Document XwY3ZZbnjXyJwOR4D2ED9B54

Tohoku J. Exp. Med., 2005,205,247-261 The Threshold Cadmium Level That Causes a Substantial Increase in/?2-Microglobulin in Urine of General Populations Masayuki Ikeda, Takafumi Ezaki, Jiro Moriguchi, Yoshinari Fukui, Hmoanco Ukai, Satoru Okamoto and Haruheco Sakurai1 Kyoto Industrial Health Association, Kyoto, and 1 Occupational Health Research and Development Center, Japan Industrial Safety and Health Association, Tokyo, Japan Ikeda, M., Ezaki, T., Moriguchi, J., Sakurai, H. The Threshold Cadmium Level That Causes a Substantial Increase in prMicroglobulin in Urine of General Populations. Tohoku J. Exp. Med., 2005,205(3), 247-261------Cadmium (Cd) is a toxic element ubiq uitous in the environment, and general populations have been exposed to this dement pri- metric mean (GM) Cd-U*, (i.e., Cd-U as corrected for creatinine [cr] concentration) and GM/Sj-MG-Ucr (jSj-MG-U as corrected for cr) of residents in polluted as well as nonpollut- ed areas in Japan were retrieved in international and domestic sources. In practice, 245 cases of the data pairs were obtained in 51 articles published since 1975. Statistical analy sis on< ,, The relation between the two parameters after double-logarithmic conver sion was in a shape of the letter J or a stick for ice hockey. ion poinkgay.ew6daife*s&4^nadwbl^ scales^ Cd-U,,levels that correspond to a^z-MG-Uc, of l,000jug / g cr were estimated to be 8-9 p.%! g cr, by ordinary and logarithmic assumption as well as by the 3rd degree re gression analysis. ----------- cadmium; general population; Japan; /82-microglobulin; urine 2005 Tohoku University Medical Press It is well known that cadmium (Cd) is a nephrotoxic metal and that exposure to Cd primanly via foods may occur among the general population because this element is ubiquitous in the environment including food materials (International Programme on Chemical Safety Received August 25,2004; revision accepted for publication December 28,2004. Address for reprints: Masayuki Ikeda, Kyoto Industrial Health Association, 67 Nishmokyo-Kitatsubmcho, Nakagyo-ku, Kyoto 604-8472, Japan. e-mail: ikeda@kyotokqjohokenkai.onjp 247 248 M. Ikeda etal. 1992a, b). It is further known that long-term en vironmental exposure to Cd even at low doses may result in dysfunction of renal tubules, and that Cd and /g2-microgIobuiin Q82-MG) levels in urine (Cd-U and /32-MG-U) are markers of expo sure to Cd and resulting health effects (tubular dysfunction), respectively (International Programme on Chemical Safety 1992a, b). Thus, it is of public health importance to know whether or not a threshold Cd-U exists in causing a sub stantial increase in /?2-MG-U, especially for the general population in Japan where dietary Cd in take has been higher than those in neighboring countries (Ikeda et al. 2000a). In order to examine the quantitative relation between Cd and /S2-MG in urine, this study group had searched for publications on the two markers, and succeeded to identify 12 articles (Saito et al. 1977; Nogawa et al. 1979; Aoshima 1987; Ishizaki et al. 1989; Iwata 1991; Aoshima et al. 1995a; Ikeda et al. 1995, 2000b; Yamanaka et al. 1998; Oo et al. 2000; Arisawa et al. 2001; Ezaki et al. 2003) by computerized retrieval systems. It was found that the increase in /?2-MG-U was not linearly related to Cd-U and that the increase was markedly accelerated when Cd-U exceeded a cer tain level, so that the over-all relation was in a shape of the letter "J" or an ice hockey stick (Ikeda et al. 2003a). In a succeeding period of further literature retrieval, it was realized that substantial amounts of data (namely, the pairs of Cd-U and /?2-MG-U) were available in domestic media such as the proceedings of the annual meetings of the Cadmium Research Group (organized by Japan Public Health Association). Compilation of such data sets was more than three times of what were published in international journals. Thus, analyses were conducted with this newly-built larger database to reconfirm the previ ous preliminary observation (Ikeda et al. 2003a) that /52-MG-U increases at a much greater rate when Cd-U is high as compared with die rate at low Cd-U. Further efforts were made to detect die point of flexion in the relation between Cd-U* and /J2-MG-Ucr The results are described in die present report Materials and Methods Ethical consideration The protocol of the present analysis was approved by the Ethics Committee of Kyoto Industrial Health Association. Literature survey In addition to search for literature since 1975 by conventional computerized retrieval systems, issues of Kankyo Hoken Report published since 1975 were sorted manually for paired data on Cd-U and/?*-MG-U of popu lations in polluted or non-poiluted areas in Japan. In ad dition, the compilation was supplemented by other data available, e.g., in the proceedings of die annual meetings of Japanese Society for Hygiene and Japanese Society of Public Health. Because of the characteristics of these materials such as meeting proceedings (for Kankyo Hoken Report, see the Discussion section), the same data appeared both in the proceedings and in a full paper in some cases; priority was given to the full paper in citing as a data source. The sorting thus gave 51 articles with 345 pairs of Cd-U and/?2-MG-U as summarized in Table 1 (Kono et al. 1976,1980; Uruno et al. 1976; Ogata et al. 1976; Saito et al. 1977,1997; Honda et al. 1978; Nogawa et al. 1979,1980, 2002; Shircashi et al. 1980; Takegawa et al. 1980; Shitomi et al. 1981; Nakano et al. 1985; Tohyama et al. 1986; Aoshima 1987, 2004; Kido et al. 1987,1988,1997,2004; Saito 1988; Ishizaki et al. 1989; Aoshima et al. 1995a, 1995b, 2000a, 2001,2002; Nishijo et al. 1990, 1991; Nishijo 1999; Iwata et al. 1991; Kasuya et al. 1992; Kodama et al. 1993; Tsuritani et al. 1994; Ikeda et al. 1995,2000b; Kido 1995,2000; Fan et al. 1998; Yamanaka et al. 1998; Yamada 1999; Oo et al. 2000; Suwazono et al. 2000; Arisawa et al. 2001; Cai et al. 2001; Ezaki et al. 2003; Kikuchi et al. 2003; Nakadaira and Nishi 2003; Horiguchi et al. 2004; Sugita et al. 2004). It should be noted that many studies report ed on both polluted and non-poiluted areas (as controls), with men and women separately or in combination. Thus, the `Total' in Table 1 does not necessarily meet with the simple summation of the numbers on the line or the column. Treatmentsfor creatinine-corrected values and geometric means In a majority of articles, Cd-U and y32-MG-U values were presented after correction for creatinine (CR) con centration (i.e., Cd-Ua and yS2-MG-Ua), and in terms of geometric mean (GM) and geometric standard deviation Threshold Cd to Increase/?2-MG in Urine 249 (GSD) for a group of individuals. Three papers (Uruno et al. 1976; Kono et al. 1980; Takegawa et al. 1980) how ever gave uncorrected values only (the values were thus assumed to be as observed), and the values were taken as if they had been corrected for creatinine. When only arithmetic means (AM) and arithmetic standard devia tions (ASD) were available, the corresponding GM and GSD values were estimated by use of the moment meth od (Sugita and Tsuchiya 199S). Medians were given in cases ofTohyama et al. (1986) and Arisawa et al. (2001), and the medians were taken as if they had been GM val ues. It should be noted that the number of individuals represented by a pair of GM Cd-Uff and /J2-MG-Ua were various depending on the articles. Statistical analysis A pair of GM Cd-U,, and GM /I2-MG-Ucr for a group of individuals will be called "a case" in the present analyses. PC software STATVIEW Version 5 was em ployed for statistical analyses including linear and cubic (tile 3rd degree) regression as well as Wilcoxon signed rank test Both Cd-Us and /3rMGa distribute log-nor mally (e.g., Ikeda et al. 1995; Sugita et al. 2004). Accordingly, logarithmic distributions on double-loga rithmic scales were considered in regression analyses, in addition to normal distributions on ordinary scales. Results Distribution of GM Cd-U,, and, p2-MG-U,, in polluted and non-polluted areas A whole data were classified into cases from polluted and nonpolluted areas following the clas sification by the original authors) of each article. and also by the gender of the subjects (men and women either separately or in combination); in practice, 6 articles gave the results for a combina tion of two genders whereas others gave the re sults for men and women separately (Table 1). The GM values for Cd-Uc,. of the cases in non polluted areas distributed in a wide range of up to 7.8 pg l g cr (Table 2). The GM Cd-U^ reported for cases in polluted areas also distributed widely down to the level of 0.8 pg / g cr, indicating that there is a substantial overlapping in the distribu tion of Cd-Uc, between the polluted and nonpol luted areas (Table 2 and Fig. 1). In an area that was selected as a control to a known polluted area (Ogata et al. 1976), the GM /S2-MG-Ucr for residents was as high as 2,000 pgl g cr (GM Cd-Uc being 3.4 pg / g cr) suggesting that the residents in the control area were also af fected by Cd exposure even though less intensive ly than those in the polluted area. Accordingly, the cases in the control area were classified in the present analyses as those in a polluted area, al though separately from those with more intensive Cd exposure. Otherwise, the highest GM @Z-MGU for those in the non-polluted areas was 386 pgf get (Table2). In contrast, the highest GM Cd-Ua and GM /S2-MG-Uer for those in polluted areas (who were in practice Itai-itai disease patients or patients suspected of the disease) were 31.6 pg / g cr and 200 mg (or 200,000 pg) f g cr, respectively (Nogawa et al. 1979). Thus, the GM /3;rMG-U,, Table 1. Number ofreports and number ofpairs ofCd-\Ja andp^MG-U^ available in literature Gender Number of reports (number of cases' cf Cd-U and /?2-MG)b in Polluted areas Non-polluted areas Total' I-I patients and Suspects*. Men Women Mixed Total' 19 (74) 35 (118) 5 (25) 42(217) 13 (49) 20 (75) 2 (4) 23(128) 23(123) 43 (193) 6 (29) 51(345) 1 (1) 8 (8) 1 (1) 9(10) * Each case has a pair of GM values for Cd-Uw andySj-GM-U^. cf a group of subjects. b Number of reports (number of cases in parenthesis). c Because some reports cover both polluted and non-polluted areas or both men and women, the total numbers for reports do not meet with simple summation of the values on the line or the columa 4 Itai-itai disease patients and the patients suspected of the disease. 250 M. Ikedaetal. Table 2. Minimum and maximum valuesfor Cd-U^ andp^-MG-U^ availble in literature Item Polluted areas Cases in Nan-polluted areas Total* I-I and Suspects4 No. of subjects per case Minimum* Maximum* Age (years) Meanb Minimum* Maximum* Cd-U,, (ug/g cr) Minimum* Maximum* /92-MG-U0(ug/gcr) Minimum* Maximum* 1 696 28-83 13 92 0.8 31.6 26 200,000 5 1,323 21-77 13 99 0.2 7.8 32 386 1 1,323 21-78 13 99 02 31.6 26 200,000 1 64 64-83 40 75 6.9 (9.4*) 29.8 4,008 (50,778*) 200,000 * The minimum and the maximum among the individuals. b The range of arithmetic (age) or geometric means (Cd-Ucr and/Sj-MG-U^ for cases. c The sum of cases in polluted and nonpolluted areas. 4 Itai-itai disease patients and subjects suspected of the disease ail in polluted areas. . ' Excluding Shiroishi et al. (1980), who studied un-hospitalized Itai-itai disease-suspected subjects. value for some exposed subjects was more than 500 times higher than the levels for residents in nonpolluted areas (Table 2). Knowing the overlapping in Cd-U^ between die polluted and nonpolluted areas, the cases of paired data on GM Cd-Ucr and GM/82-MG-Ucr from both areas are plotted together in Fig. 2. The scatter diagram on the ordinary scales showed that dots tended to disperse in a wide range when Cd-U,, was high, e.g., in excess of 5 fig I g cr. Further plotting on double-logarithmic scales shows that the over-all distribution is not linear but in a shape of the letter "J" or an ice hockey stick. Analysis with men and women separately showed that the distribution pattern after ordinary and double-logarithmic plotting was essentially the same between the two genders (figures not shown), and the results of analysis with cases of mixed gender was also similar although the num ber of cases was limited (Table 1). Thus, further analyses were conducted with a combination of men, women and the mixed gender. Analysesfor possible point offlexion in the "J" shape In order to estimate the point of flexion in Fig. 2, all cases (i.e., die combination of cases in polluted and nonpolluted areas) were divided at the various Cd-Uc, levels (from 2 to 10 /ug i g cr as the cut-off value), and the correlation with /?2-MG-U,, (in terms of the regression line) for those below the cut-off value was compared with the regression line for the relation above the cut off value, both on ordinary scales as well as on double-logarithmic scales (Table 3). The analysis on the ordinary scales (the top half in Table 3) showed that die regression lines for the cases be low the cut-off value from 2 to 5 /ug / g cr did not vary very much, although the correlation coeffi cients were generally small and insignificant (p > 0.10) in some cases. It should be noted that cases with greater j32-MG-UCT (i.e., log /Jz-MG-U,, > 3) were observed when log Cd-Ucr exceeded 0.6 (or Cd-Ucr > 4 /g / g cr) (Fig. 2B). In a sharp con trast, the regression lines for the cases above the cut-off values became steeper with smaller (i.e.. Threshold Cd to Increasey?2-MG in Urine 251 225.000 200.000 175,000 150,000 1125,000 ~t 00,000 ? 75,000 T 50,000 S. 25,000 0 A 0 5 10 15 20 25 30 35 225,000 200400 ,, 175400 t lMW "Sa 12^000 2 joaooo 1 75.000 J 50,000 ** 25000 0 A 05 10 15 20 25 30 35 Crf--U (il g/g cr) 0 5 10 IS 20 25 30 35 Cd-U (jig/g cr) Eg. 1. Relation of/Jz-MG-Uc, with Cd-Uo- A: Rela tion in polluted areas (n = 217). B: Relation in nonpolluted areas (it = 128). Note that the vertical axis in Eg. IB is magni fied by 100 times for better illustration. e.o o 5.0 4.0 B 1 l---------r n-- -T -t t 3.0 *. 2.0 *: * -V 1.0 0.0 -.75 -.5 -.25 . *--------- i i 0 .25 .S .75 1 1.25 1.5 1.75 log [Cd-U (fig/g cr)] Eg. 2. Relation of /Jz-MG-U,, with Cd- Um in pol luted and non-polluted areas in combination (n = 345). A: Plotting on ordinary scales. B: Plotting on double-logarithmic scales. greater in the absolute value) intercepts on the vertical axis as a function of an increase in the cut-off value (i.e., from 2 to 10/g Cd / g cr). For example, the slope for the group of cases with Cd-U > 10 n% / g cr was almost twice as steep as that for those with Cd-U > 2/dg / g cr, and the ab solute value of die intercept on the vertical axis was more than 4 times greater. The trends were reproduced when the analyses were made on dou ble-logarithmic scales (the bottom half in Table 3). Thus, when the point of flexion was estimat ed as a point of intersection between the two re gression lines cm ordinary scales (left half in Thble 4) (i.e., the first line with lower Cd-UCT of ^ 2 or ^ 5 fig / g cr, and the second one with higher CdUc of > 2 to > 10 fig / g cr), the Cd-U,, value at the intersection point did not vary as a function of the cut-off value for the first line (i.e., 5.1 pig Cd / g cr both for the 2 pg Cd / g cr line [the left most column] and the ^ 5 Cd / g cr line [the second left-most]), but increased substantially (e.g., from 5.1 to 10.8 pg Cd / g cr in case of the left-most column) as a function of the increase in cut-off values for the second line from 2 to 10 pig Cd / g cr (from the top line to the bottom). Such increase suggest that the relation between Cd-U,, and y^-MG-Uc, at higher Cd-U,, is not linear but slightly concave upward. The trends were essentially reproduced when similar analyses were conducted on double-loga rithmic scales (right half in Thble 4). The change in Cd-Ua at the point of intersection was less re markable and was in a relatively narrow range of 3.1 to 5.1 pig Cd / g cr. Cd-UCT of 2^g / g cr (or 252 M. Ikeda et al. Table 3. Regression line parametersfor cases in various Cd-Ua ranges Parameters n a* r On ordinary scales Cd-U^jS 2fig f gcr Cd-U^S 3/ig/gcr Cd-U,,^ 4fig!gcr Cd-UCT 5/rg/gcr Cd-Uct> 2fig/ gcr Cd-Uet> 3fig/gcr Cd-Utt > 4pg/gcr Cd-U,, > 5pg/gcr Cd-Ucr> 6fig/gcr Cd-Ucr> 1 fig t get Cd-Uer> $ fig/ get Cd-U,,> 9fig/gcr Cd-Uer> 10fig f gcr 67 112 4 108 98 17 145 81 30 174 85 27 278 -16,155 3,200 237 -20,362 3,499 200 -25,373 3,835 171 -30,897 4,189 145 -38,222 4,638 130 -45,594 5,067 116 -51,796 5,419 102 -58,186 5,769 91 -67,125 6,237 0.032 0.197 0.184 0.182 0.590 0.592 0.589 0.585 0.582 0.600 0.595 0.587 0.602 P >0.10 >0.05 >0.05 >0.05 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 On double-logarithmic scales Cd-Ua^ 2ftg/gcr Cd-U,,^ 3/Ug/gcr Cd-UCT 4jig / g cr Cd-Uw Sfiglgct Cd-U,,> 2fig/gcr Cd-Uer> 3 fig/get Cd-Utt> 4fig/get Cd-Ua> 5fig/get Cd-U^ > 6 fig/get Cd-Uw> 7 fig /gcr Cd-U^> 8ftg/get Cd-Utr> 9ftg/get Cd-Ua > 10fig /get 67 108 145 174 278 237 200 171 145 130 116 102 91 2.02 2.02 2.02 2.02 0.17 0.85 0.45 0.24 0.18 0.02 -0.04 -0.19 -0.29 0.10 0.18 022 020 3.10 2.45 2.84 3.03 3.09 323 3.27 3.40 3.48 0.114 0217 0.255 0.226 0.709 0.758 0.746 0.627 0.573 0.567 0.537 0.511 0.516 >0.10 <0.05 <0.05 <0.05 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 1 a and are parameters of a regression equation of Y = a + /8X, where X is Cd-Ucr (or log Cd-UCT) and Y is /i2-MG-Uw (or log/Sj-MG-ty on ordinary (or double-logarithmic scales), respectively. N is the number of cases. log Cd-Ucr of 0.3) is well within the levels for cases without substantial increase in prWj-\Ja (Fig. 2A) and that log Cd-U^ of 0.6 (or 4jug Cd / g cr) appeared to be the upper limit (Fig. 2B). The Cd-U,, at the point of intersection for the first regression line with > 4jwg Cd / g cr and the sec ond line with 5a 2 or ^ 5 jug Cd / g cr was 6.7 fig / g cr on the ordinary scales and it was 3.7 fig Cd / g cr on the double-logarithmic scales. Thus, 7 and 4/^gCd/g after rounding of figures appear to be Cd-Ucr for the point of flexion by analyses on ordinary and double-logarithmic scales, re spectively. The corresponding p2-MG-Ha was about 100-300 fig / g cr by the analyses on the two scales. Such/fo-MG-Us, value appeared to be somewhat conservative when compared with the Threshold Cd to IncreaseyS2-MG in Urine 253 Tabue 4. Cd-Uv andfi^MG-U^ at the cross ofthe two regression lines Regression line with lower Cd-UCT Regies- On ordinary scales sion line with Cd-U0 Ipglgct Cd-U^ S 5jUg / g cr higher Cd-U,, Cd -UCT Cd -U^ at fi,-MGU.at (ug/gcr) POP POP POP POP (pg/gcr)(pg/gcr) (ug/gcr)(pg/get) Critical Cd-U,,* On double-logarithmic scales Cd-Ua 5 2/Ug/gcr Cd-UCT 5pglga Cd-U,, aatt PUfU^aGt - POP POP (pgfgct) (pg/gcr) Cd-Ua POP POP (pg! get) (pg! get) Critical Cd-U,,* > 2 5.1 133 5.1 224 6.1 4.1 119 4.1 119 7.6 >3 5.9 136 5.9 245 6.9 3.1 116 3.1 116 7.9 >4 6.7 139 6.7 267 7.6 3.7 118 3.7 118 8.1 >5 7.4 142 7.4 287 8.5 4.0 119 4.2 140 8.2 >6 8.3 146 8.3 311 9.2 4.1 119 4.3 141 8.4 >7 9.0 149 9.1 331 9.7 43 120 4.6 142 8.5 >8 9.6 151 9.6 347 10.3 4.4 120 4.7 143 8.7 >9 10.1 153 10.1 361 10.9 4.7 120 4.9 144 8.8 >10 10.8 156 10.8 379 11.5 4.8 121 5.1 145 9.3 8.9* i'Cd-UCT at/?z-MG-U,, = 1,000/ug/gcr. b POI, point of intersection. e Calculated by the equation of the 3rd degree, Y = 1.956 + 0.006X + 0.575X* + 0.611X3 (r = 0.804), where Y is log (y52-MGa \pg / g crj), and X is log (Cd-U,, (pgfg cr]). observation in non-polluted areas (i.e., ^ 386 jMg/gcr; Table 2). Cd-Ucr levels corresponding to firMG-Ucr of 1,000jug / g cr From the regression equations summarized in Table 2, Cd-U9 was estimated for the levels at which 02-MG-Ue, was equal to 1,000 ftg / g cr. This level of 1,000 pg firMG / g cr was taken as the critical fiz-MG-Va concentration for tubular dysfunction (for rationale, see the Discussion sec tion) due to Cd exposure. As summarized in Table 4 (shown as "critical Cd-Ucr" in the table), the analyses on ordinary scales gave Cd-UCT of 7.6 pg / g cr, and that on double-logarithmic scales resulted in 8.1 pg / g cr. Separately, analyses with total cases (i.e., cases in polluted and non-polluted areas in combi nation) and with assumption of higher order re gression (up to the 6th order) were conducted for correlation coefficients between Cd-Ucr and /02-MG-Uc The analyses car ordinary scales gave the regression coefficients of 0.34, 0.47, 0.48, 0.48, 0.49 and 0.49 for the 1st, 2nd, 3rd, 4th, 5th and 6th order equation, while it was 0.51, 0.64, 0.65, 0.65, 0.65 and 0.65 in the order on doublelogarithmic scales. The results suggest that die coefficient was greater (p < 0.05 by Wilcoxon signed rank test) on double-logarithmic scales than on ordinary scales and that the improvement in the coefficient was small beyond the 3rd degree equation. Thus, the 3rd degree was selected for regression analysis. The calculation gave an equation of Y = 1.956 + 0.006X + 0.575X2 + 0.611X3 (r = 0.647), where X is log Cd-U,, and Y is log 02-MG-Uc (both in pg / g cr). X for Y = 1,000 pg! get was calculated as 8.9 pg / g cr (at the bottom in Thble 4). 254 M. Ikeda et al. 225.000 200.000 ^ 175,000 " 150,000 ^ 125,000 ^ 100J100 | 75,000 i 50,000 ^ 25,000 0 0 5 10 15 20 2$ 30 35 Cd-U (.ug/g cr) Fig. 3. Relation of/SrMG-Ue, with Cd- UM in Itaiitai disease patients and suspects (n = 10). Cd-Ue, andU,,levels in Itai-itai disease patients and the suspects Through the present literature survey, the data on Cd-Ue,. and /J2-MG-U,, levels in urine samples from Itai-itai disease patients and the suspects were available in 9 articles (Honda et al. 1978; Nogawa et al. 1979, 1980; Aoshima 1987; Shiroishi et al. 1980; Takegawa et al. 1980; Nishijo et al. 1991; Kasuya et al. 1992; Aoshima et al. 2000a); the mean ages of subjects in each article were all 65 years or older. The cases are plotted in Fig. 3, in which the two genders are shown as separate dots (for man, one case only). Except for the case reported by Shiroishi et al. (1980) (with relatively low Cd-Uw [6.9 pg / g cr as GM] and relatively low /3z-MG,, [4,008 pg I g cr as GM]) which wore on non-hospitalized sub jects who had been followed up with suspect of Itai-itai disease, Cd-Ucr levels were about 10 pg / g cr or higher and /?2-MG-U6r levels were ^ 50,000jMg (50 mg) / g cr. Whereas /?2-MG-U,, in creased as a function of increasing Cd-Ua, there was a tendency that p2-MG-U,, leveled off when Cd-Ue, was in excess of e.g. 20jug / g cr. Discussion A single journal of Kankyo Hoken Report accounted for some 37% of the articles cited as data sources, or 28% of the cases analyzed. Thus introduction might be necessary because this jour nal is not indexed in conventional publication re trieval systems. The regular issues of this series of publication (usually once a year) contain the proceedings (or expanded abstracts of presenta tions) for annual meetings organized by the Japan Public Health Association under sponsorship of Ministry of the Environment (formerly the Environment Agency) of the Government of Japan. The materials to be published in the pro ceedings are prepared after presentation and dis cussion in the annual meeting, and data were thought valid for statistical evaluation. The arti cles are written basically in Japanese, but recent articles sure usually accompanied by English ab stracts. The critical /?2-MG-Uw concentration as a marker of Cd exposure-associated tubular dys function certainly deserves discussion. Kido et al. (1988) and Iwata et al. (1993) reported that the dysfunction of renal tubules after Cd exposure may become irreversible when /32-MG-U,, levels exceed 1,000pg I g cr. In a long run, an increase in mortality and reduction in life span was also observed among the residents with ^ 1,000 pg fiz-MG / g cr in urine (Iwata et al. 1991; Nishijo et al. 1994; Arisawa et al. 2001). Aoshima et al. (1990, 2000b) proposed the same value of 1,000 pg /8Z-MG / g cr as a screening level for renal dys function among Cd-exposed populations. Similarly, Roels et al. (1997) summarized their experiences in Cd-exposed factory workers that /fc-MG-uria may not be alleviated when the level exceeds 1,500 / g cr. Taking such information together, the level of 1,000 pg 2-MG / g cr was taken in the present analysis as the critical con centration in evaluating tubular dysfunction due to Cd exposure. The reason for the difference in the estimated point of flexion between the present analysis (CdUe, of 4 to 7 pg / g cr) and the previous results (about 10-11 pg / g cr as a combination of results for two genders; Ikeda et al. 2003a), despite the fact that the database for the previous analysis was in fact a substantial part (92 cases in 12 arti cles out of 345 cases in 52 articles, or about one fourth) of the present database, apparently needs discussion. Whereas die database for the previous analysis was limited to the publications in interna tional journals, that for the present analysis in- Threshold Cd to Increase /32-MG in Urine 255 eluded meeting proceedings as described above, so that four times more cases of Cd-Uct and j^-MG-Ua- pairs were made available (Table 1). In order to examine possible sources of the difference, comparison was made between the re gression lines in terms of regression parameters. The regression line for cases in polluted areas in the present analysis was Y = -20,060 + 3.465X (r = 0.581, n = 217), or log Y = 0.862 + 2.347 log X (r = 0.716, n = 217), where Y a nd X are /fe-MG-Ue, and Cd-Uo- (both in pg / g cr), respectively. The counterpart regression lines in the previous analy sis for polluted areas (Ikeda et al. 2003a) were Y = -58,187 + 5.949X (r = 0.738, n = 44) or log Y = -0.667 + 3.755 log X (r = 0.826, n = 44). For nonpolluted areas, the present analyses gave Y = 100 + 17.13X (r = 0.348, n = 128) or log Y = 2.031 + 0.222 log X (r = 0.311, n = 128), whereas the previous analysis gave Y = 138 + 11.42X (r = 0.232, n - 53) or log Y = 2.161 + 0.087 log X(r = 0.126, n = 53). Comparison of the slopes, the in tercepts and the correlation coefficients (Ichibara 1995) of the lines for polluted areas showed that both slopes and intercepts were significantly dif ferent (p < 0.01) irrespective of logarithmic con version, whereas the correlation coefficients were not different (p g 0.10). The two lines for non polluted areas however did not show significant (p ^ 0.10) difference except that the intercepts after logarithmic conversion were different (p < 0.05). Thus, the difference came from the lines for pol luted areas and not from that for nonpolluted ar eas. - Further comparison of two data sets of cases in polluted areas (Fig. 4), one from the present analysis (small circles) and the other from previ ous study (large squares; Ikeda et al. 2003a), dis closed the presence of additional cases in four ar eas in the diagram for the present analysis. They could be classified into four groups in terms of ar eas of distribution (Table 5). The group in Area A in the upper-left comer in the figure included many cases of Itai-itai disease patients and sus pects who were all at advanced ages (AM age of cases was in a range of 65 to 82 years with the el dest at 91 years; e.g., Kasuya et al. 1992; Aoshima 2004). It is conceivable that while Cd-U,, de- 225,000 200,000 k. o 1 T5.000 bo 150,000 St 125,000 => 100/100 T 75,000 50/100 25/100 T ' ` ' 1 ' 1-------- '----- 4 * * *- ** d 9 10 15 20 25 30 35 CO-0 (.us/t cr) Fig. 4. Comparison of the cases in polluted areas for present analysis (small circles; n - 217) with that for a previous analysis (large squares; n = 44). The data for previous analysis are cited from Ikeda et al. (2003). creased during hospitalization, tubular dysfunc tion stayed unimproved or even progressed (Kido et al. 1988; Iwata et al. 1993), so that Cd-UCT lev els were relatively low although/?2-MG-UCT levels were high. In Area B which is just below Area A, there was no case of the patients or suspects (e.g., Ogata et al. 1976; Saito et al. 1997). Even below these areas and close to the horizontal axis is Area C (Table 5). The cases in the area can be better identified in Fig. 4B, as a group of dots which forms a horizontal tail to the left Whereas Cd-U,, was up to 2.9 pg / g cr to suggest moderate pollu tion, there was no increase in /^-MG-U^. The ob servation that 7 cases out of 15 appeared in do mestic sources may suggest the issue of so-called publication-bias in international journals. 256 M. Ikedaetal. Area A Table 5. Cases, by distribution area in Fig. 4, that affect the parameters ofthe regression line No. of Cd-UCT* &-MG-IV No. of AM age0 Aged GMCd-U,, GM/Sj-MG-Uct' cases (ug/gcr) (fig/gcr) 8ubject8b (years) (years) (ug/gcr) (fig/gcr) 10 >5, <17 >48,000 3 64.83 50 153 83.3 91 7.0 16.9 48,671 173,630 Area B 14 >5, 12.5 >5,000, 3 74.2 30 S 30,000 134 78.2 88 5.0 12.5 7,116 29,390 AreaC 15 <3 <1,000 5 45.1 15 80 57.3 69 0.8 2.9 53 169 AreaD 51 >10 <5,000 2 50.8 40 153 75.9 92 10.1 23.4 156 4,590 Values on the first and the second line for each area show the minimum and the maximum in the area. 1 Cd-Ua and range to define the area of distribution. bThe number of individuals in each case. c Arithmetic means were calculated for ages of the individuals in each case, of which the min. and the max. values are shown. "The max. and min. ages of the individuals. * Geometric means were calculated for Cd-U^ or /Jj-MG-Uj, of the individuals in each case, of which the min. and die max. values are shown. A fig. 5. Comparison of the low /?rMG-U,, cases for present analysis with that for the previous analysis on a magnified scale. Note that die vertical axis is magnified by 45 times as compared with that in fig. 4. Although not very clearly seen in Fig. 4A, the en largement of the lower part of the figure (Fig. 5) makes it clear that there is a heavy clustering of 51 cases in Area D, or the area in the right half of the figure and rather close to the horizontal axis (Figs. 4 and 5). The fact that Cd-U0 was >10 pg / g cr indicates heavy exposure to Cd, but the response in terms of elevation in was rather moderate, e.g., < 1,000 / g cr in 12 cas es. Cases with discrepancy in intensity between exposure to Cd and response in terms of tubular dysfunction were observed also in the previous analysis (i.e., 17 cases out of 33 [51%} with Cd- UCT > 10 (xg / g cr; Ikeda et al. 2003a), but it was even more so in the present analysis (51 cases out of 91 (56%); p < 0.10 by chi-squares test). The presence of these cases (e.g., Nakano et al. 1985; Tohyama et al. 1986; Aoshima 1987; Fan et al. 1998; Arisawa et al. 2001; Cai et al. 2001) sug gests that some subjects may be more resistant to Cd nephrotoxicity that others. Underlying mech anism certainly deserves further study. After re moval of the cases in the four areas from the data base of the present analysis, the regression analyses gave equations of Y = -26912.074 + 4111.810X (r = 0.725, n = 127), or log Y = 0.120 Threshold Cd to Increase/fc-MG in Urine 257 + 3.112 log X (r = 0.825, n - 127). The slopes and the intercepts as well as the correlation coef ficients of these equations were no longer differ ent (p > 0.10), as expected, from the results of previous analysis (Ikeda et al. 2003a) except for the intercepts on ordinary scales (p < 0.01). A major limitation of the present analysis with regard to the database is that no mechanism was available to adjust even if the same individu als were repeatedly introduced after different clas sification, e.g., grouping in terms of age in one ar ticle and by place of residence in another. It was possible to make adjustment only when the same data (e.g., GM Cd-U,,) was presented on separate occasions. It should also be noted that many, if not all, studies were conducted at the time before the practice of external quality assurance (Aitio 1996) was brought into environmental health re search. Regarding the method to estimate the point of flexion, the linear regression is apparently the simplest approach, but this approach is ham pered by the fact that the relation of /?2-MG-U,, with Cd-U,, is not linear in the high Cd-UCT range. As a result, the point of intersection varied de pending on the high Cd-Ue, range employed for calculation of the regression line (Ihble 4). Only Cd-U and /?2-MG-U as corrected for creatinine concentration were employed in the present analysis, as was in the previous analysis (Ikeda et al. 2003a). Criticism has been raised since early days on the adequacy of dividing uri nary analyte level by creatinine concentration as a measure to correct for different urine density (e.g., Alessioetal. 1985; Berlin et al. 1985). This study group also observed that bias may be caused by creatinine correction in the evaluation of /?2-MG-U (Ikeda et al. 2003b), because creatinine concentration will decrease at advanced ages (Moriguchi et al. 2004a) simultaneously with physiological increases in Cd-U and /fe-MG-U (Moriguchi et al. 2004b). It is nevertheless a common practice both in occupational and envi ronmental health to express Cd-U and /32-MG-U levels as corrected for creatinine concentration (World Health Organization 1996; American Conference of Governmental Industrial Hygienists 2003). Experiences show that a majority of data available in literature is after correction for creati nine. It is apparently desirable that data without correction or corrected for a specific gravity of urine be also made available. Furthermore, although /?2-MG-U has been most popularly employed in the epidemiology of Cd exposure-associated tubular dysfunction (Lauwerys et aL 1994; Kawada 1995), use of oth er markers in urine, such as ai-microglobulin (at-MG) (Nogawa et al. 1984; Tohyama et al. 1986; Holmqvist et al. 1993; Pless-Mulloli et al. 1998; Uchida et al. 2004) and N-acetyl-/?-Dglucosamindase (NAG) (Kawada 1995; Noonan et al. 2002; Uchida et al. 2004) have also been practiced. It is certainly important to have data on these parameters in parallel with that of Cd-U and /S2-MG-U for more precise evaluation of tubular dysfunction in relation to Cd exposure. Acknowledgements The authors are grateful to Miss F. Ohashi, Kyoto Industrial Health Association, Kyoto, Japan for her skilful PC operation, and to the staff of Japan Public Health Association, Tokyo, Japan for their support to the collection of information, particularly for die copies of articles in Kankyo Hoken Report References Aitio, A. (1996) Chapter 2. Quality assurance. In: Biological Monitoring of Chemical Exposure in the Workplace. Vol. 1. World Health Organiza tion, Geneva, pp. 20-51. Alessio, L., Berlin, A., DelFOrto, A., Toffoletto, F. & Ghezzi, L (1985) Reliability of urinary creati nine as a parameter used to adjust values of uri nary biological indicators. Int. Arch. Occup. Environ. Health, 55,99-106. American Conference of Governmental Industrial Hygienists (2003) Documentation of TLVs* and BEIs*. American Conference of Govern mental Indutrial Hygienists, Cincinnati. Aoshima, K. (1987) Epidemiology of renal tubular dysfunction in the inhabitants of a cadmiumpolluted area in the Jinzu River Basin in Toya ma Prefecture. Tohoku J. Exp. Med., 152, 151-172. Aoshima, K. (2004) Gender-related difference in bone metabolism in Cd-induced nephropathy. Kan kyo Hoken Report, in press, (in Japanese with English abstract) Aoshima, K., Kato, T., Teranishi, H. & Kasuya, M. 258 M. Ikeda et al. (1990) ^i-Microglobulin, total protein, glucose and amino acids in urine as markers for screen ing cadmium-induced kidney disturbance. Jpn. J. Publ. Health, 37,224-240. Aoshima, K., Kawanishi, Y., Fan, J.-J., Cai, Y.-Q., Katoh, T., Teranishi, H. & Kasuya, M. (1995a) Cross-sectional assessment of renal function in the inhabitants of a cadmium-polluted area. Ann. Clin. Lab. Sci., 25,493-503. Aoshima, K., Kawanishi, Y., Fan, J.-J., Tsuritani, I., Yamada, Y. & Kasuya, M. (1995b) Cadmium and lead levels in Mood and urine, and their re lation to renal tubular function in women living in a cadmium-polluted area. Arch. Complex. Environ. Studies, 7,75-79. Aoshima, K., Cai, Y.-Q., Katoh, T., Teranishi, H. & Kasuya, M. (2000a) Pathogenesis of cadmiuminduced renal tubular osteomalacia, Itai-itai dis ease. Kankyo Hoken Report, 7, 211-215. (in Japanese with English abstract) Aoshima, K., Katoh, T., Teranishi, H. & Kasuya, M. (2000b) /?a-Microglobulin, total protein, glu cose and amino acids in urine as tools of screening for Cd-induced renal dysfunction. Jpn. J. Public Health, 37, 381-387. (in Japa nese) Aoshima, K., Fan, J.-J., Katoh, T., Teranishi, H. St Kasuya, M. (2001) Evaluation of bone metabo lism by urinary markers among women in cad mium-polluted areas in Jinzu River Basin. Hokuriku J. Public Health, 28,47. (in Japanese) Aoshima, K., Fan, J.-J., Katoh, T., Teranishi, H. & Kasuya, M. (2002) Longitudinal assessments of renal tubular function and bone metabolism in cadmium nephropathy. Jpn. J. Hyg., 57,432. (in Japanese) Arisawa, K., Nakano, A., Saito, H., Liu, X.-J., Yokoo, M., Soda, M., Koba, T., Takahashi, T. & Kinoshita, K. (2001) Mortality and cancer inci dence among a population previously exposed to environmental cadmium. Int. Arch. Occup. Environ. Health, 74,255-262. Berlin, A., Alessio, L., Sesana, G., DelPOrto, A. & Ghezzi, L (1985) Problems concerning the use fulness of adjustment of urinary cadmium for creatinine and specific gravity. Int. Arch. Occup. Environ. Health, 55,107-111. Cai, Y., Aoshima, K., Katoh, T., Teranishi, H. St Kasuya, M. (2001) Renal tubular dysfunction in male inhabitants of a cadmium-polluted area in Toyama, Japan -- an eleven-year follow-up study. J. Epidemiol., 11,180-189. Ezaki, T., Tsukahara, T., Morigucbi, J., Furuki, K., Fukui, Y., Ukai, H., Okamoto, S., Sakurai, H., Honda, S. & Ikeda, M. (2003) No clear-cut evi dence for cadmium-induced renal tubular dys function among over 10,000 adult women in general Japanese population; a nation-wide large-scale survey. Int. Arch. Occup. Environ. Health, 76,186-196. Fan, J.-J., Aoshima, K., Katoh, T., Teranishi, H. & Kasuya, M. (1998) A follow-up study on renal tubular dysfunction in women living in the cad mium-polluted Jinzu River Basin in Toyama, Japan. Jpn. J. Hyg., 53, 545-557. (in Japanese with English abstract) Holmquist, L., Vesterberg, O. St Presson, B. (1993) Apoiipoprotein D and ai-microglobulin in hu man urine; effect of cadmium exposure. Int. Arch. Occup. Environ. Health, 64,469-472. Honda, R., Kobayashi, E., Nogawa, K., Ishizaki, A., Watanabe, M., Shiroishi, K. & Kato, T. (1978) Analysis of urine from Itai-itai diseasesuspected patients. Kankyo Hoken Report, 44, 132-135. (in Japanese) Horiguchi, H., Oguma, E., Sasaki, S., Miyamoto, K., Ikeda, Y., Machida, M. St Kayama, F. (2004) Dietary exposure to cadmium at close to the current provisional tolerable weekly intake does not affect renal function among female Japa nese farmers. Environ. Res., 95,20-31. Ichihara, K. (1995) Comparison of two regression pa rameters, and comparison of two correlation coefficients. In: Statistics for Bioscience. Nankodo Publishers, Tokyo, pp. pp. 20-22,23. (in Japanese) Ikeda, M., Moon, C.-S., Zhang, Z.-W., Iguchi, H., Watanabe, T., Iwami, O., Imai, Y. & Shimbo, S. (1995) Urinary armicroglobulin, /Si-micro globulin, and retinol-binding protein levels in general populations in Japan with references to cadmium in urine, blood, and 24-hour food du plicates. Environ. Res., 70,35-46. Ikeda, M., Zhang, Z.-W., Shimbo, S., Watanabe, T., Nakatsuka, H., Moon, C.-S., Matsuda-Inoguchi, N. St Higashikawa, K. (2000a) Urban popula tion exposure to lead and cadmium in east and south-east Asia. Sci. Total Environ., 249, 373-384. Ikeda, M., Zhang, Z.-W., Moon, C.-S., Shimbo, S., Watanabe, T., Nakatsuka, H., MatsudaInoguchi, N. St Higashikawa, K. (2000b) Pos sible effects of environmental cadmium expo sure on kidney function in the Japanese general population. Int. Arch. Occup. Environ. Health, 73,15-25 Ikeda, M., Ezaki, T., Tsukahara, T., Moriguchi, J., Ftuuki, K., Fhkui, Y., Ukai, S., Okamoto, S. St Sakurai, H. (2003a) Threshold levels of urinary cadmium in relation to increases in urinary Threshold Cd to Increase^2-MG in Urine 259 /?2-microglobulin among general Japanese pop ulations. Toxicol. Lett., 137,135-141. Ikeda, M., Ezaki, T., Tsukahara, T., Moriguchi, J., Furuki, K., Fukui, Y., Okamoto, S., Ukai, H., Sakurai, H. (2003b) Bias induced by the use of creatinine-corrected values in evaluation of ^2-raicroglobulin levels. Toxicol. Lett., 145, 197-207. International Programme on Chemical Safety (1992a) Environmental Health Criteria. 134. Cadmium. World Health Organization, Geneva. International Programme on Chemical Safety (1992b) Environmental Health Criteria. 135. Cadmium - Environmental Aspects. World Health Orga nization, Geneva. Ishizaki, M., Kido, T., Honda, R., Tsuritani, I., Yamada, Y., Nakagawa, H. & Nogawa, K. (1989) Dose-response relationship between urinary cadmium and jS^microglobulin in a Japanese environmentally cadmium exposed population. Toxicology, 58,121-131. Iwata, KL, Saito, H. & Nakano, A. (1991) Association between cadmium-induced renal dysfunction and mortality: Further evidence. Tohoku J. Exp. Med., 164,319-330. Iwata, K., Saito, H., Moriyama, M. & Nakano, A. (1993) Renal tubular function after reduction of environmental cadmium exposure: A tenyear follow-up. Arch. Environ. Health, 48, 157-163. Kasuya, M., Aoshima, K., Katoh, T., Teranishi, H., Horiguchi, H., Kitagawa, M. & Hagino, S. (1992) Natural history of Itai-itaj disease: A long-term observation on the clinical and labo ratory findings in patients with Itai-itai disease. In: Seventh International Cadmium Conference, edited by M.E. Cook, S.A. Hiscock & H. Morrow, R.A. Cadmium Association, London, pp. 180-192. Kawada, T. (1995) Indicators of renal effects of expo sure to cadmium: N-Acetyl-/J-D-glucosaminidase and others. J. Occup. Health, 37,69-73. Kido, T. (1995) Studies on health effects of cadmium exposure in the general environment Jpn. J. Hyg., 48, 960-972. (in Japanese with English abstract) Kido, T. (2000) Ten-year follow-up of residents in cadmium-polluted areas in Kakehashi River Basin on urinary findings. Kankyo Hoken Re port, 67,301-303. (in Japanese) Kido, T., Honda, R., Tsuritani, I., Yamaya, H., Ishizaki, M., Yamada, Y- & Nogawa, K. (1987) An epi demiological study on renal dysfunction of in habitants in Cd-exposed areas in the Kakehashi River Basin in Ishikawa prefecture. Jpn. J. Hyg., 42, 964-972. (in Japanese with English abstract) Kido, T., Honda, R, Tsuritani, I., Yamaya, H., Ishizaki, M., Yamada, Y. & Nogawa, K. (1988) Progress of renal dysfunction in inhabitants environmen tally exposed to cadmium. Arch. Environ. Health, 43,213-217. Kido, T., Hayano, M., Kobayashi, E., Nogawa, K., Nishijo, M., Tabata, S., Nakagawa, H. & Tsuritani, I. (1997) Human intestinal alkaline phosphatase in urine as a marker of cadmium- induced renal dysfunction. Kankyo Hoken Re port, 61,208-210. (in Japanese) Kido, T., Sunaga, K., Nakagawa, H., Nishijo, M., Nogawa, K. & Kobayashi, E. (2004) Epidemi ological studies on cadmium-induced adverse health effects of inhabitants in Kakehashi River Basin, Ishikawa. Kankyo Hoken Report, in press, (in Japanese with English abstract) Kikuchi, Y., Nomiyama, T., Kumagai, N., Dekio, F., Uemura, T., Takebayashi, T., Nishiwaki, Y., Matsumoto, Y., Sano, Y., Hosoda, K., Watanabe, S., Sakurai, H. & Omae, K. (2003) Uptake of cadmium in meals from the digestive tract of young non-smoking Japanese female volun teers. J. Occup. Health, 45,43-52. Kodama, Y., Matstmo, K., Kawamoto, T., Suenaga, R., Igisu, H. & Matsuoka, M. (1993) Relationship between cadmium and ^2-microglobulin in urine. Kankyo Hoken Report, 69, 262-266. (in Japanese) . Kono, T., Sakai, Y., Ida, N., Nishi, M., Tono, S., Inagi, K., Matsuda, H. & Katoh, M. (1976) Health survey on residents in the Kakehashi River Ba sin. Kankyo Hoken Report, 38, 104-117. (in Japanese) Kono, S., Ohmura, T., Nakagawa, H., Toga, H., Nishi, M. & Matsuo, Y. (1980) Tubular dysfunction and heavy metal excretion in urine. Kankyo Hoken Report, 46,248-254. (in Japanese) Lauwerys, R.R, Bernard, A.M., Roels, H.A. & Buchet, J.-P. (1994) Cadmium: Exposure markers as predictors of nephrotoxic effects. Clin. Chem., 40,1391-1394. Moriguchi, J., Ezaki, T., Tsukahara, T., Fukui, Y,, Ukai, H., Okamoto, S., Shimbo, S., Sakurai, H. & Ikeda, M. (2004a) Decrease in urine specific gravity and urinary creatinine in elderly women. Int. Arch. Occup. Environ. Health, in press. Moriguchi, J., Ezaki, T., Tsukahara, T., Fukui, Y., Ukai, H., Okamoto, S., Shimbo, S., Sakurai, H. & Ikeda, M. (2004b) Effects of aging and tubular dysfunction markers in urine from adult women in non-polluted areas. Int. Arch. Occup. Envi ron. Health, in press. 260 M. Ikedaetal. Nakadaira, H. & Nishi, S. (2003) Effects of low-dose cadmium exposure on biological examinations. Sci. Total Environ., 308,49-62. Nakano, A., Saito, H. & Wakisaka, I. (1985) Studies on urinary cadmium and /?2-microglobulin of residents in cadmium-polluted areas. Res. Rep. Natl. Inst. Environ. Study, Jpn., 84,13-30. (in Japanese with English abstract) Nishijo, M. (1999) Oinical futures of Itai-itai disease patients and cases of renal tubular dysfunction among the inhabitants living in Cd polluted Kakehashi River Basin. Kankyo Hoken Report, 65,80-88. (in Japanese with English abstract) Nishijo, M., Teranishi, H., Morikawa, Y., Iwata, K., Katoh, T., Aoshima, K. & Kasuya, M. (1990) Case-control study on kidney dysfunction among residents in a cadmium-polluted area in Jinzu River Basin. Hokuriku J. Public Health, 17,253-259. (in Japanese) Nishijo, M., Nakagawa, H., Morikawa, Y., Senma, M., Tabata, S., Kitagawa, Y., Kono, T., Teranishi, H., Kido, T. & Kato, T. (1991) Mortality among patients with Itai-itai disease and relevant fac tors. Kankyo Hoken Report, 58, 69-74. (in Japanese) Nishijo, M., Nakagawa, H., Morikawa, Y., Tabata, S., Senma, M., Kitagawa, Y., Kawano, S., Ishizaki, M., Sugita, N., Nishi, M., Kido, T. & Nogawa, K. (1994) Prognostic factors of renal dysfunc tion induced by environmental cadmium pollu tion. Environ. Res., 64,112-121. Nogawa, K., Kobayashi, E, Honda, R. & Ishizaki, A. (1979) Clinico-chemical studies on chronic cadmium poisoning; Part 1. Results of urinary examinations. Jpn. J. Hyg., 34, 407-414. (in Japanese with English abstract) Nogawa, K., Kobayashi, E., Honda, R., Shinoda, A. & Kato, T. (1980) Renal dysfunction among Itaiitai disease patients and suspects. Kankyo Ho ken Report, 46,217-220. (in Japanese) Nogawa, K., Kido, K., Yamada, Y., Tsuritani, I., Honda, R., Ishizaki, M. & Terahata, K. (1984) arMicroglobulin in urine as an indicator of re nal tubular damage caused by environmental cadmium exposure. Toxicol. Lett., 22,63-68. Nogawa, K., Kobayashi, E., Okubo, Y., Suwazono, Y. & Nakagawa, H. (2002) Comparison of the ex cretion of cadmium in 24-hour urine in 1985, 1993 and 1998 in the general population and study on the renal effects of exposure to low cadmium level in a non-cadmium polluted area. Kankyo Hoken Report, 68, 322-335. (in Japa nese with English abstract) Noonan, C.W., Sarasua, S.M., Campagna, D., Kathman, SJ., Lybarger, J.A. & Mtiller, P.W. (2002) Effects of exposure to low levels of en vironmental cadmium on renal biomarkers. Environ. Health Perspect., 110,151-155. Ogata, H., Hara, K., Matsuda, G., Kondo, A., Kazuya, H., Ariyoshi, T., Otsuka, K., Yajima, H., Watanabe, K., Matsuo, R. & Azuma, F. (1976) Kidney dysfunction among residents in SasuShine River Basin, Tsushima, Nagasaki Prefec ture. Kankyo Hoken Report, 38, 118-122. (in Japanese) Oo, Y.K., Kobayashi, E., Nogawa, K., Okubo, Y., Suwazono, Y., Kido, T. & Nakagawa, H. (2000) Renal effects of cadmium intake of a Japanese general population in two areas unpolluted by cadmium. Arch. Environ. Health, 55,98-103. Pless-Mulloli, T., Boettcher, M., Steiner, M. & Berger, J. (1998) ai-Microglobulin: epidemiological indicator for tubular dysfunction induced by cadmium? Occup. Environ. Med., 55,440-44. Roels, H., VanAsche, F.J., Oversteyne, M., De Groof, M., Lauwerys, R.R. & Lison, D. (1997) Re versibility of microglobulinuria in cadmium workers with incipient tubular dysfunction after reduction of exposure. Br. J. Ind. Med., 50, 37-48. Saito, H. (1988) Health of residents in a cadmiumpolluted area after environmental remediation. Kankyo Hoken Report, 54, 134-137. (in Japa nese) Saito, H., Shioji R., Hurukawa, Y., Nagai, K., Arikawa T., Saito, T., Sasaki, Y., Furuyama, T. St Yoshinaga, K. (1977) Cadmium-induced proxi mal tubular dysfunction in a cadmium-polluted area. Contr. Nephrol., 6,1-12. Saito, H., Iwata, T., Moriyama, M,, Sumitomo, M., Ito, K. & Yamazaki, T. (1997) Health survey on residents in Sasu Area, Izuhara, Nagasaki Pre fecture. Kankyo Hoken Report, 61,334-345. (in Japanese) Shiroishi, K., Nakata, J., Watanabe, M-, Shoji, M. & Kitsunetsuka, H. (1980) yS2-Microglobulin in serum and urine of Itai-itai disease patients. Kankyo Hoken Report, 46, 223-227. (in Japa nese) Shitomi, K., Saito, H., Nakano, A., Unakami, H., Takada, K., Sato, T., Furuyama, T., Yoshinaga, K., Arikawa, T. & Nagai, K. (1981) Urinary /Jz-microglobulin in an environmentally cadmi um-polluted area: Studies of generational and sexual differences, and a comparison with the results of proximal tubular function tests. Jpn.J. Nephrol., 23,45-62. (in Japanese with English abstract) Sugjta, M. & Tsuchiya, K. (1995) Estimation of varia tion among individuals of biological half-times Threshold Cd to Increase/?2-MG in Urine 261 of cadmium calculated from accumulation data. Environ. Res., 68,31-37. Sugita, M., Izuno, T., Otahara, Y., Osada, M., Kobayashi, M. & Shimbo, S. (2004) Renal effect of low level cadmium exposure-effect of rice intake and smoking on renal dysfunction in Japan. Kankyo Hoken Report, in press, (in Japanese with English abstract) Suwazono, Y., Kobayashi, E., Okubo, Y., Nogawa, K., Kido, T. & Nakagawa, H. (2000) Renal effects of cadmium exposure in cadmium nonpolluted areas in Japan. Environ. Res., 84,44-55. Takegawa, A., Nogawa, K. & Hagino, N. (1980) Bone biopsy on Itai-itai disease patients and suspects. Kankyo Hoken Report, 46, 203-216. (in Japa nese) Tohyama, C., Kobayashi, E., Saito, H., Sugahara, N., Nakano, A. & Mitane, Y. (1986) Urinary ai-microglobulin as an indicator protein of re nal tubular dysfunction caused by environmen tal cadmium exposure. J. Appl. Toxicol., 6, 171-178. Tsuritani, I., Honda, R., Ishizaki, M., Yamada, Y., Aoshima, K. & Kasuya, M. (1994) Serum bone-type alkaline phosphatase activity in women living in a cadmium-polluted area. Toxicol. Lett., 71,209-216. Uchida, M., Teranishi, H., Aoshima, K., Katoh, T., Kasuya, M. & Inadera, H. (2004) Reduction of erythrocyte catalase and superoxide dismutase activities in male inhabitants of cadmium-pol luted area in Jinzu river basin, Japan. Tbxicol. Lett., 151,451-457. Uruno, T., Otoyama, S., Nakano, N. & Miyatake, K. (1976) Health survey of a population in a cad mium-polluted area in the Yoshino River Basin, Yamagata prefecture. Kankyo Hoken Report, 38,90-104. (in Japanese) World Health Organization (1996) Chapter 3. Select ed metals 3.1 Cadmium. In: Biological Moni toring of Chemical Exposure in Workplace, Vol. I. World Health Organization, Geneva, pp. 52-90. Yamada, Y. (1999) Bone mass decrease and calcium in urine after environmental exposure to cadmi um. Kankyo Hoken Report, 65,89-90. (in Japa nese) Yamanaka, O., Kobayashi, E, Nogawa, K., Suwazono, Y., Sakurada, I. & Kido, T. (1998) Association between renal effects and cadmium exposure in cadmium-nonpolluted area in Japan. Environ. Res., A77,1-8.