This study is the first to report significant positive associations between kidney Cd levels in biopsies from living donors and urinary calcium excretion. We observed that donors with high kidney Cd (above the median) had significantly higher U-Ca excretion normalized for creatinine than those with low kidney Cd, mainly due to a significant difference in women but not men. We also found a significant positive association between U-Ca excretion in the 24-hour sample and kidney Cd in a multiple linear regression model, including age, sex, weight, menopause, ionized serum calcium, urinary flow rate and vitamin D (25(OH)D3). We believe that the association was stronger in 24-hour urine because that sample was collected during a longer time than the overnight sample, and should therefore give a more reliable measure of the mean urinary calcium concentration.
Men excreted significantly more U-Ca per hour than women, in accordance with previous studies . However, after normalizing for creatinine (Ca/creatinine ratio), there was a trend for women to excrete more calcium than men. The tendency towards stronger associations between kidney Cd and U-Ca when using dichotomized instead of continuous kidney Cd could indicate a threshold effect, very low kidney cadmium levels not affecting U-Ca.
Cadmium is known to affect both kidney function and the skeleton, as was first described as the Itai-Itai disease in Japan . Later studies have confirmed an increased risk for osteoporosis and/or fractures after Cd exposure [8–13]. We focused on the relationship between kidney Cd and urinary calcium excretion, but we also measured BMD in part of the study population. The effect of Cd on calcium metabolism is interesting as urinary calcium is a marker of tubular function, but also because calcium is the most important mineral in the bone tissue. Some previous studies have shown a positive association between urinary Cd, as a marker of Cd body burden, and calcium excretion [17, 19]. However, urinary Cd is not an ideal indicator of the body burden because Cd excretion may increase when kidney function is impaired, and Cd may be co-excreted with urinary proteins [21, 26, 27]. One of the strengths with this study is the use of kidney Cd, which eliminates the risk of a non-causal association due to diuresis or co-excretion of Cd and protein.
The effect of kidney Cd on urinary calcium excretion was mainly seen in women. This is supported by a study of residents in Thailand with elevated urinary Cd, where the fractional excretion of calcium increased with rising cadmium exposure (urinary Cd) especially in women . In that study, there was also a positive correlation between fractional calcium excretion and the excretion of bone resorption markers, and the levels of those markers were significantly higher in women. This could indicate that women are at higher risk for bone effects resulting from cadmium exposure than men.
The mechanism behind the increased urinary calcium excretion in donors with high kidney Cd in this study is not clear, but it could be an effect of early renal tubular dysfunction  or a direct effect on bone, possibly by increased bone resorption [2, 5, 12, 14, 15, 20]. A third possibility is that calcium deficiency, following calcium loss in the urine, has led to an increased expression of essential metal transporters in the intestine, thereby also increasing the absorption of cadmium [6, 7]. An increased calcium excretion can be both the cause of, and an effect of, increased bone resorption, which can result in low BMD (osteopenia or osteoporosis) and increased risk of fractures [10, 11, 13].
In the present study, we found significant negative correlations between kidney Cd and BMD in the univariate analyses, but no significant associations in the multiple regression analyses, where body weight was the predominant factor and positively associated to BMD. Other possible confounders or effect modifiers included in the model were age, sex, menopause, smoking and vitamin D. The absence of significant associations could indicate that the increased Ca excretion is not associated with low BMD at these low Cd levels. However, with BMD measured in only 67 of the 109 donors with data on kidney Cd levels, our small group size provides limited power to find an association between kidney Cd and BMD. In addition, most of our subjects may have been too young to have an effect of Cd on BMD (median age 51 years).
Increasing age, female gender, estrogen deficiency, e.g. at menopause, low BMI and weight, osteoporosis in the family history, smoking and previous fractures are known risk factors for low BMD . There is also a positive association between physical activity and bone mass. We had no information about menopausal age, osteoporosis in the family or level of physical activity, which is a limitation of the study. However, menopause was assumed for women aged >51 years, which is the median age for menopause . We also had no information about dietary calcium intake, which may contribute to variability in calcium levels. Vitamin D is usually affected by season, which was confirmed in this study (data not shown). These sources of variability in covariates could lead to some misclassification, but such misclassification should be non-differential.
The biopsies in the study were taken from living kidney donors. In other circumstances it would not be ethical to take kidney biopsies from healthy people, but at this transplantation center a “baseline” biopsy is normally taken from the donated kidney, as part of the routine at transplantation. Thus the study group does not include subjects with diabetes or other serious systemic diseases. However, we consider them representative for the healthy part of the general population.