Incarcerated individuals represent a marginalized and vulnerable population, yet incarceration can also represent an opportunity to reduce health inequalities by providing consistent medical support to a population that often lacks access to care in the community [1–9], as well as opportunities to promote health through preventive measures such as more balanced nutrition and access to physical activity [10, 11].

Although institutional meals are typically provided in prisons, the actual nutritional intake can vary widely. Many incarcerated people rely on supplementary foods from canteens or external sources, often choosing items that are energy-dense but nutritionally poor [5, 12, 13]. This pattern can contribute to both weight gain and nutritional imbalances, increasing the risk of chronic diseases like obesity, diabetes, and cardiovascular disease. Weight gain is further compounded by the frequent use of psychotropic medications, particularly antipsychotics, which are known to induce metabolic changes and increase appetite [14]. Additionally, the social and psychological dynamics of prison life can significantly shape eating habits, with food serving as a source of comfort and a form of resistance to institutional control [15].

Given these factors, it remains unclear to what extent the dietary patterns observed in prisons influence the nutritional status of incarcerated persons, particularly regarding essential trace elements like zinc, selenium, and copper. These micronutrients are critical for immune function, oxidative stress control, and metabolic regulation, making their assessment particularly relevant for this population [16–19]. While national programs, coordinated by the Federal Food Safety and Veterinary Office, monitor these trace elements in the general population [20–22], comparable data for people in detention are lacking, underscoring the importance for such evaluation.

Canton de Vaud has six custodial facilities: four adult prisons (three for men, one for women, covering remand and sentence execution), one for minors/young adults, and one semi-detention establishment: three sites were included in the present study. In November 2023, Vaud accounted for 10.7% of Switzerland’s total detention population, with 110 per 100 000 inhabitants, above the national average (∼76/100 000) [23] and near the European median (∼105/100 000) [24]. Like most European countries, Switzerland, and Vaud in particular, has a predominantly male prison population, with men accounting for 94.1% of this population [25]. Vaud faces chronic overcrowding with 117% occupancy in 2023 versus 94% nationally [23]. This high-volume and high-occupancy context provides a pertinent setting for interpreting health and nutrition indicators in incarcerated population.

During the COVID-19 pandemic, Vaud cantonal health authorities launched the “SérocoVID” project to estimate prior SARS-CoV-2 infection, assess immunity, and identify transmission factors within its population. As part of this effort, three prison facilities located in the canton of Vaud (French-speaking Switzerland) were included due to their outbreak vulnerability, providing a unique opportunity to collect broader health data from incarcerated individuals, including plasma levels of zinc, selenium, and copper, beyond the study’s primary focus on viral transmission and immunity.

The primary aim of this study is to describe plasma zinc, selenium, and copper levels among incarcerated men in Canton of Vaud. We then compare these findings with data from men in the general Swiss adult population of the same region to identify potential differences and disparities. We also explore whether selected sociodemographic and behavioral factors are associated with plasma trace element levels among incarcerated individuals. Describing plasma trace element concentrations and their distribution in this population may help inform future monitoring and generate hypotheses for further research on nutritional and health determinants in prison settings.

The study included 206 men with a mean age of 37.8 years (SD = 12.8). Among them, 61.0% (114/206) reported a pre-incarceration household income below 3,000 CHF, and 51.2% (104/203) had no education beyond compulsory schooling. A total of 72.1% (142/197) were of foreign origin, representing 50 different nationalities. These figures are consistent with national prison statistics from the FSO, particularly regarding age and the proportion of foreign nationals. Additional details are presented in Supplementary Table S1.

Plasma concentrations of selenium, copper and zinc among participants had median values of 106 ng/mL for Se, 911 ng/mL for Cu and 824 ng/mL for Zn (see Table 1). Substantial interindividual variability was observed, particularly for zinc and copper, with wide dispersion between the 5th and 95th percentiles (see Supplementary Table S2 for full distribution).

To place these values in context, we compared them to data from the NutriSerocovid study. As shown in Figure 1, the zinc distribution among incarcerated men appears shifted to the right and more dispersed compared to the general population. For copper and selenium, distributions were more similar, with overlapping medians and less pronounced differences in shape.

CRP levels, measured for all participants, were low and normal, except for one individual with a value of 20 mg/L: the 75th percentile was 2.33 mg/L. Given this limited variability, no adjustment for CRP was performed in the analysis.

Comparison of plasma concentrations between incarcerated men and community men revealed nuanced differences depending on the element and the adjustment factors applied (see Table 1). For selenium, median levels were nearly identical between groups (106 ng/mL in prison vs. 105 ng/mL in the general population), and neither unadjusted nor adjusted comparisons of medians showed significant differences. However, after adjusting for age and BMI, mean selenium levels were significantly lower among incarcerated men (adjusted mean difference: −4.66 ng/mL; 95% CI: −8.97 to −0.35; p = 0.03), a contrast not apparent in the unadjusted comparison (p = 0.77), suggesting a confounding effect of age.

For copper, both unadjusted and adjusted median and mean levels were slightly higher in the incarcerated population (median: 911 vs. 905 ng/mL; mean: 916 vs. 907 ng/mL), but the differences were not statistically significant until adjusting for age and BMI (see Table 1). After adjustment, mean copper levels were significantly higher among incarcerated individuals (adjusted mean difference: +51.8 ng/mL; 95% CI: 8.78 to 94.8; p = 0.02), again underscoring the influence of age on group comparisons.

For zinc, the difference was more pronounced. Incarcerated men had significantly higher median (824 vs. 749 ng/mL, p < 0.0001) and mean levels (830 vs. 750 ng/mL, p < 0.0001) compared to the general population (see Table 1). However, this difference was attenuated after adjustment for age and BMI, and only the adjusted mean remained statistically significant (adjusted mean difference: +33.1 ng/mL; 95% CI: 1.31 to 65.0; p = 0.04), while the adjusted median difference lost significance (p = 0.18).

These discrepancies between mean and median findings are visually supported by Figure 1, which shows a broader and right-skewed distribution of zinc values in the incarcerated population. Such distributional asymmetries likely contribute to the divergence in statistical significance between parametric and non-parametric tests. For copper and selenium, distributions appear more symmetric and overlapping, consistent with the less pronounced group differences observed.

The prevalence of concentrations below the Swiss lower reference limits did not differ significantly between groups for any element.

Using alternative cut-offs (700 ng/mL), the proportion of men with low plasma zinc was significantly higher in the community (17.3% vs. 32.3%; p < 0.001). For copper (636 ng/mL), low plasma copper was more frequent among detained men (7.65% vs. 1.59%; p = 0.01). For selenium (60 ng/mL), the proportion below the cut-off was low in both groups (2.55% vs. 0%; p = 0.08). Details are presented in Supplementary Table S3.

We also examined differences in sociodemographic characteristics and micronutrient levels across the three detention facilities (Table 2). Men in Prison 3 were significantly older and had a higher BMI compared to those in Prison 1 (p < 0.001 and p = 0.040, respectively), while no significant differences were observed between Prisons 1 and 2. The proportion of foreign-origin individuals was highest in Prison 2, which differed significantly from both other sites (p = 0.019 vs. Prison 1; p = 0.007 vs. Prison 3).

Micronutrient concentrations also varied across facilities (Table 2). Zinc levels were significantly lower in Prison 2 compared to both Prisons 1 and 3 (p = 0.038 and p = 0.004, respectively). Conversely, copper levels were significantly higher in Prison 3 than in Prisons 1 and 2 (p = 0.045 and p = 0.021). No significant differences were found for selenium.

Finally, to identify factors associated with micronutrient status, we conducted univariable and multivariable regression analyses. Tables 3–5 summarize associations between selenium, copper, and zinc levels and variables such as age, education, income, BMI, self-rated health, smoking, and prison site.

For selenium, lower levels were associated with older age (β = −0.29, p = 0.02), low income (β = −7.65, p = 0.02), and current smoking (β = −11.1, p = 0.02). A borderline association was also observed for education (β = 5.86, p = 0.05). Selenium levels initially appeared higher in Prison 2, but this difference was not significant after adjustment for confounders.

For copper, no strong associations were identified. Although men in Prison 3 had higher copper levels in unadjusted models (β = 81.1, p = 0.02), this association did not remain significant after adjustment.

Zinc concentrations were significantly associated with age and prison site. Older men had lower zinc levels (β = −3.30, p = 0.002), and individuals in Prison 2 had significantly lower levels than those in Prison 1 (β = −66.9, p = 0.03). No other variables showed significant associations.

This study provides the first descriptive assessment of plasma zinc, copper, and selenium concentrations among incarcerated men in Switzerland. Overall, few men had plasma concentrations below reference limits and proportions did not differ significantly from those observed in men from the general population.

After adjustment for age and BMI, subtle differences emerged. Selenium levels were slightly lower, while zinc and copper were slightly higher among incarcerated men. These shifts were not significant in median-based comparisons, indicating distributional differences rather than consistent population-wide disparities. Zinc showed a more dispersed, right-skewed distribution, with greater interindividual variability, a pattern, confirmed by kernel density plots and Wilcoxon rank-sum test, suggesting heterogeneity within the study population.

Sensitivity analyses using alternative lower cut-offs made some between-group differences statistically significant, consistent with the distributional patterns described above. Selenium values below the cut-off were rare in both groups, copper was more often below the cut-off among detained men, and zinc below the cut-off increased in both groups but remained more prevalent among community men. These unadjusted findings should be interpreted cautiously, as they illustrate how the prevalence of values below selected thresholds depends on the cut-offs applied and highlight the need for clearer, outcome-oriented reference ranges.

Plasma selenium is considered as a suitable biomarker for ranking subjects according to long-term selenium intake [42]. Its informative value as a biomarker depends on overall selenium status but additional, yet costly, biomarkers such as selenoprotein would be needed for a comprehensive understanding of selenium status [43]. Lower selenium in the prison sample was associated with older age and smoking, consistent with community-based studies [44–49]. Lower socio-economic position, measured via pre-incarceration income, was also linked to lower selenium concentrations. However, we found no prior studies directly relating plasma selenium to income. Socioeconomic differences in diet quality are well documented and may contribute to differences in selenium intake. [50–53]. Slightly lower mean selenium concentrations compared with the general population may plausibly reflect the much higher smoking prevalence in prisons (76.8% current smoker in our sample), given its known negative association with selenium status [44–47, 49]. Smoking has also been linked to broader alterations in micronutrient status [49, 54–57].

In addition, dietary patterns and environmental factors, including the selenium content of food sources, influence selenium exposure. These factors may contribute to variability in selenium concentrations across populations. The absence of dietary intake data, detailed health indicators, and institutional dietary assessments limit the interpretation of observed micronutrient concentrations. In addition, plasma concentrations of trace elements, particularly zinc and copper, are influenced by physiological and inflammatory processes and therefore may not accurately reflect individual nutritional status; however, inflammation was not present in this case (median CRP = 2.3 mg/L).

There is currently no specific, reliable biomarker that indexes habitual zinc intake or total body zinc; marginal deficiency at the individual level is difficult to detect [35, 58]. Plasma zinc, the most used indicator, reflects only a small mobilizable pool and is highly sensitive to non-nutritional factors (infection, stress, inflammation) as well as fasting status and sampling time, with post-prandial decreases of up to ∼19% [31, 59–61]. Because fasting status and draw time were not recorded (CRP was generally low), some measurement variability cannot be excluded.

In our sample, zinc declined with age and was significantly lower in Prison 2. This facility primarily houses men in pretrial detention (a smaller proportion serving sentences), a stricter regime, and has been characterized by chronic overcrowding and suboptimal infrastructure, conditions likely to heighten stress. Shorter stays also mean the effects of pre-incarceration vulnerabilities (e.g., poor diet, substance use) may still be present. These factors should be interpreted cautiously, as similar constraints affect individuals detained in Prison 1. Dietary differences could also contribute: Prisons 1 and 3 share a centrally prepared menu, unlike Prison 2, which operates its own. Together, these contextual features may help explain lower zinc measured in Prison 2. It is possible that unmeasured factors, such as incarceration duration, stress and canteen use, explain part of the observed differences between prisons.

Plasma copper was not associated with available covariates in our analyses. Slightly higher mean levels among incarcerated men, but still within normal ranges, can be considered as non-clinically relevant and may reflect behavioral or health factors. Smoking, strongly prevalent in prisons, has been linked to higher plasma copper through increased ceruloplasmin synthesis [49, 57], although in our data smoking was not a statistically significant predictor. Higher values are associated with cardiovascular disease [62], but this study is not adequately powered to explore such link. Other unmeasured factors such as inflammation (which was not present in the study) [37] or liver disease [63–65] also contribute, and the clinical relevance of this finding remains uncertain.