Diabetes with early kidney involvement may shorten life expectancy by 16 years

This study aimed to identify the excess risks associated with diabetic patients with early kidney involvement (early diabetic kidney disease). The mortality risks of early diabetic kidney disease, defined as diabetes in early stages 1-3 chronic kidney disease (CKD), were assessed from a cohort of 512,700 adults in Taiwan participating in a health surveillance program from 1994-2008. Three related groups were identified and compared: diabetes without CKD, early diabetic kidney disease, and early CKD without diabetes. Deaths were ascertained through the National Death Registry. One-third of diabetics had early kidney disease, and approximately two-thirds of patients were classified with early CKD due to proteinuria. Patients with early diabetic kidney disease had more lifestyle risks such as inactivity or obesity, which characteristically amplified excess mortality by up to five times. The three-fold increase in all-cause mortality (hazard ratio 3.16) and a 16-year loss in life expectancy made early diabetic kidney disease a serious and yet often overlooked disease, with most patients unaware of their kidney involvement. Mortality for early diabetic kidney disease was nearly twice as high as that for early CKD (hazard ratio 2.01) or diabetes without CKD (hazard ratio 1.79). The 16-year life span loss is much worse than individually from early CKD (six years) or diabetes (ten years). Thus, identifying early proteinuria among diabetic patients and realizing the importance of reducing lifestyle risks like inactivity is a clinical challenge, but can save lives. This study aimed to identify the excess risks associated with diabetic patients with early kidney involvement (early diabetic kidney disease). The mortality risks of early diabetic kidney disease, defined as diabetes in early stages 1-3 chronic kidney disease (CKD), were assessed from a cohort of 512,700 adults in Taiwan participating in a health surveillance program from 1994-2008. Three related groups were identified and compared: diabetes without CKD, early diabetic kidney disease, and early CKD without diabetes. Deaths were ascertained through the National Death Registry. One-third of diabetics had early kidney disease, and approximately two-thirds of patients were classified with early CKD due to proteinuria. Patients with early diabetic kidney disease had more lifestyle risks such as inactivity or obesity, which characteristically amplified excess mortality by up to five times. The three-fold increase in all-cause mortality (hazard ratio 3.16) and a 16-year loss in life expectancy made early diabetic kidney disease a serious and yet often overlooked disease, with most patients unaware of their kidney involvement. Mortality for early diabetic kidney disease was nearly twice as high as that for early CKD (hazard ratio 2.01) or diabetes without CKD (hazard ratio 1.79). The 16-year life span loss is much worse than individually from early CKD (six years) or diabetes (ten years). Thus, identifying early proteinuria among diabetic patients and realizing the importance of reducing lifestyle risks like inactivity is a clinical challenge, but can save lives. Diabetes continues to be an important global health challenge despite increased public awareness, intensive monitoring, and aggressive management. The current number of more than 415 million people with diabetes estimated worldwide may increase to over 642 million by 2040.1International Diabetes FederationIDF Diabetes Atlas.7 ed. International Diabetes Federation, Brussels, Belgium2015Google Scholar, 2World Health OrganizationDiabetes Fact Sheets. World Health Organization, Geneva, Switzerland2015Google Scholar Diabetes is associated with increased mortality and morbidity, particularly for cardiovascular disease, stroke, or renal failure, but also for several types of cancer.3Giovannucci E. Harlan D.M. Archer M.C. et al.Diabetes and cancer: a consensus report.Diabetes Care. 2010; 33: 1674-1685Crossref PubMed Scopus (1396) Google Scholar, 4Seshasai S.R. Kaptoge S. et al.Emerging Risk Factors CollaborationDiabetes mellitus, fasting glucose, and risk of cause-specific death.N Engl J Med. 2011; 364: 829-841Crossref PubMed Scopus (1884) Google Scholar Our knowledge and efforts in managing diabetes-related complications have brought about a substantial decline in their rate of occurrence in the past 2 decades,5Jansson S.P.O. Andersson D.K.G. Svardsudd K. Mortality Trends in Subjects With and Without Diabetes During 33 Years of Follow-up.Diabetes Care. 2010; 33: 551-556Crossref PubMed Scopus (61) Google Scholar, 6Fox C.S. Coady S. Sorlie P.D. et al.Trends in cardiovascular complications of diabetes.JAMA. 2004; 292: 2495-2499Crossref PubMed Scopus (466) Google Scholar but a large burden of disease still persists as diabetes prevalence continues to increase. The major challenge, however, is to make these patients aware of the seriousness of the condition so that early intervention is possible. Diabetic kidney disease (DKD) is, by definition, diabetes with kidney involvement, presenting as albuminuria and/or an impaired glomerular filtration rate (GFR).7National Kidney FKDOQI Clinical Practice Guideline for Diabetes and CKD: 2012 Update.Am J Kidney Dis. 2012; 60: 850-886Abstract Full Text Full Text PDF PubMed Scopus (945) Google Scholar, 8Altemtam N. Russell J. El Nahas M. A study of the natural history of diabetic kidney disease (DKD).Nephrol Dial Transplant. 2012; 27: 1847-1854Crossref PubMed Scopus (91) Google Scholar, 9Atkins R.C. Zimmet P. 2010 International Society of Nephrology/International Federation of Kidney Foundations World Kidney Day Steering Committee (RA); International Diabetes Federation (PZ)Diabetic kidney disease: act now or pay later.J Bras Nefrol. 2010; 32: 7-10Crossref PubMed Scopus (6) Google Scholar, 10de Boer I.H. Rue T.C. Hall Y.N. et al.Temporal trends in the prevalence of diabetic kidney disease in the United States.JAMA. 2011; 305: 2532-2539Crossref PubMed Scopus (671) Google Scholar, 11Reutens A.T. Epidemiology of diabetic kidney disease.Med Clin North Am. 2013; 97: 1-18Abstract Full Text Full Text PDF PubMed Scopus (184) Google Scholar DKD, the most common cause of end-stage renal disease (ESRD), has increased by 34% in the US in the last 20 years, despite increasing use of medications that lower blood sugar or blood pressure.10de Boer I.H. Rue T.C. Hall Y.N. et al.Temporal trends in the prevalence of diabetic kidney disease in the United States.JAMA. 2011; 305: 2532-2539Crossref PubMed Scopus (671) Google Scholar Advanced cases of DKD are irreversible and beyond the opportunity for prevention. On the other hand, early DKD, defined as diabetes with CKD stages 1 to 3, clinically expressed as proteinuria and mild loss of kidney function, has a high chance of being medically controlled or even reversed to normal. For the medical community outside the nephrology specialty, however, there is limited vigilance to detect early DKD because its serious nature is not well-known. With the availability of a large cohort of half a million individuals,12Wen C.P. Cheng T.Y.D. Tsai M.K. et al.All-cause mortality attributable to chronic kidney disease: a prospective cohort study based on 462 293 adults in Taiwan.Lancet. 2008; 371: 2173-2182Abstract Full Text Full Text PDF PubMed Scopus (730) Google Scholar we analyzed the characteristics and mortality outcomes of early DKD. Life expectancy for individuals with DKD was then calculated and compared with that for those with “diabetes without CKD” or “CKD without diabetes.” Analysis of CKD subjects from this same cohort has been reported previously.13Fox C.S. Matsushita K. Woodward M. et al.Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without diabetes: a meta-analysis.Lancet. 2012; 380: 1662-1673Abstract Full Text Full Text PDF PubMed Scopus (737) Google Scholar, 14Hallan S.I. Matsushita K. Sang Y.Y. et al.Age and association of kidney measures with mortality and end-stage renal disease.JAMA. 2012; 308: 2349-2360Crossref PubMed Scopus (428) Google Scholar, 15Mahmoodi B.K. Matsushita K. Woodward M. et al.Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without hypertension: a meta-analysis.Lancet. 2012; 380: 1649-1661Abstract Full Text Full Text PDF PubMed Scopus (329) Google Scholar A total of 512,700 subjects were identified; among them, 27,455 (5.4%) had diabetes. One-third of those with diabetes (9067 or 33.3%) had early DKD, whereas two-thirds (18,388 or 66.7%) had no kidney involvement. Approximately 50,977 participants (9.9%) had early CKD without diabetes, and 434,268 participants (84.7%) constituted a reference group with neither diabetes nor CKD. Their baseline characteristics are summarized in Table 1. Early DKD patients were older (mean age: 59.3 years) and 54.1% were men. Two-thirds (64.6%) had hypertension, and one-quarter (27.6%) had high serum cholesterol levels. Compared with early CKD, early DKD exhibited more proteinuria as a whole (72.3% vs. 60.8%), and twice as much advanced or serious proteinuria (17.5% vs. 8.5% for 1+ and 13.8% vs. 4.1% for ≥2+). An overwhelming majority (98%) of individuals with early DKD were unaware of their kidney involvement, because a large proportion of them came from less-educated groups. The lifestyle risks of DKD subjects were more substantial than those of other groups, with more than 50% physically inactive, 25% smoking, and 10% obese.Table 1Baseline demographics and clinical characteristics of participants by diabetes and chronic kidney diseaseTotal participantsReference populationaParticipants with neither diabetes nor CKD.Diabetes without CKDEarly CKDbStages 1 to 3. without diabetesEarly DKDcDiabetes with CKD stages 1 to 3.N512,700434,26818,38850,9779,067Mean age (SD)41.6 (14.0)39.7 (12.8)51.1 (11.8)50.2 (16.5)59.3 (11.7)Age (%) 20–3953.158.312.932.46.9 40–5933.232.552.830.938.5 ≥6013.89.234.336.754.6Male (%)49.849.352.252.354.1Lower education level (middle school or below, %)26.322.254.643.566.3Chronic kidney disease (%) Stage 11.80015.214.4 Stage 24.50038.040.0 Stage 35.00046.845.6 Total proteinuria7.50060.872.3 Trace (±)5.60048.241.1 1+1.2008.517.5 ≥2+0.8004.113.8HypertensiondDefined as blood pressure ≥ 140/90 mm Hg, self-reported history of hypertension, or use of antihypertensive drugs. (%)18.013.745.535.564.6High serum cholesteroleDefined as total cholesterol ≥ 240 mg/dl. (%)10.79.420.115.227.6Lifestyle risk factors (anyone of the following, %)61.861.962.660.562.6 Physical inactivity53.553.948.552.150.8 Smoking23.923.524.027.525.9 DrinkingfDefined as regular drinking with ≥ 3 times/wk and 2 drinks/time.17.417.119.019.718.4 ObesitygDefined as body mass index > 30.4.03.39.96.012.3Unawareness of disease statusN/AN/A44.496.898.1CKD, chronic kidney disease; DKD, diabetic kidney disease; N/A, not applicable.a Participants with neither diabetes nor CKD.b Stages 1 to 3.c Diabetes with CKD stages 1 to 3.d Defined as blood pressure ≥ 140/90 mm Hg, self-reported history of hypertension, or use of antihypertensive drugs.e Defined as total cholesterol ≥ 240 mg/dl.f Defined as regular drinking with ≥ 3 times/wk and 2 drinks/time.g Defined as body mass index > 30. Open table in a new tab CKD, chronic kidney disease; DKD, diabetic kidney disease; N/A, not applicable. During an average follow-up period of 8 years, a total of 18,263 deaths occurred. Age-adjusted mortality rates for early DKD participants (1182/100,000) were 3 times higher than for the reference group (365.5/100,000) (Table 2). (The Kaplan-Meier survival curve is shown in Supplementary Figure S1.) After controlling for potential confounders, when compared with the reference group, early DKD patients had a 3-fold increase in all-cause mortality (hazard ratio [HR]: 3.16; 95% confidence interval [CI]: 3.0, 3.4) compared with participants with diabetes and no CKD (HR: 1.76; 95% CI: 1.6, 1.9) or participants with early CKD and no diabetes (HR: 1.58; 95% CI: 1.5, 1.7). The increased mortality was also found for cardiovascular disease mortality, with an HR of 1.74 (95% CI: 1.6, 1.9) for early CKD without diabetes, 1.40 (95% CI: 1.2, 1.7) for diabetes without CKD, and 2.78 (95% CI: 2.4, 3.2) for early DKD. In early DKD patients, mortality for all cancers (HR: 1.75; 95% CI: 1.5, 2.0), lung cancer (HR: 1.39; 95% CI: 1.1, 1.8), liver cancer (HR: 3.18; 95% CI: 2.5, 4.0), and colorectal cancer (HR: 2.28; 95% CI: 1.6, 3.2) were all significantly increased, with greater increases than for other groups. Additional causes of increased mortality were diabetes, kidney diseases, respiratory system disorders, and infectious diseases (Table 2).Table 2Hazard ratios for all causes and cause-specific mortality by diabetes and chronic kidney disease statusAdjusted mortality ratedExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.ReferenceaParticipants with neither diabetes nor CKD.N = 434,268Diabetes without CKDN = 18,388Early CKD without diabetesbCKD stages 1 to 3.N = 50,977Early DKDcDiabetes with CKD stages 1 to 3.N = 9067365.5616.4606.21182.4Deaths (n)HReAdjusted by 11 variables: age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use.Deaths (n)HReAdjusted by 11 variables: age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use.(95% CI)Deaths (n)HReAdjusted by 11 variables: age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use.(95% CI)Deaths (n)HReAdjusted by 11 variables: age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use.(95% CI)All causes9440116001.76dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.6, 1.9)51861.58dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.5, 1.7)20373.16dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(3.0, 3.4)CVD164512841.40dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.2, 1.7)13291.74dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.6, 1.9)4312.78dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(2.4, 3.2) Stroke71711231.26(0.9, 1.7)5641.64dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.4, 1.9)1822.74dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(2.2, 3.4) CHD3941941.97dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.4, 2.7)3562.04dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.7, 2.5)1403.86dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(3.0, 5.0)Diabetes154133522.8dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(17.2, 30.3)1641.87dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.3, 2.7)53349.66dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(37.6, 65.6)Kidney diseases671284.69dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(2.6, 8.6)2699.56dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(6.4, 14.4)12119.79dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(12.4, 31.5)All cancer429215351.42dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.3, 1.6)17341.28dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.2, 1.4)4411.75dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.5, 2.0) Lung cancer9481931.02(0.8, 1.3)3651.04(0.9, 1.2)791.39dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.1, 1.8) Liver cancer87311482.19dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.7, 2.8)3751.74dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.5, 2.1)1233.18dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(2.5, 4.0) Colorectal cancer4061611.57dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.1, 2.2)1721.24dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.0, 1.6)622.28dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.6, 3.2) Bladder cancer441121.83(0.6, 5.4)502.54dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.3, 4.9)62.55(0.9, 7.0)Respiratory system6001691.28(0.9, 1.8)4221.42dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.2, 1.7)1052.46dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.9, 3.2)Infectious disease1351261.58(0.8, 3.1)851.59dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.1, 2.4)242.34dExpressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.(1.2, 4.5)CI, confidence interval; CHD, coronary heart disease; CKD, chronic kidney disease; CVD, cardiovascular disease; DKD, diabetic kidney disease; HR, hazard ratio.a Participants with neither diabetes nor CKD.b CKD stages 1 to 3.c Diabetes with CKD stages 1 to 3.d Expressed as per 100,000 person-years, with sex and age standardized to the 2009 Taiwanese population.e Adjusted by 11 variables: age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use. Open table in a new tab CI, confidence interval; CHD, coronary heart disease; CKD, chronic kidney disease; CVD, cardiovascular disease; DKD, diabetic kidney disease; HR, hazard ratio. A direct comparison of mortality risks, with the use of a forest plot in Figure 1, between early DKD and early CKD without diabetes (left) or diabetes without CKD (right) yielded the following findings. All-cause mortality for early DKD was nearly twice as high compared with early CKD without diabetes (HR: 2.01) or diabetes without CKD (HR: 1.79). Cardiovascular disease mortality for early DKD was also higher (HR: 1.61 vs. 1.96), including stroke (HR: 1.72 vs. 2.14), coronary heart disease (HR: 1.85 vs. 1.93), and kidney disease (HR: 1.97 vs. 4.03). All-cancer mortality for early DKD was also significantly increased (HR: 1.37 vs. 1.22), including liver cancer (HR 1.85 vs. 1.40) and colorectal cancer (HR 1.87 vs. 1.39). These significant increases for early DKD compared with the other groups remained twice as high when DKD status was reassessed for those participants returning for second or third screenings (Supplementary Table S1). For early DKD, all-cause mortality risks were significantly higher across all levels of estimated GFR (eGFR; Figure 2a), proteinuria (Figure 2b), fasting glucose (Figure 2c), and blood pressure (Figure 2d) when compared with other groups. This significantly higher mortality persisted even when levels of eGFR, proteinuria, fasting glucose, and systolic blood pressure were at normal values. At age 30 years, the life expectancy of participants with early DKD was 14.8 years shorter for men and 16.9 years shorter for women compared with the reference group. At age 50 years, the life expectancy of participants with early DKD was 11.5 and 14.1 years shorter for men and women, respectively, compared with the reference group. In comparison, at age 30 years life expectancy was 10.2 years (men) and 11.7 years (women) shorter for the diabetes without CKD group, and 5.7 years (men) and 6.7 years (women) shorter for the CKD without diabetes group (Figure 3, Supplementary Table S2). Four lifestyle risk factors—physical inactivity, smoking, drinking and obesity—were more prevalent among participants with early DKD compared with the reference population. Two-thirds of participants with early DKD had at least 1 of the 4 risk factors. Each was associated with increased mortality among the reference group (Figure 4 and Supplementary Figure S2), but amplified such risks among those with early DKD. For example, the excess mortality risk of physical inactivity in the reference group was 80.4/100,000, the difference between mortality risk for inactive (361.8) versus active (281.4) lifestyles. This excess mortality risk was 5 times higher for physical inactivity among participants with early DKD, 446.5/100,000 (difference between inactive [1317.2/100,000] and active [870.7/100,000]). Similar observations were seen for obesity, smoking, and drinking, with 5-fold, 50%, and 2-fold amplified mortality risk differences, respectively, for participants with early DKD. In this study, we found that life expectancy for patients with diabetes complicated by early CKD was shortened by an average of 16 years. Mortality risks for this population were doubled compared with those for diabetic patients without CKD or early CKD patients without diabetes, and tripled compared with those for the reference population. Notably, however, an overwhelming majority of participants with early DKD (98%) were not aware of having such a serious condition. Several characteristics of early DKD were noteworthy when compared with the other 3 groups, diabetes without CKD, early CKD without diabetes, and the reference group. First, up to three-quarters of early DKD patients had proteinuria, a condition more prevalent and more serious than early CKD, and early detection of proteinuria is a cornerstone of early DKD management.9Atkins R.C. Zimmet P. 2010 International Society of Nephrology/International Federation of Kidney Foundations World Kidney Day Steering Committee (RA); International Diabetes Federation (PZ)Diabetic kidney disease: act now or pay later.J Bras Nefrol. 2010; 32: 7-10Crossref PubMed Scopus (6) Google Scholar, 16Bowman B.T. Kleiner A. Bolton W.K. Comanagement of diabetic kidney disease by the primary care provider and nephrologist.Med Clin North Am. 2013; 97: 157-173Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar Checking GFR alone may be inadequate for detecting most early DKD cases. Diabetes patients often focused on checking urine for sugar levels but not regularly checking it for protein levels, a process easily achieved with the urine dipstick method. Second, given the same level of proteinuria, early DKD had more ominous outcomes and higher mortality than early CKD (Figure 2b). Furthermore, heavier proteinuria was found 2 to 3 times more in early DKD and had made DKD outcome even worse. Third, given the same fasting blood glucose level, the mortality outcome was consistently worse for early DKD than for the other groups (Figure 2c). This suggests that good glycemic control alone, reducing fasting glucose level as low as possible among early DKD patients, was inadequate for best survival outcomes. This is in line with the observation in clinical trials that intensive lowering of blood glucose levels among patients with diabetes did not result in significant mortality reduction.17Gerstein H.C. Miller M.E. et al.Action to Control Cardiovascular Risk in Diabetes Study GEffects of intensive glucose lowering in type 2 diabetes.N Engl J Med. 2008; 358: 2545-2559Crossref PubMed Scopus (6456) Google Scholar, 18Kelly T.N. Bazzano L.A. Fonseca V.A. et al.Systematic review: glucose control and cardiovascular disease in type 2 diabetes.Ann Intern Med. 2009; 151: 394-403Crossref PubMed Scopus (292) Google Scholar Similarly, controlling systolic blood pressure in hypertension per se was inadequate for the best mortality results in early DKD patients (Figure 2d). The inadequacy of treating early DKD with 1 risk factor at a time marks DKD as a unique disease. The important role of lifestyle risks in early DKD has not been fully appreciated. First, lifestyle risks are ubiquitous in early DKD, with two-thirds of patients having at least 1 of the 4 lifestyle risks (i.e., physical inactivity, smoking, alcohol drinking, and obesity). Each of these risks would amplify the mortality risk of early DKD beyond its impact on the general public. For example, the excess mortality risk of physical inactivity was 5 times greater in early DKD than in the non-DKD group, as shown in Figure 4. Similar observations, but with less impact on absolute mortality, were seen for obesity, smoking, and drinking, where early DKD amplified mortality risk differences (Supplementary Figure S2). Through these amplifications, the role of lifestyle risks became particularly important in early DKD. At the same time, the degree of benefit that could be gained through regular exercise was 5-times greater in those with early DKD compared with an average individual. Such a gain has been shown among fully active DKD participants in this cohort.19Wang I.K. Tsai M.K. Liang C.C. et al.The role of physical activity in chronic kidney disease in the presence of diabetes mellitus: a prospective cohort study.Am J Nephrol. 2013; 38: 509-516Crossref PubMed Scopus (11) Google Scholar Because more than one-half of participants with early DKD were physically inactive, engaging in regular exercise should be emphasized as a clinical routine. Incidentally, exercise has been reported to demonstrate the added benefit of reducing proteinuria in diabetes.20Robinson E.S. Fisher N.D. Forman J.P. Curhan G.C. Physical activity and albuminuria.Am J Epidemiol. 2010; 171: 515-521Crossref PubMed Scopus (40) Google Scholar Similar to physical inactivity, smoking was reported in a sizable portion of participants with early DKD, and quitting smoking should also have been a clinical priority. However, in reality, the majority of participants with early DKD were unaware of it, and the few who were aware of their diabetes status were preoccupied with blood sugar control or blood pressure control, with little attention paid to the importance of modifying their lifestyle risks. Thus, reducing lifestyle risks should become a centerpiece of treatment for every early DKD case discovered in the primary care setting. Reducing lifestyle risks will also help lower the cancer risk increase found in early DKD. Early DKD demonstrated significantly increased cancer risks (1.75) compared with diabetes (1.42) and early CKD (1.28). To our knowledge, this is the first study highlighting the large life-shortening effect and strong cancer relationship with DKD. Viewing early DKD as a unique and serious condition on its own will facilitate its detection and management. In summary, early DKD is, in this study, characterized by 3 distinct observations: highly prevalent proteinuria with 16-year loss of life, larger-than-average benefits if lifestyle changes are made, and unsatisfactory survival outcome when the treatment is targeted at 1 risk at a time, lowering blood sugar or lowering blood pressure. There are important limitations in this study. First, proteinuria was determined by a single urine sample measurement, not meeting the requirement of persistent albuminuria in the Kidney Disease Outcomes Quality Initiative definition of kidney damage.21Eknoyan G. Levin N.W. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification - foreword.Am J Kidney Dis. 2002; 39: S14-S266Abstract Full Text Full Text PDF Scopus (9) Google Scholar Therefore, DKD prevalence in our study population might have been overestimated. However, up to 75% of proteinuria was reported to persist.10de Boer I.H. Rue T.C. Hall Y.N. et al.Temporal trends in the prevalence of diabetic kidney disease in the United States.JAMA. 2011; 305: 2532-2539Crossref PubMed Scopus (671) Google Scholar The estimated hazard ratios for early DKD, however, would probably be biased toward the null hypothesis due to nondifferential misclassification when we applied the same criteria to the reference group. Second, the classification of early DKD was based on the blood and urine tests at the initial examination and ignored the temporal changes in subsequent time periods. However, major results for early DKD persisted when second or third returning examinations were considered, as shown in Supplementary Table S1. Third, urinary albumin-creatinine ratio (ACR), the gold standard for measuring albuminuria, is preferred by most nephrologists in establishing the diagnosis of DKD. However, the higher cost and time-consuming nature of ACR testing, and the low prevalence of proteinuria in the general public (7%),22Khwaja A. Throssell D. A critique of the UK NICE guidance for the detection and management of individuals with chronic kidney disease.Nephron Clinical Practice. 2009; 113: C207-C212Crossref PubMed Scopus (12) Google Scholar, 23Wen C.P. Yang Y.C. Tsai M.K. Wen S.F. Urine dipstick to detect trace proteinuria: an underused tool for an underappreciated risk marker.Am J Kidney Dis. 2011; 58: 1-3Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar made ACR impractical for screening healthy subjects in general practice.24White S.L. Yu R. Craig J.C. et al.Diagnostic accuracy of urine dipsticks for detection of albuminuria in the general community.Am J Kidney Dis. 2011; 58: 19-28Abstract Full Text Full Text PDF PubMed Scopus (126) Google Scholar Instead, we used the machine-read dipstick method to detect proteinuria. The comparability of the dipstick method and ACR for determining albuminuria has been well-reported.23Wen C.P. Yang Y.C. Tsai M.K. Wen S.F. Urine dipstick to detect trace proteinuria: an underused tool for an underappreciated risk marker.Am J Kidney Dis. 2011; 58: 1-3Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar, 25Matsushita K. van der Velde M. et al.Chronic Kidney Disease Prognosis CAssociation of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis.Lancet. 2010; 375: 2073-2081Abstract Full Text Full Text PDF PubMed Scopus (2830) Google Scholar Finally, the serious outcome of early DKD may be a health disparity or inequality issue, with more DKD found among the less-educated population. Our cohort comprised fee-paying participants for medical screening, with higher socioeconomic status and health consciousness, and therefore a lower rate of early DKD. Thus, due to possible selection bias, the public health implications of early DKD in Taiwan may be much more serious. In conclusion, the higher mortality and shorter life expectancy associated with early DKD are worse than those commonly perceived for overall diabetes. Due to the more serious outcome of DKD even at its early stage, we propose that early DKD be considered a disease in its own right to attract public attention. Special efforts are needed for prevention, early identification, and proper management of early DKD. This prospective cohort study consisted of 543,412 adults who participated in a self-paying comprehensive health surveillance program offered by a private firm (MJ Health Management Institution, Taiwan) between 1994 and 2008. Data including blood test or urine sample such as serum creatinine, eGFR, and urine protein test were nearly 95% complete for 513,926 individuals. The majority of participants returned for repeated examinations in subsequent years, but only results from the initial tests were used in our analysis. Those who returned for second and third screening evaluations were included in a sensitivity analysis to validate the results. A detailed description of the study has been documented elsewhere.12Wen C.P. Cheng T.Y.D. Tsai M.K. et al.All-cause mortality attributable to chronic kidney disease: a prospective cohort study based on 462 293 adults in Taiwan.Lancet. 2008; 371: 2173-2182Abstract Full Text Full Text PDF PubMed Scopus (730) Google Scholar, 26Wen C.P. Wai J.P. Tsai M.K. et al.Minimum amount of physical activity for reduced mortality and extended life expectancy: a prospective cohort study.Lancet. 2011; 378: 1244-1253Abstract Full Text Full Text PDF PubMed Scopus (1225) Google Scholar The protocol was approved by the institutional review board of the National Health Research Institutes. Consent was obtained from all participants. Data related to individual identification were removed, and participants remained anonymous for the entire study period. In addition to a self-administered questionnaire for medical history, each participant underwent a standard panel of medical tests including blood tests, urine tests, body measurements, functional tests, and physical examinations. Overnight fasting blood and first morning voided urine samples were collected and analyzed. Diabetes was defined as fasting plasma glucose level ≥ 126 mg/dl (7 mmol/l), self-reported history of diabetes, or use of any anti-diabetes medication. CKD was defined based on the Kidney Disease Outcomes Quality Initiative definition,21Eknoyan G. Levin N.W. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification - foreword.Am J Kidney Dis. 2002; 39: S14-S266Abstract Full Text Full Text PDF Scopus (9) Google Scholar with eGFR estimated by the CKD-EPI Study equation,27Levey A.S. Stevens L.A. Schmid C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-612Crossref PubMed Scopus (15959) Google Scholar and/or albuminuria determined by dipstick method and reported as negative, trace (+/–), 1 (+), and ≥ 2 (+), equivalent to ACR <10, 10 to 29, 30 to 299, and ≥ 300 mg/ml, respectively.25Matsushita K. van der Velde M. et al.Chronic Kidney Disease Prognosis CAssociation of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis.Lancet. 2010; 375: 2073-2081Abstract Full Text Full Text PDF PubMed Scopus (2830) Google Scholar Early CKD was defined as CKD stages 1 to 3 without diabetes. Early DKD was defined as diabetes with CKD stages 1 to 3. CKD stage 1 was defined as eGFR ≥ 90 ml/min per 1.73 m2 with positive urinary protein, and stage 2 was defined as eGFR ≥ 60 ml/min per 1.73 m2 with positive urinary protein. Stage 3 was defined as eGFR 30 to 59 ml/min per 1.73 m2. Participants were classified into the following 4 groups: diabetes without CKD, early CKD without diabetes, diabetes with early CKD (early DKD), and a reference group with neither CKD nor diabetes. Individuals with CKD stage 4 (n = 817) and 5 (n = 409), either with or without diabetes, were excluded from the analysis. All the classifications were based on one-off measurements. We assessed the early CKD status and associated mortality for those who returned in 1 or 2 years for repeated screening. One of 3 (34%) had 2 tests, and 1 of 5 (18%) had 3 tests. Questions regarding awareness of nephritis or kidney disease and awareness of diabetes or currently taking diabetes medication were used to assess CKD and diabetes awareness, respectively. National death file or National Cancer Registry file were matched with this cohort for mortality and cancer incidence statistics. Participants were monitored from the date of baseline measurement until 31 December 2008 or death. A Cox proportional hazards model was used for HRs with 11 variables adjusted, including age, sex, body mass index, education level, systolic blood pressure, cholesterol, smoking, drinking, physical activity, Chinese herbal medicine use, and analgesic use. The proportional hazard assumption was examined and met by plotting the log-minus-log survival curves and survival times against cumulative survival. The age-standardized mortality rate was adjusted for the 2009 Taiwanese general population. The life table method was used to estimate remaining years in life or life expectancy.28Chiang C.L. The Life Table and its Applications.1st ed. Krieger Publishing, Malabar, FL1984Google Scholar, 29Wen C.P. Tsai S.P. Chung W.S. A 10-year experience with universal health insurance in Taiwan: measuring changes in health and health disparity.Ann Intern Med. 2008; 148: 258-267Crossref PubMed Scopus (287) Google Scholar All the authors declared no competing interests. CPW, CHC, MKT, and XW conceptualized and designed the study. MKT, JHL, PJL, and CPW analyzed and interpreted the data. CPW and CHC drafted the article and submitted it for publication. CPW, CHC, SPT, CW, CHC, CWK, and XW critically revised the article for important intellectual content. SPT, MKT, JHL, and PJL provided administrative, technical, and logistical support. CKT was responsible for collection and assembly of the data. All authors had final approval of the article. This study is supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial Center (MOHW106-TDU-B-212-113004). Download .docx (.01 MB) Help with docx files Table S1Hazard ratios for all-cause mortality among study participants who obtained second and third screening tests. Download .docx (.02 MB) Help with docx files Table S2Shortened life expectancy for early diabetic kidney disease and comparisons with the reference, early chronic kidney disease without diabetes, and diabetes without chronic kidney disease groups by gender. Download .docx (.04 MB) Help with docx files Figure S1Kaplan-Meir survival curve for early diabetic kidney disease (DKD) and comparisons with the reference, early chronic kidney disease (CKD) without diabetes, and diabetes without CKD groups. Download .docx (.13 MB) Help with docx files Figure S2Relative risk (hazard ratios) increased in early diabetic kidney disease (DKD) compared with participants with neither diabetes nor chronic kidney disease (CKD) by different lifestyle risk factors.