Association of Ultraprocessed Food Consumption with Risk of Cardiovascular Disease Among Individuals with Type 2 Diabetes: Findings from the UK Biobank.

SCOPE: Among patients with diabetes, who have modified nutritional behavior and a higher risk of cardiovascular disease (CVD), the influence of ultraprocessed foods (UPFs) on CVD remains unknown. The study aims to evaluate the association between UPF intake and the risk of CVD among individuals with type 2 diabetes (T2D) and further examine the potential biological pathways linking the association. METHODS AND RESULTS: This study includes 5405 participants with T2D who provided at least one 24-h dietary recall from the UK Biobank study. In the fully adjusted models, a 10% increase in the proportion of UPFs is associated with higher hazards of overall CVD (hazard ratio [HR]: 1.10; 95% confidence interval [CI]: 1.04, 1.15), coronary heart disease (HR: 1.10; 95% CI: 1.04, 1.16), heart failure (HR: 1.14; 95% CI: 1.05, 1.25), but not stroke (HR: 1.01; 95% CI: 0.90, 1.12). Cystatin C, high-density lipoprotein cholesterol (HDL-C), apolipoprotein A, C-reactive protein, and body mass index collectively explain 26.9% (12.8%, 48.5%) of the association between UPF intake and the risk of overall CVD. CONCLUSION: Higher UPF intakes are associated with increased hazards of CVD among individuals with T2D, and the association is partly mediated through worsening biomarkers of renal function, lipid metabolism, inflammation, and body weight.

Trends in Self-Reported Adherence to Healthy Lifestyle Behaviors Among US Adults, 1999 to March 2020.

Importance: Adherence to a healthy lifestyle is associated with lower risks of adverse outcomes. However, trends in multiple lifestyle factors and overall healthy lifestyle status among US adults in recent years are unknown. Objective: To examine trends in multiple lifestyle factors and overall healthy lifestyle among US adults. Design, Setting, and Participants: This serial cross-sectional study used nationally representative data from 10 National Health and Nutrition Examination Survey (NHANES) cycles (nine 2-year cycles from 1999 to 2016 and 1 combined cycle from 2017 to March 2020) among adults 20 years or older. Data were analyzed from December 10, 2021, to January 11, 2023. Exposure: Survey cycle. Main Outcomes and Measures: Five healthy lifestyle factors: never smoking, moderate or lighter alcohol consumption (for women: ≤7 drinks/wk; for men: ≤14 drinks/wk), healthy diet (Healthy Eating Index-2015 scores ≥60.0), sufficient physical activity (≥150 min/wk of equivalent moderate physical activity), and healthy weight (body mass index [calculated as weight in kilograms divided by height in meters squared] 18.5-24.9). Results: A total of 47 852 adults were included in this study. The weighted mean [SE] age was 47.3 [0.2] years; 24 539 (weighted proportion, 51.5%) were women. From the 1999-2000 cycle to the 2017 to March 2020 cycle, the estimated prevalence of the 5 lifestyle factors showed divergent trends, with increasing prevalence of never smoking (from 49.4% [95% CI, 46.4%-52.4%] to 57.7% [95% CI, 55.5%-59.9%]; difference, 8.2% [95% CI, 4.5%-12.0%]), healthy diet (from 19.3% [95% CI, 16.0%-22.6%] to 24.5% [95% CI, 21.5%-27.5%]; difference, 5.2% [95% CI, 0.8%-9.7%]), and sufficient physical activity (from 55.7% [95% CI, 51.8%-59.6%] to 69.1% [95% CI, 67.2%-71.1%]; difference, 13.4% [95% CI, 9.0%-17.8%]), while prevalence of healthy weight decreased from 33.1% (95% CI, 30.5%-35.6%) to 24.6% (95% CI, 22.6%-26.7%; difference, -8.4% [95% CI, -11.8% to -5.1%]) (all P < .001 for trend). Meanwhile, there was no significant trend in moderate or lighter alcohol consumption. Overall, the estimated prevalence of at least 4 healthy lifestyle factors increased from 15.7% (95% CI, 12.8%-18.7%) to 20.3% (95% CI, 17.8%-22.7%; difference, 4.5% [95% CI, 0.7%-8.4%]; P < .001 for trend). Disparities in healthy lifestyle were widened by age group, with little improvement among adults 65 years and older (difference, 0.04% [95% CI, -4.28% to 4.35%]). There were persistent disparities in healthy lifestyle by race and ethnicity, educational level, and income level. Conclusions and Relevance: The findings of this cross-sectional study of NHANES data over a 22-year period suggest diverse change patterns across 5 healthy lifestyle factors and a modest improvement in overall lifestyle existed among US adults, with worsening or persistent disparities in lifestyle.

Trends in the Prevalence of Metabolically Healthy Obesity Among US Adults, 1999-2018.

Importance: Improved understanding of trends in the proportion of individuals with metabolically healthy obesity (MHO) may facilitate stratification and management of obesity and inform policy efforts. Objectives: To characterize trends in the prevalence of MHO among US adults with obesity, overall and by sociodemographic subgroups. Design, Setting, and Participants: This survey study included 20 430 adult participants from 10 National Health and Nutrition Examination Survey (NHANES) cycles between 1999-2000 and 2017-2018. The NHANES is a series of cross-sectional and nationally representative surveys of the US population conducted continuously in 2-year cycles. Data were analyzed from November 2021 to August 2022. Exposures: National Health and Nutrition Examination Survey cycles from 1999-2000 to 2017-2018. Main Outcomes and Measures: Metabolically healthy obesity was defined as a body mass index of 30.0 (calculated as weight in kilograms divided by height in meters squared) without any metabolic disorders in blood pressure, fasting plasma glucose (FPG), high-density lipoprotein cholesterol (HDL-C), or triglycerides based on established cutoffs. Trends in the age-standardized prevalence of MHO were estimated using logistic regression analysis. Results: This study included 20 430 participants. Their weighted mean (SE) age was 47.1 (0.2) years; 50.8% were women, and 68.8% self-reported their race and ethnicity as non-Hispanic White. The age-standardized prevalence (95% CI) of MHO increased from 3.2% (2.6%-3.8%) in the 1999-2002 cycles to 6.6% (5.3%-7.9%) in the 2015-2018 cycles (P < .001 for trend). There were 7386 adults with obesity. Their weighted mean (SE) age was 48.0 (0.3) years, and 53.5% were women. The age-standardized proportion (95% CI) of MHO among these 7386 adults increased from 10.6% (8.8%-12.5%) in the 1999-2002 cycles to 15.0% (12.4%-17.6%) in the 2015-2018 cycles (P = .02 for trend). Substantial increases in the proportion of MHO were observed for adults aged 60 years or older, men, non-Hispanic White individuals, and those with higher income, private insurance, or class I obesity. In addition, there were significant decreases in the age-standardized prevalence (95% CI) of elevated triglycerides (from 44.9% [40.9%-48.9%] to 29.0% [25.7%-32.4%]; P < .001 for trend) and reduced HDL-C (from 51.1% [47.6%-54.6%] to 39.6% [36.3%-43.0%]; P = .006 for trend). There was also a significant increase in elevated FPG (from 49.7% [95% CI, 46.3%-53.0%] to 58.0% [54.8%-61.3%]; P < .001 for trend) but no significant change in elevated blood pressure (from 57.3% [53.9%-60.7%] to 54.0% [50.9%-57.1%]; P = .28 for trend). Conclusions and Relevance: The findings of this cross-sectional study suggest that the age-standardized proportion of MHO increased among US adults from 1999 to 2018, but differences in trends existed across sociodemographic subgroups. Effective strategies are needed to improve metabolic health status and prevent obesity-related complications in adults with obesity.

Differences in HDL-related mortality risk between individuals with and without hypertension: a prospective cohort study in UK Biobank.

BACKGROUND AND AIMS: Very high levels of high-density lipoprotein cholesterol (HDL-C) have been paradoxically linked to increased mortality risk. The present study aimed to examine associations of HDL-C and varied sizes of the HDL particle (HDL-P) with mortality risk stratified by hypertension. METHODS AND RESULTS: This prospective cohort study included 429 792 participants (244 866 with hypertension and 184 926 without hypertension) from the UK Biobank. During a median follow-up of 12.7 years, 23 993 (9.8%) and 8142 (4.4%) deaths occurred among individuals with and without hypertension, respectively. A U-shaped association of HDL-C with all-cause mortality was observed in individuals with hypertension after multivariable adjustment, whereas an L-shape was observed in individuals without hypertension. Compared with individuals with normal HDL-C of 50-70 mg/dL, those with very high HDL-C levels (>90 mg/dL) had a significantly higher risk of all-cause mortality among individuals with hypertension (hazard ratio, 1.47; 95% confidence interval, 1.35-1.61), but not among those without hypertension (1.05, 0.91-1.22). As for HDL-P, among individuals with hypertension, a larger size of HDL-P was positively whereas smaller HDL-P was negatively associated with all-cause mortality. After additional adjustment for larger HDL-P in the model, the U-shaped association between HDL-C and mortality risk was altered to an L-shape among individuals with hypertension. CONCLUSIONS: The increased risk of mortality associated with very high HDL-C existed only in individuals with hypertension, but not in those without hypertension. Moreover, the increased risk at high HDL-C levels in hypertension was likely driven by larger HDL-P.

This study examined the potential modification of hypertension on associations of high-density lipoprotein cholesterol (HDL-C), especially at a very high level, and varied sizes of HDL particle (HDL-P) with the risk of mortality.Very high HDL-C levels were associated with increased risk of mortality in individuals with hypertension, but not in those without hypertension.In individuals with hypertension, the increased risk at a high HDL-C level was attributed to a larger size of HDL-P, which was directly associated with mortality risk. An inverse association with mortality was observed for a smaller size of HDL-P.

Association of Combined Healthy Lifestyles With Cardiovascular Disease and Mortality of Patients With Diabetes: An International Multicohort Study.

OBJECTIVE: To prospectively examine the associations of combined lifestyle factors with incident cardiovascular disease (CVD) and mortality in patients with diabetes. PATIENTS AND METHODS: Patients with prevalent diabetes were included from 5 prospective, population-based cohorts in China (Dongfeng-Tongji cohort and Kailuan study), the United Kingdom (UK Biobank study), and the United States (National Health and Nutrition Examination Survey and National Institutes of Health-AARP Diet and Health Study). Healthy lifestyle scores were constructed according to non-current smoking, low to moderate alcohol drinking, regular physical activity, healthy diet, and optimal body weight; the healthy level of each lifestyle factor was assigned 1 point, or 0 for otherwise, and the range of the score was 0 to 5. Cox proportional hazards models were used to estimate hazard ratios for incident CVD, CVD mortality, and all-cause mortality adjusting for sociodemographic, medical, and diabetes-related factors, and outcomes were obtained by linkage to medical records and death registries. Data were collected from October 18, 1988, to September 30, 2020. RESULTS: A total of 6945 incident CVD cases were documented in 41,350 participants without CVD at baseline from the 2 Chinese cohorts and the UK Biobank during 389,330 person-years of follow-up, and 40,353 deaths were documented in 101,219 participants from all 5 cohorts during 1,238,391 person-years of follow-up. Adjusted hazard ratios (95% CIs) comparing patients with 4 or 5 vs 0 or 1 healthy lifestyle factors were 0.67 (0.60 to 0.74) for incident CVD, 0.58 (0.50 to 0.68) for CVD mortality, and 0.60 (0.53 to 0.68) for all-cause mortality. Findings remained consistent across different cohorts, subgroups, and sensitivity analyses. CONCLUSION: The international analyses document that adherence to multicomponent healthy lifestyles is associated with lower risk of CVD and premature death of patients with diabetes.

Associations of combined healthy lifestyles with cancer morbidity and mortality among individuals with diabetes: results from five cohort studies in the USA, the UK and China.

AIMS/HYPOTHESIS: Cancer has contributed to an increasing proportion of diabetes-related deaths, while lifestyle management is the cornerstone of both diabetes care and cancer prevention. We aimed to evaluate the associations of combined healthy lifestyles with total and site-specific cancer risks among individuals with diabetes. METHODS: We included 92,239 individuals with diabetes but without cancer at baseline from five population-based cohorts in the USA (National Health and Nutrition Examination Survey and National Institutes of Health [NIH]-AARP Diet and Health Study), the UK (UK Biobank study) and China (Dongfeng-Tongji cohort and Kailuan study). Healthy lifestyle scores (range 0-5) were constructed based on current nonsmoking, low-to-moderate alcohol drinking, adequate physical activity, healthy diet and optimal bodyweight. Cox regressions were used to calculate HRs for cancer morbidity and mortality, adjusting for sociodemographic, medical and diabetes-related factors. RESULTS: During 376,354 person-years of follow-up from UK Biobank and the two Chinese cohorts, 3229 incident cancer cases were documented, and 6682 cancer deaths were documented during 1,089,987 person-years of follow-up in the five cohorts. The pooled multivariable-adjusted HRs (95% CIs) comparing participants with 4-5 vs 0-1 healthy lifestyle factors were 0.73 (0.61, 0.88) for incident cancer and 0.55 (0.46, 0.67) for cancer mortality, and ranged between 0.41 and 0.63 for oesophagus, lung, liver, colorectum, breast and kidney cancers. Findings remained consistent across different cohorts and subgroups. CONCLUSIONS/INTERPRETATION: This international cohort study found that adherence to combined healthy lifestyles was associated with lower risks of total cancer morbidity and mortality as well as several subtypes (oesophagus, lung, liver, colorectum, breast and kidney cancers) among individuals with diabetes.

Associations of Urinary Trichloroacetic Acid Concentrations with Spermatozoa Apoptosis and DNA Damage in a Chinese Population.

Exposure to trichloroacetic acid (TCAA) has been associated with impaired semen quality; however, its association with spermatozoa apoptosis and DNA damage remains unclear. We, therefore, collected single semen and repeated urine samples from male partners of couples attending a reproductive center, which were measured for spermatozoa apoptosis and DNA damage parameters and TCAA concentrations, respectively. Multivariable linear regression models were used to explore the associations between urinary TCAA concentrations and spermatozoa apoptosis (n = 462) and DNA damage parameters (n = 512). After adjusting for potential confounders, positive dose-response relationships were found between urinary TCAA concentrations and percentage of tail DNA (tail%) and tail-distributed moment (TDM) (both p for trend <0.10). Compared with men in the lowest tertile of urinary TCAA concentrations, men in the highest tertile had a greater tail% and TDM of 6.2% (95% CI: 0.7, 12.2%) and 8.9% (95% CI: -1.9, 20.5%), respectively. Urinary TCAA concentrations were unrelated to spermatozoa apoptosis parameters in a dose-response manner. However, urinary TCAA concentrations were positively associated with the percentage of Annexin V+/PI- spermatozoa (apoptotic cells), when urinary TCAA concentrations were modeled as continuous variables. Our results suggest that exposure to TCAA at concentrations in real-world scenarios may be associated with spermatozoa apoptosis and DNA damage.

Trajectories of metabolic risk factors during the development of type 2 diabetes in Chinese adults.

AIMS: China has the largest number of adults with diabetes. Although multiple metabolic risk factors (MRFs) are implicated in the development of diabetes, it remains unclear how they progress during the development of diabetes among Chinese. We examined trajectories of multiple MRFs among Chinese and identified the critical period when drastic changes occurred during the development of diabetes. METHODS: This prospective cohort study included participants since 2006-2007 in the Kailuan study. People attended biennial examinations until 2017 with additions of new participants at each examination cycle. The time when a participant first completed the examination was served as the baseline. A total of 122,659 participants without prevalent diabetes at baseline and with complete follow-up data were included. MRFs were collected via biennial physical examinations and laboratory measures. Incident diabetes cases were identified via biennial fasting glucose tests and self-reported physician-diagnosis. RESULTS: During up to 12 years of follow-up, 14,922 incident diabetes cases were identified. Compared with participants who did not develop diabetes, those who developed diabetes had more adverse levels of most MRFs at baseline and during follow-up. Abrupt increases in multiple MRFs (including fasting glucose, surrogate insulin resistance indicators, lipids, systolic blood pressure, pulse pressure, heart rate, alanine aminotransferase, and C-reactive protein) were observed 3 years before the diagnosis of diabetes. CONCLUSIONS: We identified 3 years before diabetes diagnosis as a critical period when multiple MRFs experienced drastic changes. This would have implications for early monitoring and timely prevention for individuals who experience sudden adverse progression of multiple MRFs.

Association of handgrip strength with semen characteristics: a study with repeated measurements among healthy Chinese men.

Accumulating epidemiological evidence shows that handgrip strength provides predictive potential in physical, mental, and reproductive health status. However, the associations between handgrip strength and semen characteristics have not been explored. We recruited 1382 eligible men at the Hubei Province Human Sperm Bank (Wuhan, China) who had their handgrip strength measured at recruitment and provided 6458 repeated semen specimens within a 6-month period. Semen characteristics, including semen volume, sperm motility parameters (immotility, nonprogressive motility, and progressive motility), and sperm concentration, were assessed. Mixed-effect models and restricted cubic spline functions were applied to investigate the relationship of handgrip strength with repeated measurements of semen characteristics. After adjusting for confounding factors, the mixed-effect models revealed that handgrip strength was positively associated with semen volume, sperm concentration, progressive motility, total motility, and total count (all P for trend < 0.05). Compared to men in the lowest quartile, those in the highest quartile of handgrip strength had higher semen volume, sperm concentration, progressive motility, total motility, and total count, with measurements of 14.2% (95% confidence interval [CI]: 5.9%-23.2%), 19.5% (95% CI: 7.3%‒33.1%), 9.5% (95% CI: 3.4%‒15.9%), 8.8% (95% CI: 3.2%‒14.6%), and 36.4% (95% CI: 18.9%‒56.5%), respectively. These positive dose-response relationships were further confirmed in restricted cubic splines, where handgrip strength was modeled as a continuous variable. Handgrip strength, as an indicator of muscular function and strength, was positively associated with semen characteristics in a dose-dependent manner.

Low Concordance Between Blood Pressures Measured in Periodic Health Examinations and in a Workplace-Based Hypertension Management Program.

Poor adherence to standard protocols of blood pressure (BP) measurement in routine clinical practice leads to higher readings than "research-quality" measurements. Whether this phenomenon exists in periodic health examinations was unknown. We aimed to explore the concordance between BP measurements in periodic health examinations and those measured following a standard measurement protocol. We used data from the Kailuan Study, an ongoing longitudinal cohort study in China, of which participants received biennial health examinations in health management centers. In addition, BPs were measured following standard protocols in a workplace-based hypertension management program nested in the Kailuan Study. We compared BP readings of the same person between the two settings using generalized linear mixed-effects models. A total of 3988 men (the mean age was 44.9 years) had at least two BP measurements both in health examinations and management program with a time interval between the two settings that less than 90 days. The mean systolic blood pressures (SBP) and diastolic blood pressures (DBP) in health examinations were 4.2 (95% CI 3.9-4.5) mm Hg and 3.3 (95% CI 3.1-3.5) mm Hg higher than those in the management program, respectively. Bland-Altman analyses showed the wide agreement intervals ranging from - 27.7- to 36.5-mm Hg for SBP and - 18.3- to 24.7-mm Hg for DBP. In conclusion, BP measurements in periodic health examinations were generally higher than BPs measured following a standard protocol. Our findings highlight the importance of standard BP measurement to avoid overestimation of hypertension prevalence and treatment initiation. Supplementary Information: The online version contains supplementary material available at 10.1007/s43657-022-00067-w.

Effectiveness of a Workplace-Based, Multicomponent Hypertension Management Program in Real-World Practice: A Propensity-Matched Analysis.

Limited studies had investigated the potential benefits of workplace-based hypertension management programs on long-term blood pressure (BP) control and health outcomes. This study used the propensity score matching to examine the effectiveness of a workplace-based hypertension management program on BP control and risks of major adverse cardiovascular events and all-cause mortality. Within the Kailuan study, a workplace-based hypertension management program was initiated in 2009 among men with hypertension, which included regular BP measuring (twice a month), free antihypertensive medications, and individualized health consultation. Participants were followed until loss to follow-up, death, or December 31, 2019. Among 17 724 male hypertensives aged 18 to 60 years, 6400 participated in the program. The propensity score matching yielded 6120 participants in the management group and 6120 participants in the control group. Both systolic and diastolic BPs were significantly lower in the management group than in the control group over follow-up, and the mean between-group difference at the 10th year was -7.83 (95% CI, -9.06 to -6.62) mm Hg for systolic BP and -4.72 (95% CI, -5.46 to -3.97) mm Hg for diastolic BP. Participants in the program were more likely to achieve BP control (odds ratio, 1.70 [95% CI, 1.41-2.06]) and had significantly lower risks of major adverse cardiovascular events (hazard ratio, 0.83 [95% CI, 0.72-0.94]) and all-cause mortality (hazard ratio, 0.71 [95% CI, 0.58-0.86]), compared with those who were not in the program. A workplace-based hypertension management program was related to reduced BP levels and lower risks of major adverse cardiovascular events and mortality in Chinese men with hypertension.

Identifying windows of susceptibility to essential elements for semen quality among 1428 healthy men screened as potential sperm donors.

BACKGROUND: Essential elements such as iron (Fe), cobalt (Co), copper (Cu), zinc (Zn), selenium (Se), rubidium (Rb), strontium (Sr), and molybdenum (Mo) are necessary for reproductive health. However, their associations with human semen quality remain inconclusive. OBJECTIVES: To investigate the associations of urinary Fe, Co, Cu, Zn, Se, Rb, Sr, and Mo concentrations with semen quality in healthy men screened as potential sperm donors and identify critical windows of susceptibility. METHODS: 1428 healthy men provided 3766 urine and 6527 semen samples, which were measured for urinary essential element concentrations and sperm quality parameters, respectively. Linear mixed models and cubic spline curves were used to evaluate associations between urinary essential elements and semen quality. Multiple informant models were used to identify potential critical windows of susceptibility. RESULTS: Linear mixed models and cubic spline curves showed positive dose-response relationships between urinary Zn and sperm concentration and total count and between urinary Mo and total sperm count [all False Discovery Rate (FDR) adjusted p-value for trend < 0.05]. In the multiple-element linear mixed models, the men in the highest versus lowest quartiles of urinary Zn and Mo had a higher sperm concentration of 17.5% (95% CI: 2.8%, 34.2%; p-value for trend = 0.006) and total sperm count of 18.3% (95% CI: 1.4%, 38.0%; p-value for trend = 0.027), respectively. Urinary Zn was also positively associated with total sperm count in a dose-dependent manner (p-value for trend = 0.036), though the percentile difference in total sperm count between men in the highest and lowest quartile was not statistically significant (16.4%, 95% CI: -1.7%, 37.9%). These associations appeared to be stronger when urinary Zn and Mo were measured at 0-9 days before the date of semen examination (i.e., corresponding to epididymal storage). CONCLUSIONS: Higher urinary Zn and Mo, particularly during the period of epididymal storage, were associated with greater sperm production.

Temporal variability of organophosphate flame retardant metabolites in spot, first morning, and 24-h urine samples among healthy adults.

A single measurement of organophosphate flame retardant (OPFR) metabolites in a spot sample is often used in epidemiological studies to estimate individual exposures. Over seven consecutive days, we collected 661 spot samples, including 127 first morning voids (FMVs) and 123 simulated 24-h collections, from 20 healthy adults and analyzed for eight OPFR metabolites. Intraclass correlation coefficients (ICCs) were calculated to evaluate the variability of the analyzed metabolites. In spot samples group, serial measurements of OPFR metabolites showed poor reproducibility (0.0422 ≤ ICC ≤ 0.349), and the within-day variability was the main contributor of the total variability. The estimated ICCs based on different correction methods for urine dilution (i.e., specific gravity-adjusted, creatinine-adjusted, and creatinine as a covariate) were similar, but varied according to gender and body mass index. Uniformly low sensitivities (0.417-0.633) were observed when using a single FMV or spot sample to predict the 1-week highly (top 33.0%) exposed volunteers. Therefore, using a single urinary measurement to predict chronic exposure to OPFRs can lead to a high degree of classification errors. When multiple urine samples are collected, considering the sampling type, the time of collection, and demographic characteristics may provide a more complete approach to assess exposure to diverse OPFRs.

Reproducibility of essential elements chromium, manganese, iron, zinc and selenium in spot samples, first-morning voids and 24-h collections from healthy adult men.

Evaluation of Cr, Mn, Fe, Zn and Se in humans is challenged by the potentially high within-individual variability of these elements in biological specimens, which are poorly characterised. This study aimed to evaluate their within-day, between-day and between-month variability in spot samples, first-morning voids and 24-h collections. A total of 529 spot urine samples (including eighty-eight first-morning voids and 24-h collections) were collected from eleven Chinese adult men on days 0, 1, 2, 3, 4, 30, 60 and 90 and analysed for these five elements using inductively coupled plasma-MS. Intraclass correlation coefficients (ICC) were utilised to characterise the reproducibility, and their sensitivity and specificity were analysed to assess how well a single measurement classified individuals' 3-month average exposures. Serial measurements of Zn in spot samples exhibited fair to good reproducibility (creatinine-adjusted ICC = 0·47) over five consecutive days, which became poor when the samples were gathered months apart (creatinine-adjusted ICC = 0·33). The reproducibility of Cr, Mn, Fe and Se in spot samples was poor over periods ranging from days to months (creatinine-adjusted ICC = 0·01-0·12). Two spot samples were sufficient for classifying 60 % of the men who truly had the highest (top 33 %) 3-month average Zn concentrations; for Cr, Mn, Fe and Se, however, at least three specimens were required to achieve similar sensitivities. In conclusion, urinary Cr, Mn, Fe, Zn and Se concentrations showed a strong within-individual variability, and a single measurement is not enough to efficiently characterise individuals' long-term exposures.