The role of serum uric acid (UA) as an independent risk factor for several health outcomes remains controversial. However, given the accumulating evidence that UA is related to predictors of occupational disability such as obesity, excessive alcohol consumption, hypertension, and diabetes mellitus, UA may represent a relevant risk indicator for occupational disability, which has emerged as an important public health problem.
The association between UA and occupational disability was examined in a cohort of 16,532 male construction workers in Germany who underwent occupational health examinations from 1986 to 1992 and were followed until 2005. Cox regression analysis was employed with adjustment for established risk factors.
A total of 3,002 cases of disability pension occurred during the followup. Risk of all-cause occupational disability was significantly increased for UA concentration in the top quartile (hazard ratio [HR] 1.25, 95% confidence interval [95% CI] 1.12–1.40) after adjustment for potential confounders when compared with UA levels in the lowest quartile (≤5.23 mg/dl). Cause-specific analysis revealed a significant association between increased UA and occupational disability due to cardiovascular diseases (HR 1.62, 95% CI 1.20–2.17). In addition, positive but statistically nonsignificant associations were observed for occupational disabilities due to respiratory diseases (HR 1.78, 95% CI 0.99–3.23), musculoskeletal disorders (HR 1.16, 95% CI 0.98–1.37), diseases of the digestive system (HR 1.59, 95% CI 0.69–3.67), and mental disorders (HR 1.40, 95% CI 0.95–2.06).
Our study is the first to our knowledge to indicate that increased UA might also serve as a potential independent risk indicator for occupational disability, in particular due to cardiovascular diseases.
Serum uric acid (UA) is known as a predictor of gout (1) and is associated with the metabolic syndrome, which covers the development of obesity, hypertension, hyperlipidemia, and diabetes mellitus (DM) (2, 3). Although an association between hyperuricemia and cardiovascular risk has been described for over 50 years (4), the role of UA as a mediator of vascular damage has gained widespread interest only very recently (5). Furthermore, the relationship between UA and cardiovascular disease was seen even at UA levels considered to be in the normal-to-high range (>5.5 mg/dl) (6). However, there is an ongoing debate about whether hyperuricemia represents an independent risk factor for cardiovascular diseases or whether increased UA is merely associated with other risk factors such as hypertension and renal disease (7). The biologic action of UA in humans is not fully understood. UA is considered an antioxidant compound, but preclinical evidence suggests a proinflammatory action (8).
Studies assessing the prognostic value of UA levels with respect to all-cause and cardiovascular mortality have revealed conflicting results. In most of them, but not in all, a positive association was observed (9–11). Two large population-based cohort studies of middle-aged men in Austria and China indicated strong dose-response relationships for these outcomes, which remained substantial after stratification for hypertension and DM (12, 13). In addition, hyperuricemia has been associated with an increased cancer risk (14, 15), and it has been discussed whether elevated levels of UA increase the risks of dementia and lowered cognitive function (8, 16, 17).
Given this background, serum UA may represent a promising marker for a number of chronic conditions that lead to disability and premature mortality. Risk of occupational disability, which is an important end point that causes major social and economic problems (18), was found to be related to factors such as obesity (19, 20), excessive alcohol consumption (21), hypertension (22), and DM (23), all of which are related to hyperuricemia.
To our knowledge, no studies have evaluated the role of UA as a predictor of occupational disability so far, although disability in working-age people is a major public health issue globally. In Germany, there are currently nearly 1.6 million people receiving disability pensions from the German pension fund (6.3% of all pensions), of whom >160,000 were granted pensions in 2008 (13.1% of all incident pensions) (24).
In the present study, we assessed the association of detailed dose-response patterns of UA with all-cause occupational disability and the association of UA with occupational disability due to different causes in a large scale cohort of manual workers over a 10-year followup period.
MATERIALS AND METHODS
The baseline study population comprised 19,421 male construction workers, ages 25 to 59 years, who were working as bricklayers (n = 6,204), painters (n = 2,947), laborers (n = 2,874), plumbers (n = 2,804), carpenters (n = 2,594), or plasterers (n = 1,998). They underwent routine occupational health surveillance by the Workmen's Compensation Board for Construction Workers in Württemberg, Southern Germany, between August 1986 and December 1992. This occupational health surveillance is based on legislation on health and safety at work, and regular examinations are offered to all construction workers. Participation is voluntary, but ∼75% of all invited employees attended the medical examination during the period of recruitment and were eligible for followup. All participants were members of the statutory pension fund and had not received a disability pension as of the baseline examination. They were representative of the underlying source population of all construction workers with respect to age, nationality, and type of occupation. The study was approved by the local and regional ethics committees and by the Ministry of Social Affairs Baden-Württemberg.
Health examinations at baseline included a physician-explored working and lifestyle factors history, a physical examination (including measurements of height, weight, and blood pressure), functional measurements, and blood and serum analysis. The examinations were conducted by experienced occupational health physicians and documented according to a standardized protocol. UA levels in this study were measured photometrically with a Hitachi 705/717 system (Hitachi). Information regarding UA levels was missing in 702 cases because some workers either rejected having a blood sample taken or provided external laboratory-analyzed findings from a recent blood analysis, which were not included in the medical records used for our study.
The cohort was linked to the pension register of the German Pension Fund Baden-Württemberg in March 2006 to identify construction workers who had been granted a disability pension during followup. The pension register provided information regarding vital status and whether the individual was still working, had retired due to age, was unemployed or under vocational retraining, or whether a disability pension (permanent or temporary) was granted. In the case of multiple temporary or transient periods of occupational disabilities, only the first-incident disability was considered.
The criteria for being considered work disabled and receiving a disability pension are under continuous change. Up through the year 2000, a disability pension was granted in Germany when one's ability to earn a living had been permanently reduced by ≥50% due to illness, injury, or impairment, irrespective of whether the injury was work related or not, and whether the worker was able to engage in any substantial gainful employment. In the year 2001, the threshold was set to 3 hours of work ability per day for complete work disability and to 6 hours per day for partial work disability. Disability pensions are granted contingent on thorough medical examination by the medical service of the pension fund and are based on a medical diagnosis.
Because some cases of occupational disability were ascertained in retrospect, the cutoff date for followup was set in June 2005 to capture all cases that had occurred during the specified time period. Cause of occupational disability was coded according to the International Classification of Diseases, Ninth Revision (ICD-9), by trained medical officers from the pension fund. We had to exclude 2,187 men (11.3%) who had either moved to a different region or had changed employment. The very strict confidentiality rules in Germany did not allow us to follow these people further. Also, the 702 men (3.6%) with missing measures of UA at baseline had to be excluded. Therefore, the final study population for this analysis comprised 16,532 construction workers who had been successfully linked to the pension register.
In this analysis, UA concentrations were classified into quartiles with cut points at 5.23, 6.05, and 6.97 mg/dl, which correspond to the 25th, 50th, and 75th percentiles computed over the entire sample, respectively. Additionally, the highest group was subdivided at a cut point of 7.91 mg/dl, which corresponds to the 90th percentile. The normal range of UA in the common clinical setting is defined as all values <7.0 mg/dl (25).
The risk of occupational disability according to levels of UA was assessed using the Cox proportional hazards model with the date of the baseline examination as the start of followup time, whereas the occurrence of occupational disability was defined as an event. A person was classified as censored if they were known not to have been granted a disability pension according to pension fund records, or if their pension fund insurance was terminated due to another reason (such as retirement pension, death, 65th birthday [the legal retirement age], or change to another insurer), or if (in analyses of cause-specific disabilities only) they had occupational disability due to another cause.
Potential variation of the association of UA with occupational disability after control for auxiliary factors known to be related to UA, such as age, alcohol consumption, body mass index (BMI), cholesterol, and smoking, was assessed by multivariate analysis. After crude analysis, we first included age as a covariate into the model. Further adjustment for nationality, profession, BMI, alcohol consumption, smoking, and cholesterol at baseline as covariates was done in multivariable analysis. For cholesterol and age a linear term, and for the latter also a quadratic term, were simultaneously entered into the model while index variables were created for the other, categorical variables. In light of the mounting evidence that increased UA may induce hypertension, hypertension was not considered as a potential confounder in the multivariate analysis.
To explore potential differences in the predictive value of UA in the short and long run, supplemental specific analyses were performed for the initial 3 years and for the subsequent years of followup. Additional analyses were conducted in subgroups according to the presence or absence of defined types of comorbidity at baseline (prevalence of gout, cardiovascular diseases, and DM as well as diseases of the liver, bile, and pancreas) and with respect to cause-specific occupational disability.
To overcome the statistical drawbacks of treating UA as a categorical variable, UA was also entered as a continuous variable with 4.5 mg/dl as the reference value in supplementary regression models using fractional polynomials as described by Royston et al (26). The proportional hazards assumption was checked by log–log survival plots. All statistical analyses were carried out with the SAS statistical software package, version 9.1 (SAS Institute).
Characteristics of the study population (overall and grouped by UA) at baseline are shown in Table 1. The mean age of the 16,532 workers at baseline was 42.0 years. More than three-fourths of the cohort members were of German nationality. Bricklayers represented the largest professional group, composing nearly one-third of the study population. A considerable share of the study participants (∼63%) were overweight or obese (BMI ≥25 kg/m2). The majority (52%) reported daily alcohol consumption, and 58% of the study population were current smokers at the time of the baseline examination.
Table 1. Characteristics of the study population at the baseline examination by category of serum uric acid (UA)*
Serum UA, mg/dl
Values are the percentage unless otherwise indicated.
Information was missing regarding body mass index in 322 cases (2%), alcohol consumption in 2,198 cases (13%), cholesterol in 3 cases, and smoking in 2,658 cases (16%).
Prevalence adjusted to the age distribution in the entire cohort by the direct method. Diagnoses have been classified according to International Classification of Diseases, Ninth Revision, code: gout, 274; cardiovascular disease, 390–459; diabetes mellitus, 250; musculoskeletal disorders, 710–739; respiratory diseases, 460–519; diseases of the digestive system, 520–579; and mental disorders, 290–319.
Elevations of UA levels were strongly associated with German nationality. The factors most strongly related to increased UA concentration were obesity (BMI ≥30 kg/m2), self-reported excessive alcohol consumption, and high cholesterol levels (>254 mg/dl; for trend tests, P < 0.001 for each). No major differences in UA concentrations could be observed with respect to age or professional groups. Baseline prevalence of gout, cardiovascular diseases, and diseases of the digestive system (predominantly diseases of the liver, bile, and pancreas) was strongly associated with hyperuricemia (for trend tests, P < 0.001 for each factor).
All-cause occupational disability according to UA.
Overall, there were 3,002 incident cases of disability pension during the followup period (mean duration 10.8 years). The association between UA levels and all-cause occupational disability with the lowest quartile of UA concentration as the reference category is shown in Table 2.
Table 2. All-cause occupational disability by category of serum uric acid in mg/dl*
HR = hazard ratio; 95% CI = 95% confidence interval; ref = reference.
Per 100,000 person-years.
Adjusted for age, body mass index, nationality, smoking status, cholesterol, profession, and alcohol consumption.
In crude analysis, a modest positive relationship between UA concentration and all-cause occupational disability (for trend test, P < 0.001) was observed, with a significantly increased risk in the 2 highest quartiles. Risk was elevated by nearly 50% in the top quartile compared with the reference group. After adjustment for age, the association of UA concentration with occupational disability was even stronger, resulting in significantly elevated relative risks (RRs) in all 3 upper quartiles of UA. Risk of occupational disability was elevated nearly 2-fold in the category of UA values >7.91 mg/dl (upper 10%) compared with the lowest quartile. Further adjusting for BMI, nationality, smoking status, cholesterol, profession, and alcohol consumption weakened this association to some extent, but the RR of occupational disability remained significantly elevated for the highest quartile compared with the lowest quartile and was still elevated by nearly 50% for subjects with UA values >7.91 mg/dl. Likewise, a significant association between increased UA and the risk of occupational disability was observed using 7.0 mg/dl as the threshold for the normal range (RR 1.24, 95% confidence interval [95% CI] 1.14–1.35).
Disability pension was granted during the first 3 years of the followup period in 502 (17%) of the 3,002 cases. Associations of baseline UA levels with occupational disability were very similar during the first 3 years of followup and during later years, even though the adjusted hazard ratio (HR) for the top quartile was statistically significant for the later years only.
A more comprehensive picture of the association of UA with occupational disability is provided by the dose-response analysis, treating UA concentration as a continuous variable using a first-degree fractional polynomial with a quadratic term (β = 0.00559, SE = 0.00102) (Figure 1). The increase in HRs was negligible at UA concentrations in the normal range (<7.0 mg/dl), but a more substantial increase was seen in the higher regions.
Additional analyses with stratification of the study population by presence or absence of gout, cardiovascular diseases (including hypertension), and DM as well as diseases of the liver, bile, and pancreas at the baseline examination, all of which might be responsible for hyperuricemia (>7.0 mg/dl), did not reveal any relevant confounding or interaction by these conditions (data not shown).
Occupational disability according to UA and cause of disability.
Information on the cause of occupational disability could be obtained for 2,715 (90.4%) of 3,002 cases of disability pension. The most common main cause of occupational disability was musculoskeletal disorders, occurring in 1,243 (45.8%) of the cases, half of which were due to dorsopathies and more than one-fourth of which were due to arthrosis. Cardiovascular diseases represented the second most frequent cause of occupational disability (occurring in 17.3% of the cases), followed by mental disorders (9.0%) and cancer (8.0%). Frequencies and HRs (age- and multiple-adjusted) of cause-specific occupational disability are shown in Table 3.
Table 3. Cause-specific occupational disability by category of serum uric acid (UA)*
Serum UA, mg/dl
HR = hazard ratio; 95% CI = 95% confidence interval; ref = reference.
Per 100,000 person-years.
Adjusted for age, body mass index, nationality, profession, smoking status, cholesterol, and alcohol consumption.
Serum UA concentrations were positively associated with the risk of occupational disability due to cardiovascular diseases, musculoskeletal disorders, respiratory diseases, diseases of the digestive system, mental disorders, and other diseases in age-adjusted analysis. The RR of occupational disability due to cardiovascular diseases was significantly elevated for all 3 upper quartiles, with a >2-fold risk in the highest quartile, whereas risk of occupational disability due to musculoskeletal disorders was significantly increased for the 2 highest UA quartiles compared with the reference group. Furthermore, a significantly increased risk of occupational disability due to mental disorders (HR 1.65, 95% CI 1.17–2.35) was found for workers with UA concentrations in the highest quartile. Despite a clear positive association of UA with risk of occupational disability due to respiratory diseases as well as diseases of the digestive system, the differences in risk across all levels of UA were not statistically significant given the low frequencies of disability pensions granted for these reasons. No significant association was observed for occupational disability due to cancer. Splitting the top quartile into 2 subgroups strengthened the association and revealed a significantly increased risk of occupational disability due to all of the assessed diagnoses except for cancer and respiratory diseases for workers with UA values >7.91 mg/dl.
The increase in risk of occupational disability associated with higher levels of UA substantially declined for all of the assessed diagnoses except respiratory diseases after further adjustment for BMI, nationality, smoking status, cholesterol, profession, and alcohol consumption. While the strongest, albeit statistically nonsignificant, risk increase was seen for occupational disability due to respiratory diseases, risk of occupational disability due to cardiovascular diseases persisted with an HR of 1.62 (95% CI 1.20–2.17) for the highest versus the lowest quartile of UA. Considering solely the top category above the 90th percentile, the strongest albeit again statistically nonsignificant risk increase was observed for occupational disability due to diseases of the digestive system. Risk of occupational disability due to cardiovascular diseases as well as musculoskeletal disorders persisted in being significantly elevated for this most exposed group, with a more than 2-fold elevated risk for occupational disability due to cardiovascular diseases and an RR of 1.36 (95% CI 1.11–1.67) for occupational disabilities due to musculoskeletal disorders. The latter association was stronger for osteopathies, chondropathies, and acquired musculoskeletal deformities (ICD-9 codes 730–739) as causes of disability than it was for arthrosis, dorsopathies, and rheumatic disorders. A modest significant association between increased UA and occupational disability due to injuries (130 of the 382 cases with other causes of occupational disability) could also be observed in the top quartile, with a RR of 1.87 (95% CI 1.06–3.30) (data not shown).
To our knowledge, this is the first study that has assessed the role of UA as a predictor of occupational disability. Whereas the association of UA with all-cause mortality among middle-aged men appears to be weak or even nonsignificant (9, 10, 12, 14), we found a consistent dose- response relationship of UA with occupational disability, which suggests that UA may represent an important predictor of occupational disability. In addition, cause- specific analysis revealed positive dose-response relationships between UA and occupational disability due to musculoskeletal and mental disorders as well as cardiovascular and respiratory diseases, which represent the major causes of disability in this occupational group.
Elevated UA concentration was clearly associated with the risk of all-cause occupational disability in age-adjusted analysis. This association was reduced by some extent but remained significant after further adjustment for potential confounding factors, suggesting that hyperuricemia may act as an independent risk factor for disability. Likewise, the association between increased UA and occupational disability due to cardiovascular diseases was especially strong and remained significant in the top quartile of UA concentration after adjustment for multiple covariates. This is in line with the established association between UA and cardiovascular risk (7), although the association among middle-aged men appears to be weaker than among the elderly or among women (9–11). This discrepancy might be explained by the fact that hypertension is a major cause of occupational disability and that several studies have reported that hyperuricemia may increase the risk for hypertension independent of other risk factors (2, 3, 27). Recent experimental investigations in animals indicate that hypertension may be caused by UA-mediated renal vasoconstriction resulting from a reduction in endothelial levels of nitric oxide, with activation of the renin–angiotensin system (28). Likewise, clinical research revealed that increased UA in humans also correlates with endothelial dysfunction and increases renin activity (29).
Musculoskeletal disorders cause the largest proportion of all cases of disability pension among manual workers. Age-adjusted analysis revealed a strong positive association between elevated UA concentration and occupational disability due to musculoskeletal disorders. Adjustment for covariates such as BMI, which is an important risk factor for osteoarthritis (19), substantially attenuated this association, but an increased risk persisted for the top category above the 90th percentile. It cannot be ruled out that workers with lower UA levels take on more strenuous job assignments and thus are more prone to non–gout-related musculoskeletal disability. However, the association remained significant when the only diagnoses considered were osteopathies, chondropathies, and acquired musculoskeletal deformities. This pattern may be partly explained by the well-established manifestation of intraarticular crystal deposition due to hyperuricemia, which eventually leads to osteopathy and joint damage. In addition, UA has been associated with unspecific chronic musculoskeletal symptoms, which were mainly ascribed to chronic widespread weather-dependent symptoms (30).
Our study also revealed that hyperuricemia is associated with mental disorders. This is in line with results from other epidemiologic studies that have reported a relationship between serum UA and vascular dementia due to the proinflammatory potency of UA, which may lead to cognitive impairment (8). Although there are conflicting results about whether hyperuricemia increases or decreases the risk of dementia and impaired cognitive function (8, 16, 17), our results, which in contrast to the previous studies were not confined to elderly subjects, suggest that increased UA may be positively associated with the development of mental diseases.
A modest significant association between increased UA and occupational disability due to injuries was found in the top quartile. However, the proportion of occupational disability due to injuries (ICD-9 codes 800–999) was negligible, because occupational injuries are usually covered by the workers' compensation board rather than by the pension fund.
Previous clinical studies proposed UA as a clinical marker for hypoxia, chronic respiratory disease, and chronic respiratory failure (31–33). These findings might explain the observed modest significant association between increased UA and occupational disability due to respiratory diseases for UA values in the top category above the 90th percentile.
In contrast to recent findings indicating a clear positive relationship between UA concentration and cancer mortality (14) as well as cancer incidence (15), we could not find any association of UA with cancer as the main diagnosis for occupational disability. It has been suggested that the antioxidant properties of UA may play an important role in cancer etiology by preventing formation of oxygen radicals, thereby protecting against carcinogenesis (34). Further work needs to be done to investigate whether hyperuricemia might be a relevant marker for the incidence of cancer or, rather, represent an early manifestation of the carcinogenic process (15).
Several potential limitations of the present study should be considered. Our cohort was restricted to middle-aged men performing physically heavy work with mostly low socioeconomic status, job autonomy, and control. Several studies revealed stronger associations between UA and health outcomes such as cardiovascular diseases and all-cause death among women, suggesting that the relationship between UA and occupational disability could be stronger in a general population (9, 11). The generalizability of the study results might be limited by a particular unhealthy lifestyle with respect to smoking status, alcohol use, and obesity. However, adjustment for these factors did not affect the observed patterns. A further potential limitation is that the cause-specific analysis was restricted to the main diagnoses of occupational disability. No information regarding auxiliary causes of occupational disability was available. Therefore, the possibility has to be considered that unmeasured comorbidity at the time of ascertainment of occupational disability may mask the true relationship of UA with cause-specific occupational disability. However, this possible limitation does not invalidate the prognostic significance of UA as a risk indicator with respect to multiple specific end points. In addition, all diagnoses responsible for work disability were defined by trained medical officers from the pension fund. Therefore, the validity of the cause of disability can be assumed to be high.
Major strengths of our study were the length and completeness of followup and the size and relative homogeneity of the study population. The large case number of disability pensions enabled us to assess the association of UA with occupational disability in great detail. Although the sample sizes of some of the cause-specific main outcomes (such as respiratory diseases, diseases of the digestive system, and cancer) were too low to detect significant associations, our analysis gave a clear insight in the pattern of major cause-specific dose-response relationships. In contrast to alcohol consumption and smoking, UA concentration can be routinely measured in clinical laboratories, with reliable results and only small, short-term individual fluctuations that do not follow age-specific or diurnal patterns. Given these characteristics, UA represents an easily and routinely measured marker that can feasibly be used as a helpful prognostic indicator in both clinical and occupational health practice.
In conclusion, our study is the first to indicate that increased UA might also serve as a potential independent risk indicator for occupational disability, in particular due to cardiovascular diseases.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Arndt had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Brenner, Arndt.
Acquisition of data. Brenner, Drath, Arndt.
Analysis and interpretation of data. Claessen, Brenner, Arndt.
We thank the German Pension Fund Baden-Württemberg for providing the followup data, as well as Claudia El-Idrissi Lamghari (German Cancer Research Center, Division of Clinical Epidemiology and Aging Research, Heidelberg, Germany) and Jürgen Banzhaf (Workmen's Compensation Board for Construction Workers, Böblingen, Germany) for technical assistance over the course of this study.