses for standing intensity, that is, the average number of hours spent standing/walking within each 5-year exposure time window, in order to convert our findings to measures with direct relevance for prevention.
Any potential for prevention, in case of increased risks, was evaluated by calculating the excess number of inguinal hernia repairs above the expected number, by multiplying the rate fraction (relative risk-1/relative risk) for each exposure category with the number of cases within the exposure category, and then summing up the excess number of inguinal hernia repairs across all exposure groups.
Results
Figure 1 displays the establishment of the study cohort. A total of 1 545 987 persons were eligible for the study. Total follow-up time amounted to 14 699 465 person-years, corresponding to an average of 9.5 years per person. During follow-up, 40 395 first-time events of lateral or medial inguinal hernia repair occurred: 22 926 lateral hernias (12 841 right-sided, 10 081 left-sided, four not side-specified), 15 877 medial hernias (8763 right-sided, 7109 left-sided, five not side-specified), 1493 pantaloon hernias, and 99 hernias with missing information on type. The incidence rate (IR) of lateral hernia repair was 15.6 per 10 000 person-years. The IR of medial hernia repair was 10.8 per 10 000 person-years. Figure 2 presents the age distribution. The mean age at surgery for lateral hernia was 46.6 years (SD 12.2), and for medial hernia 49.4 years (SD 9.8).
Flowchart from national birth cohort to study cohort.
Distribution of age at time of first-time surgery for lateral (n=22 926) and medial (n=15 877) inguinal hernia. The mean age at surgery for lateral hernia was 46.6 years, and for medial hernia 49.4 years.
Table 1 shows characteristics of the cohort at start of follow-up according to categories of ton-years. Across the exposed categories, the mean age increased from 31.2 to 39.4 years with increasing exposures. The unexposed category had the same age as the most highly exposed category. As would be expected, SES showed a skewed distribution with increasing prevalence of employees at intermediate and basic levels in the higher-exposure categories. Table 1 also shows a strong positive correlation between ton-years, standing-years, and frequent-heavy-lifting-years. The correlation coefficient between ton-years and standing-years was 0.7677, and between the two lifting exposures, it was 0.9588.
Characteristics of the risk cohort at start of the follow-up period for each individual according to cumulative load lifted within the preceding 5 years.
Table 2 shows results of age-adjusted and fully adjusted analyses of risk of first-time lateral inguinal hernia repair, as well as results of analyses including mutual adjustment for standing-years, and either of the two lifting exposures. All three exposures showed exposure–response patterns when analysed one at a time in both age-adjusted and fully adjusted models with ORs of up to around 1.4. When exposure to standing-years was included together with either ton-years or frequent-heavy-lifting-years, only standing-years remained as a risk factor (table 2). In stratified analyses, the adjusted ORs for increasing numbers of standing-years were similar for participants with ≤2 ton-years and >2 ton-years (results not shown). Analyses using the variable that combined standing-years with ton-years did not indicate interaction, since the ORs for low, medium and high numbers of standing-years were similar, irrespective of whether these categories were combined with ≤2 ton-years or >2 ton-years (results not shown). When the standing intensity increased from 0 h/day to a maximum of 7.3 h/day with increments of 1 h, the adjusted ORs gradually increased to around 1.3 for >3–≤7 h/day, and then increased to 1.77 (95% CI 1.54 to 2.03); the OR already reached significance for >1–≥2 h/day ((OR 1.06 (95% CI 1.01 to 1.12)). According to aetiologic fractions estimated based on standing-years (table 2, adjusted model I), 3449 cases (15%) out of at total of 22 814 first-time lateral hernia repairs would—in theory—be preventable.
Risk of lateral inguinal hernia repair in relation to specific occupational mechanical exposures.
Table 3 shows results of analyses of risk of first-time surgery for medial inguinal hernia. A small increased risk was seen for those with exposures above six standing-years (OR=1.14 (95% CI 1.06 to 1.23)), but in general, no increased risk was seen in relation to occupational mechanical exposures.
Risk of medial inguinal hernia repair in relation to specific occupational mechanical exposures.
Discussion
In this population-based male cohort study, we found that the risk of lateral inguinal hernia repair increased with increasing cumulative exposures to daily lifting activities (total loads, frequent heavy lifting), and prolonged standing/walking at work. The risk of medial inguinal hernia repair was unrelated to these exposures, except for a relation with cumulative standing/walking at work above a threshold of six standing-years, corresponding to 6 years of work that entailed standing/walking for 6 h per day.
The study was based on nationwide high-quality longitudinal registers. Particular strengths were the distinction between medial and lateral repairs, and the independent quantitative assessment of specific mechanical occupational exposures. It may be questioned if hernia repair can be considered a measure of hernia formation. Since watchful waiting is an option in asymptomatic or minimally symptomatic cases, symptom aggravation due to occupational mechanical exposures could explain exposure–response relations. However, it seems improbable that symptom aggravation would occur in case of a lateral hernia, but not in case of a medial hernia. Therefore, we think that the exposures are most likely related to hernia formation rather than aggravation of symptoms. Socioeconomic status could affect the propensity to seek medical attention and the probability of referral for evaluation at a hospital, but again, this could hardly explain the difference in work-relatedness between the two types of hernias
We arbitrarily decided on a 5-year time window for accumulation of exposure effects; this may not be the ideal choice. However, time-window analyses require exposures to vary over time at the individual level, which tends not to be the case—people with heavy physical work are likely to have this type of work for years, just as employees at upper and intermediate levels are likely to stay at these levels over time. Therefore, we have not attempted time-window analyses so far. Inflation of the effect measures due to recall bias could not explain the observed associations because our exposure estimates did not rely on self-report. We may have overlooked high risks for small but highly exposed occupational groups (or subsets of occupational groups) that could be hidden in less exposed job groups of the JEM. In general, exposure misclassification with erroneous inclusion of highly exposed occupations in job groups with low exposures and/or occupations with low exposures in job groups with high exposures would lead to underestimation of effects. Denmark's public hospital system means that it should not be necessary for people to change jobs or leave the labour market because of a hernia that demands surgery. Thus, underestimation of effects due to healthy worker selection seems highly unlikely.
Two external experts originally evaluated the ranking of the exposures in the JEM, and in general, they agreed with the ranking.27 However, the exact exposure estimates may be inaccurate. Information on DISCO-88 codes was much more likely to be missing for employees at basic level than for self-employed persons, top managers and employees at upper and intermediate levels, table 1. This could lead to an underestimation of the associations, but the overall percentage of men with missing exposure information was only 0.8% and, therefore, negligible.
The main weakness of our register-based study was that we had no information on body mass index, smoking and physical activity outside of work.9 ,12 ,15 ,30 We did control for age and SES. As our next step, we plan to take the just-mentioned potential confounders into account using data from the Musculoskeletal Research Database at the Danish Ramazzini Centre. The database contains around 40 000 persons (50% men) who have provided questionnaire data in nine previous studies of musculoskeletal symptoms in the general Danish population or selected occupational groups.
For lateral hernias, we found that only standing-years retained significance in analyses that also included either ton-years or frequent-heavy-lifting-years (due to correlation, ton-years and frequent-heavy-lifting years could not be included in the same model). The effect of standing-years was also found in analyses stratified by lifting activities in terms of ton-years. Thus, standing-years remained as the most robust risk factor. A possible mechanism for work-related formation of inguinal hernias is increased intra-abdominal pressure during work activities.7 Significant increases in intra-abdominal pressure have been observed during a variety of occupational tasks. In 20 healthy young people, measured mean pressures while sitting, standing and walking up a flight of stairs were 17 mm Hg, 20 mm Hg, and 69 mm Hg, respectively.17 Intra-abdominal pressures above 100 mm Hg have been measured during heavy lifting in a stooping position.16 Our results agree with the hypothesis that formation of lateral hernias can be related to increased abdominal pressure that leads to protrusion of abdominal contents through an opening of the inguinal canal (maybe through a patent processus vaginalis which is an embryological evagination of the peritoneum). While standing, pressure due to gravity will pass down to the inguinal region, and standing may be a prerequisite for the canal to open, whereas, sitting may preclude the protrusion of a hernia. The last-mentioned theory would comply with our results that indicated an effect of standing for only short periods a day with no obvious threshold. To the extent that our lifting variables did not accurately reflect intra-abdominal pressures, we could underestimate effects of specific lifting activities. Hence, our finding that standing-years remained as the most robust risk factor must be interpreted with caution. Our results do not support the hypothesis that physical exposures lead to formation of medial hernias through a mechanism involving connective tissue degradation and weakening of the transversalis fascia. In accordance with this, a recent study found no differences in key mechanical properties of the transversalis fascia in patients with or without inguinal hernias (unfortunately, the study did not distinguish between medial and lateral hernias).31 Interestingly (and as noted in the aforementioned recent systematic review,7), two previous studies have observed that lateral hernias were more likely to occur among patients with heavy work.32 ,33
The incidence of inguinal hernia repair in Denmark is comparable with reports from Sweden.30 ,34 It seems likely that our results can be generalised to other countries with working conditions comparable with the Danish. Our study was restricted to men, and the results may not be generalisable to women who are much less susceptible to inguinal hernia formation. The ORs were significantly elevated, but did not exceed 1.5. Nevertheless, we estimated that—in theory—around 15% of all lateral hernia repairs among men would be preventable by reducing the daily time spent standing/walking. If this is assumed to be correct, the question arises whether efforts should be made to reduce the number of hours per day spent standing/walking across all exposure categories above zero, or if the maximum accepted number of hours per day should be set at—for example, 5 or 6 h of standing/walking. In view of the negative effects of a sedentary life style, the last-mentioned strategy seems more attractive, but then of course, the difference that we observed regarding the influence of standing/walking on risk needs to be corroborated in further studies. The same is true of the difference in work-relatedness of medial and lateral inguinal hernias.
In conclusion, we found that the risk of lateral inguinal hernia repair increased with increasing cumulative exposures to daily lifting activities (total loads, frequent heavy lifting), and prolonged standing/walking at work. In general, the risk of medial hernia repair was unrelated to the exposures. Combination of medial and lateral inguinal hernias in aetiologic research may mask specific relations with risk factors. For lateral hernias, the results suggested that preventive effects might be achieved by reducing mechanical exposures at work.
Acknowledgments
We thank Tine Steen Rubak Erichsen who made her Job Exposure Matrix available for the study.
References
Footnotes
-
Contributors All authors contributed to the conception and design of the study and to the interpretation of the results. MVV and PF conducted the analyses. MVV drafted the manuscript and SWS prepared the final version that has been approved by all authors.
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.