Results

Age and sex as well as urinary excretion of cadmium and protein HC in the study population are shown in table 1.

A positive, highly significant, linear relation was found between dose (cadmium in urine) and effect (urinary protein HC) after adjustment for age for both sexes (table 2). The table shows the regression coefficients for the independent variables age and urinary cadmium, with protein HC as the dependent variable.

A total of 171 people had increased protein HC concentrations in urine with a clear dose-response relation between urinary cadmium and the prevalence of increased protein HC in urine as shown in table 3. During the course of investigation it was discovered thatseveral (n=105) environmentally exposed people also had worked in the battery plant. These people were therefore transferred to the occupationally exposed group, which thus increased to 222 people.

The dose-response relation remained even when the occupationally exposed people were excluded. The figure shows the ORs for increased urinary protein HC relative to increasing urinary cadmium after adjustment for age and sex.

• Download figure
• Open in new tab
• Download powerpoint

Odds ratios (95% CIs) for prevalence of tubular proteinuria related to urinary cadmium in Swedish people exposed to low concentrations of cadmium.

Multiple logistic regression analysis, including age and urinary cadmium as independent variables and normal or increased urinary protein HC as the dependent variable was used to establish dose response relations at different ages. Given the mean age of the study population (53 years) an increased prevalence of tubular proteinuria of 10% was calculated to occur at a urinary cadmium concentration of 1.0 nmol/mmol creatinine. Here we have accounted for a normal background prevalence of increased urinary HC of around 5%, and thus the total measured prevalence of tubular proteinuria at a urinary cadmium concentration of 1.0 nmol/mmol creatinine would be 15%.

Discussion

The present study shows that tubular proteinuria occurs in environmentally exposed people at lower concentrations of cumulative cadmium doses than was previously realised. People with cadmium in urine of around 1 nmol/mmol creatinine (which is in the upper part of the normal range) had a threefold increase in risk of having increased urinary protein HC (figure). The OR for increased urine HC is increased more than fivefold at urinary cadmium concentrations above 5 nmol Cd/mmol creatinine, which has been recommended as a health based limit by an expert group within the World Health Organisation (WHO).

Both age and urinary cadmium affect urinary protein HC excretion (table2) and therefore age should be considered when dose-effect and dose-response relations are being analysed. Based on our data, a 10% increase in urinary protein HC would be expected in a population, with a mean age of 53, at a cadmium concentration in urine of 1.6 nmol/mmol creatinine.

One strength of our study is that we have been able to collect and analyse urine samples, with sensitive methods, from a large population with exposures to cadmium close to the normal range. In most countries the normal urinary excretion of cadmium is in the range 0.1–0.6 nmol/mmol creatinine.1 7 When more sensitive methods for detecting tubular damage are used, early and subtle effects from the accumulation of cadmium are shown.14 When a very sensitive biomarker of effect (such as protein HC) is used, the clinical implication of increased urinary excretion is limited. However, as will be discussed later, subtle tubular damage is the beginning of a pathological process, which may end up in renal failure and secondary manifestations on the bone and mineral metabolism.7

Most previous studies have used β₂-microglobulin as the marker of tubular dysfunction. A major problem with β₂-microglobulin as a biomarker in large epidemiological studies is the instability of the protein in acidic urine. More sensitive markers have, however, been found. In particular, protein HC has been shown to be a sensitive and reliable marker of early tubular damage.15-17 In a recent European collaborative study aimed at identifying early indicators of cadmium toxicity, the most sensitive indicators of nephrotoxicity were found to be protein HC, NAG, thromboxane B2, and RBP.18 The cut off points for tubular dysfunction were taken from a rather limited reference material, based on a population living in a region not far from the study area.13 It should be noted that the cut off points correspond to the 95th percentile in the reference population, and thus 5% of the study population should be expected to have values exceeding the cut off point. This has been taken into consideration in the interpretation of the results.

To obtain a wide range of cadmium exposures (facilitating the dose-response analyses) we included environmentally as well as occupationally exposed people. However, as shown in table 3, exclusion of all people with previous occupational exposure did neither affect the overall results nor their interpretation. The lower response rate among the battery plant workers might have introduced a bias, if the non-participants had different characteristics than the participants. However, as noted in the methods section, a survey of a random sample of the non-participants gave no indication that they differed from the participants in a systematic way for age, sex, or morbidity. Furthermore, the ORs for the environmentally exposed group were similar to those obtained when the occupationally exposed workers were included in the analysis (table 3).

In other parts of the world, where cadmium has contaminated the environment, increased concentrations of cadmium in urine and renal tubular effects have been recorded among residents.1-3 7During the past decade, several studies of both occupationally and environmentally exposed populations have shown that tubular proteinuria occurs at doses of urinary cadmium of 2–4 nmol/mmol creatinine.9 11 19 20 In a large study on environmentally exposed people in Belgium a 10% prevalence of tubular proteinuria was found at 2 nmol Cd/mmol creatinine, not adjusting for age.9 The present study is in agreement with these findings with a 10% prevalence of tubular proteinuria at 1.6 nmol Cd/mmol creatinine adjusted to the mean age of the study population.

Kidney function, at least glomerular filtration, normally deteriorates with age21 and pathological changes of the renal vessels, nephron, and the interstitium become more prevalent with increasing age.22 Cadmium accumulates in the kidney with increasing age, and age may thus constitute an important effect modifier when examining long term health effects of cadmium exposure. The effect of aging itself on normal urinary excretion of β₂-microglobulin, however, seems to be limited.23 The covariance between age and urinary cadmium in protein HC excretion (table 2) nevertheless suggests an increased vulnerability to cadmium in elderly people. Probably, normal degenerative changes of the renal tubuli in elderly people cause an increased sensitivity to cadmium.

Most studies of exposure to cadmium and renal tubular function have been performed on male workers. The present study showed only minor differences in tubular dysfunction between men and women, in accordance with a previous Japanese study of environmentally exposed people.24

All studies of exposure to cadmium and tubular dysfunction have been cross sectional. The tubular damage may have occurred a long time before the study was performed. If the concentration of urinary cadmium at the time of onset of tubular damage was higher, the dose-response curve would shift to the left in subsequent studies.17This problem, with a variable shape of the dose-response curve, is probably less relevant for environmentally exposed people, who have a continuous relatively low level of exposure compared with workers who have experienced high exposure to airborne cadmium in the past.17

It is, therefore, particularly noteworthy that tubular proteinuria occurs at relatively low concentrations of cadmium in urine and in the environmentally exposed group, in which the tubular damage appears at even lower concentrations than in the occupationally exposed group. It may well be that the occupationally exposed group comprises a selection of relatively healthy people, whereas the general population also includes more susceptible people. Furthermore, the group of environmentally exposed people also includes some who were very young when the exposure started and therefore are perhaps more susceptible to cadmium. In view of the relatively stable and low exposure to cadmium in the population examined, we think that the dose-response data presented here are valid and may be used to develop risk estimates in other environmentally exposed populations.

Research on health effects from cadmium has focused on the tubular damage, which from a health point of view may have limited importance, as the tubular proteinuria in itself does not give rise to any subjective symptoms or manifestations of disease.1 6 14It has recently been argued that the subclinical renal effects found in people with an increased cadmium body burden are not associated with progressive renal dysfunction, and that these early renal effects may even be reversible.25 It has been shown, however, that the cadmium induced tubular damage is irreversible in most cases even if exposure ends.7 Instead the tubular damage may become worse and glomerular damage with a drop in the GFR occurs. This has been seen among occupationally exposed workers as well as in people environmentally exposed to cadmium.7

Uraemia was common among Japanese farmers with itai-itai disease and end stage renal disease has been observed in workers heavily exposed to cadmium.3 An overall increased mortality among residents with cadmium induced tubular damage in areas of Japan polluted with cadmium has been reported, indicating that serious health consequences may eventually result from the tubular damage.3 7

Cadmium accumulation in the kidney is also related to negative health effects on the bone and mineral metabolism.1 3 7 An increased prevalence of kidney stones in workers exposed to cadmium has been noted by several investigators.7 Possibly this is related to the increased urinary excretion of calcium, secondary to tubular damage. A recent follow up of a large population of environmentally exposed people in Belgium suggests that cadmium accumulation increases the risk of development of osteoporosis.26 This is in agreement with the findings of the OSCAR study, in which we found age adjusted negative correlations between urinary cadmium, or urinary protein HC, and bone density.11a

In conclusion, the results from the present study indicate that renal tubular damage due to exposure to cadmium may develop at lower concentrations than previously anticipated. Even if the tubular dysfunction in itself is a benign condition, it is an indicator of a renal effect that eventually may lead to serious health consequences. Therefore, measures should be taken to reduce exposure to cadmium in the general population, including lowering of current standards for intake of cadmium in food.