Age-related variations in thyroid function hormones indicate the risk of prostate cancer: evidence from a nationally representative dataset

Abstract

Background

The risk of prostate cancer (PCa) is closely related to age and influenced by endocrine hormone imbalances. The study aimed to discuss the association between thyroid function hormone variations across different age groups and PCa as well as prostate-specific antigen (PSA) levels.

Methods

We analyzed data from two National Health and Nutrition Examination Survey (NHANES) cycles (2007–2008, 2009–2010). Serum thyroid function tests were conducted using immunoenzymatic assays. PCa was identified through self-reported diagnosis, and highly-probable PCa was identified by elevated PSA (>10 ng/mL or 4–10 ng/mL with a PSA ratio <10%). Weighted multivariable logistic regression assessed the relationship between thyroid hormone levels and the risk of PCa and highly-probable PCa.

Results

Among 2,169 participants, the prevalence of PCa was 5.35%, higher in those aged ≥65 years (12.13%) compared to those <65 years (1.52%). Elevated total triiodothyronine (TT3) and free triiodothyronine (FT3) were significantly associated with increased PCa risk in older adults [TT3: odds ratio (OR) =1.481, P<0.001; FT3: OR =1.452, P<0.001]. Among those without PCa, elevated total thyroxine (TT4) and free thyroxine (FT4) were associated with higher highly-probable PCa, with TT4 indicating early risk in individuals younger than 65 years, and FT4 in those 65 years or older. Higher TSH levels were associated with increased PSA ratio, while TT3 and FT3 showed stronger inverse age-dependent associations with total and free PSA levels. FT4 was only negatively associated with PSA ratio in older adults.

Conclusions

Thyroid function hormone levels were associated with PCa risk in older adults and also indicated an increased early risk of PCa in younger adults, as reflected by PSA markers.

Highlight box.

Key findings

• Elevated total triiodothyronine (TT3) and free triiodothyronine (FT3) increased prostate cancer (PCa) risk in older adults.

• Higher total thyroxine (TT4) and free thyroxine (FT4) suggested early highly-probable PCa, with TT4 relevant in younger and FT4 in older adults.

What is known and what is new?

• PCa risk grows with age and is typically assessed by prostate-specific antigen (PSA).

• This study newly demonstrates age-specific associations between thyroid hormones and PCa/PSA.

What is the implication, and what should change now?

• TT4 may aid early detection in younger adults, while FT4, TT3, and FT3 provide additional insights in older adults.

• Incorporating thyroid function into risk stratification could improve early detection and personalized monitoring.

Introduction

Prostate cancer (PCa) is one of the most prevalent cancers among men in the United States, with both its incidence and mortality rates on the rise in recent years. According to the most recent data from the American Cancer Society, the total number of PCa cases in the United States reached 268,490 in 2022, leading to 34,500 deaths (1). As an age-related chronic disease (2), with the aging of the population, PCa has become a significant public health concern, prompting the need to investigate its early risk prediction factors across different age groups. While previous studies have identified risk factors for PCa, such as age, race/ethnicity, and family history (3), these factors fail to fully explain the residual risk associated with PCa.

Thyroid function-related hormones including thyroid-stimulating hormone (TSH), triiodothyronine (T3), and thyroxine (T4) play a crucial role in various physiological processes, including proliferation, growth, differentiation, apoptosis, metabolism, and energy regulation (4), even the male reproductive system (5). Several studies have also indicated the significant impact of thyroid function-related hormones on cancer development, metastasis, and cancer-related mortality, such as renal and breast cancers (6,7). However, existing research on their association with PCa risk remains limited and inconclusive. For example, a systematic review and meta-analysis integrating nine cohort and case-control studies showed that although higher T4 levels were associated with tumor differentiation, no significant or definitive association could be established between thyroid diseases and the risk of developing PCa (8).

Both T3 and T4 are believed to contribute to carcinogenesis (9), stimulate tumor-induced angiogenesis (10), and even increase PCa cell proliferation in vitro (11,12). Thyroid hormones can promote prostate cell proliferation and tumor angiogenesis through nuclear receptor- or integrin-mediated signaling pathways. They may also exert a synergistic effect in the early stages of androgen-dependent PCa by regulating the expression or activity of androgen receptors (ARs) and related proteins (8,13,14). A prospective study on the relationship between iodine status, thyroid function, and the risk of PCa only revealed that men with self-reported thyroid disorder had a higher risk of developing PCa, but it did not investigate the relationship between the levels of thyroid function indicators and PCa (15). Another study linked higher T3 levels to increased pathological invasiveness in PCa, such as tumor staging and percentage of tumor involvement, while free T4 (FT4) and TSH showed no such associations (16). TSH, a key thyroid status indicator, has been linked to various pathologies like metabolic disorders, obesity, and cancer (17). A randomized controlled study suggested that men with elevated TSH levels and those categorized as hypothyroid had a reduced risk of PCa, but the study was limited to a specific population of smokers (18).

Despite these findings, there is currently a lack of comprehensive research evaluating the link between overall thyroid function indicators and PCa risk across different age groups, and the role of prostate-specific antigen (PSA) has also been overlooked. PSA testing can identify treatable early-stage PCa, significantly reducing disease-specific mortality and improving the detection of asymptomatic, well-differentiated PCa (19,20). Therefore, it is crucial to consider PSA levels when examining the relationship between thyroid function indicators and PCa. Our primary objective is to assess the association between various thyroid function indicators, PCa risk, and PSA indices, as well as early risk indicated by PSA, in a nationally representative population, and to consider the modifying effect of aging on the relationship between these factors. We present this article in accordance with the STROBE reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-411/rc).

Methods

Data and study participants

National Health and Nutrition Examination Survey (NHANES), an ongoing, cross-sectional, probability sampling, multistage survey run by the National Center for Health Statistics (NCHS), is conducted at 2-year intervals to primarily assess the health and nutritional status of the American population (21). The study obtained ethical approval from the NCHS Ethics Review Board, and written informed consent was obtained from all participants in the NHANES study. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

A total of 20,686 individuals from the NHANES 2007–2010 cycle were initially screened. Inclusion criteria encompassed male participants with complete thyroid function test data [including TSH, total T3 (TT3), free T3 (FT3), total T4 (TT4), and FT4], completed PCa questionnaire data, and available sampling weights. After applying these criteria, 2,304 individuals were identified as primarily eligible. Subsequently, individuals using medications known to affect thyroid function (e.g., levothyroxine, liothyronine, methimazole, propylthiouracil) were excluded (n=101). Individuals with those reporting a diagnosis of prostatitis were excluded (n=34). After exclusion, 2,169 participants were eligible for PCa-associated analyses, while 2,022 participants with complete PSA test data (total PSA, free PSA, and PSA ratio) were included in PSA-related analyses. The detailed population selection process is shown in Figure 1.

Figure 1.

Flowchart of study population selection. FT3, free triiodothyronine; FT4, free thyroxine; NHANES, National Health and Nutrition Examination Survey; PCa, prostate cancer; PSA, prostate-specific antigen; TSH, thyroid-stimulating hormone; TT3, total triiodothyronine; TT4, total thyroxine.

Assessment of thyroid profiles, PCa diagnosis and PSA indices

In the analysis, we incorporated serum levels of five thyroid function indicators, including TSH (µIU/mL), total T3 (TT3; ng/dL), free T3 (FT3; pg/mL), total T4 (TT4; µg/dL), FT4 (ng/dL). The serum TSH were assessed using the HYPERsensitive human TSH (hTSH) assay, which is a 3rd generation, two-site immunoenzymatic (“sandwich”) assay. Competitive binding immunoenzymatic assays were utilized for TT3, FT3, and TT4, while a two-step enzyme immunoassay was employed for FT4 analysis.

PCa was identified by self-reported previous diagnosis based on the answer of “ever been told by a doctor or health professional that you had PCa?” derived from questionnaire items KIQ201 in the PSA detection section of the Laboratory Data.

It is important to acknowledge that self-reported data may be affected by recall bias, potentially influencing the data analysis and interpretation. Here, we also introduced PSA assessment to identify individuals at high risk of developing PCa. PSA testing is applicable to male participants aged 40 years and older, provided they do not have any of the following conditions: current infection or inflammation of the prostate gland, rectal exam in the past week, prostate biopsy in the past month, cystoscopy in the past month, or a history of PCa. The serum-free PSA concentration was detected using The Access Hybritech free PSA assay, which is a two-site immunoenzymatic “sandwich” assay and the serum total PSA concentration were obtained using the Hybritech PSA method on the Beckman Access. The percent free PSA compared to total PSA enhances the specificity in predicting disease (22). Therefore, the PSA ratio was calculated by dividing the free PSA by the total PSA. For participants without PCa, the levels of various PSA indices were comprehensively considered, defining those with total PSA levels exceeding 10 ng/mL or ranging between 4–10 ng/mL with a PSA ratio of less than 10% as highly-probable PCa.

Statistical analysis

Given that the NHANES survey employed a complex, multistage probability sampling design, sample weights corresponding to each survey period were incorporated to generate nationally representative estimates. In the description of baseline characteristics, continuous variables were presented as medians with interquartile ranges (IQRs) and compared using the weighted Wilcoxon rank-sum test, while categorical variables were expressed as weighted percentages and compared using the weighted Chi-squared test.

Furthermore, after adjusting for covariates, including age, high body mass index (BMI), smoking, and alcohol consumption, history of diabetes and hypertension (well-established risk factors for PCa), the associations between thyroid function indicators and PCa risk, as well as PSA indices, were assessed using weighted multivariable logistic regression and weighted linear regression models, respectively. BMI was calculated as weight in kilograms divided by height in centimeters squared. Triglyceride levels were measured using a timed-endpoint method included in the Standard Biochemistry Profile. Smoking status was categorized as 0, 1, and 2, corresponding to “Not at all”, “Some days”, and “Every day”, respectively. Individuals who had consumed at least 12 alcoholic beverages in the past year were categorized as alcohol drinkers (assigned a value of 1), while the rest were considered non-drinkers (assigned a value of 0).

To explore potential age-related differences in the impact of thyroid hormone levels on PCa risk, the study population was further divided into younger (<65 years) and older (≥65 years) groups for subgroup analyses. Results were reported as odds ratios (ORs) and beta coefficients, along with their 95% confidence intervals (CIs) and P values. A two-sided P value <0.05 was considered statistically significant. All statistical analyses were performed using R software version 4.3.3.

Results

Baseline characteristics of the study participants

A total of 2,169 participants were included in this study, of whom 783 (36.1%) were aged 65 years and above. The unweighted prevalence of PCa was 5.35% (n=116), with a marked age-related disparity: only 1.52% of participants under 65 years had PCa, compared to 12.13% in those aged ≥65 years (P<0.001, Table 1). After applying sampling weights, the estimated population-level prevalence of PCa was 3.18% overall, 0.85% in the younger group, and 10.64% in the older group (Table S1). Older adults exhibited significantly higher levels of TSH (1.82 vs. 1.57 µIU/mL, P<0.001) and lower levels of both TT3 (TT3: 102 vs. 113 ng/dL, P<0.001) and FT3 (3.00 vs. 3.20 pg/mL, P<0.001). FT4 75^th percentile levels were slightly but significantly higher in the elderly (0.90 vs. 0.80 ng/dL, P<0.001), whereas TT4 showed no significant difference (P=0.11, Table 1). Older individuals had lower levels of low-density lipoprotein cholesterol (LDL-C) and total cholesterol but higher fasting blood glucose, blood urea nitrogen, serum creatinine levels (all P<0.001) and they were less likely to be current smokers (9.83% vs. 24.46%) (all P<0.001).

Table 1. Baseline characteristics of the study participants.

Variable	Overall (n=2,169)	Age <65 years (n=1,386)	Age ≥65 years (n=783)	P value
Age, years	60.00 (50.00, 70.00)	52.00 (46.00, 59.00)	73.00 (69.00, 79.00)	<0.001
BMI, kg/m2	28.27 (25.33, 31.70)	28.42 (25.40, 32.03)	28.00 (25.26, 31.28)	0.057
TG, mmol/L	1.39 (0.98, 2.01)	1.42 (1.02, 2.09)	1.35 (0.95, 1.94)	0.13
LDL-C, mmol/L	2.97 (2.38, 3.60)	3.10 (2.56, 3.72)	2.77 (2.12, 3.28)	<0.001
TC, mmol/L	4.99 (4.29, 5.77)	5.17 (4.55, 5.95)	4.65 (4.03, 5.35)	<0.001
HDL-C, mmol/L	1.16 (0.98, 1.42)	1.16 (0.98, 1.40)	1.22 (1.01, 1.45)	0.006
FBG, mmol/L	5.88 (5.44, 6.66)	5.83 (5.38, 6.44)	6.16 (5.61, 6.94)	<0.001
ALT, U/L	24.00 (19.00, 32.00)	27.00 (21.00, 36.00)	21.00 (17.00, 26.00)	<0.001
AST, U/L	25.00 (22.00, 30.00)	26.00 (22.00, 31.00)	24.00 (21.00, 28.00)	<0.001
BUN, mmol/L	5.00 (3.93, 6.07)	4.64 (3.57, 5.71)	5.71 (4.64, 7.50)	<0.001
Creatinine, μmol/L	86.63 (74.26, 99.01)	81.33 (72.49, 93.70)	91.94 (81.33, 107.85)	<0.001
History of diabetes	358 (16.51)	188 (13.56)	170 (21.71)	<0.001
History of hypertension	955 (44.03)	513 (37.01)	442 (56.45)	<0.001
Smoking	416 (19.18)	339 (24.46)	77 (9.83)	<0.001
Drinking	1,714 (79.02)	1,114 (80.38)	600 (76.63)	0.045
PCa	116 (5.35)	21 (1.52)	95 (12.13)	<0.001
Thyroid function profiles
TSH, μIU/mL	1.67 (1.11, 2.44)	1.57 (1.06, 2.34)	1.82 (1.22, 2.67)	<0.001
TT4, μg/dL	7.50 (6.70, 8.50)	7.50 (6.70, 8.50)	7.60 (6.70, 8.60)	0.11
FT4, ng/dL	0.80 (0.70, 0.87)	0.80 (0.70, 0.80)	0.80 (0.70, 0.90)	<0.001
TT3, ng/dL	110.00 (96.00, 123.00)	113.00 (101.00, 128.00)	102.00 (89.00, 115.00)	<0.001
FT3, pg/mL	3.10 (2.90, 3.40)	3.20 (3.00, 3.50)	3.00 (2.80, 3.20)	<0.001

Continuous variables were presented as median (interquartile range) and categorical variables were expressed as count (weighted %). ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BUN, blood urea nitrogen; FBG, fasting blood glucose; FT3, free triiodothyronine; FT4, free thyroxine; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; PCa, prostate cancer; TC, total cholesterol; TG, triglyceride; TSH, thyroid-stimulating hormone; TT3, total triiodothyronine; TT4, total thyroxine.

Association of thyroid function indicators with risk of PCa and highly-probable PCa

Weighted multivariable logistic regression revealed that among the total population, higher levels of TT3 were significantly associated with an increased risk of PCa (OR =1.312, 95% CI: 1.017–1.694, P=0.04) (Table 2). In stratified analysis by age, the association between TT3 and PCa risk was more pronounced in participants aged ≥65 years (OR =1.481, 95% CI: 1.238–1.772, P<0.001), along with a strong positive association observed for FT3 in the older group (OR =1.452, 95% CI: 1.218–1.731, P<0.001) (Table 2). Although TT4 and FT4 were not significantly associated with overall PCa risk, FT4 was significantly associated with highly-probable PCa (OR =1.428, 95% CI: 1.137–1.794, P=0.004), with the association remaining significant in participants aged ≥65 years (OR =1.485, 95% CI: 1.064–2.074, P=0.02) (Table 2). Moreover, TT4 showed a significant association with highly-probable PCa in the subgroup aged <65 years (OR =1.771, 95% CI: 1.144–2.741, P=0.01) (Table 2). No statistically significant associations were observed for TSH with either PCa risk or highly-associated PCa in both age groups. These findings underscore age-dependent heterogeneity in the relationship between thyroid function and PCa development, with stronger associations observed in the elderly.

Table 2. Weighted multivariable logistic regression association of thyroid function indicators with risk of PCa and highly-associated PCa defined by PSA indices.

Population/thyroid function indicators	PCa risk			Highly-probable PCa
	OR (95% CI)	P value		OR (95% CI)	P value
Total population
TSH	1.042 (0.800–1.357)	0.75		0.684 (0.419–1.115)	0.12
TT3	1.312 (1.017–1.694)	0.04		1.203 (0.750–1.928)	0.43
FT3	1.276 (0.992–1.641)	0.06		1.012 (0.637–1.609)	0.96
TT4	1.028 (0.805–1.313)	0.82		1.470 (1.098–1.967)	0.01
FT4	1.131 (0.959–1.333)	0.14		1.428 (1.137–1.794)	0.004
Age <65 years
TSH	0.869 (0.415–1.819)	0.70		0.420 (0.156–1.128)	0.08
TT3	0.649 (0.307–1.375)	0.25		1.206 (0.740–1.967)	0.44
FT3	0.590 (0.237–1.468)	0.24		1.141 (0.748–1.742)	0.53
TT4	0.982 (0.580–1.664)	0.95		1.771 (1.144–2.741)	0.01
FT4	1.024 (0.613–1.712)	0.92		1.178 (0.765–1.813)	0.44
Age ≥65 years
TSH	1.072 (0.794–1.447)	0.64		0.907 (0.597–1.378)	0.63
TT3	1.481 (1.238–1.772)	<0.001		1.260 (0.654–2.426)	0.47
FT3	1.452 (1.218–1.731)	<0.001		0.978 (0.531–1.802)	0.94
TT4	1.026 (0.771–1.366)	0.86		1.244 (0.812–1.908)	0.30
FT4	1.174 (0.980–1.406)	0.08		1.485 (1.064–2.074)	0.02

Models were adjusted for age, BMI, triglycerides, diabetes, smoking status, alcohol consumption, history of diabetes and hypertension. Thyroid function indicators levels were standardized. Highly-probable PCa was identified by elevated PSA (>10 or 4–10 ng/mL with a PSA ratio <10%). BMI, body mass index; CI, confidence interval; FT3, free triiodothyronine; FT4, free thyroxine; OR, odds ratio; PCa, prostate cancer; PSA, prostate-specific antigen; TSH, thyroid-stimulating hormone; TT3, total triiodothyronine; TT4, total thyroxine.

Association of thyroid function indicators with PSA indices

In the total population, higher TSH levels were significantly associated with lower total PSA (β=−0.048, 95% CI: −0.089 to −0.007, P=0.02), while a positive association was observed with PSA ratio (β=0.139, 95% CI: 0.075 to 0.203, P<0.001) (Figure 2). TT3 showed a modest inverse association with free PSA (β=−0.053, P=0.04), and this inverse relationship with total PSA became more pronounced among younger participants (<65 years). The positive association between TSH and PSA ratio remained consistent across age groups. Among older adults (≥65 years), FT3 exhibited significant inverse associations with both total PSA and free PSA (β=−0.088, P=0.03; β=−0.142, P=0.02, respectively). Moreover, FT4 was significantly negatively associated with PSA ratio in the older population (β=−0.188, 95% CI: −0.286 to −0.090, P<0.001).

Figure 2.

Weighted multivariable linear regression association of thyroid function indicators with PSA indices. Models were adjusted for age, BMI, smoking status, alcohol consumption, history of diabetes and hypertension. Thyroid function indicators levels were standardized. BMI, body mass index; CI, confidence interval; FT3, free triiodothyronine; FT4, free thyroxine; PSA, prostate-specific antigen; TSH, thyroid-stimulating hormone; TT3, total triiodothyronine; TT4, total thyroxine.

Discussion

In this study, we comprehensively investigated the association of various thyroid function indicators with PCa risk and various PSA indices. We found that in the elderly population, elevated levels of TT3 and FT3 were independent risk factors for PCa. Even among individuals without PCa, elevated TT4 and FT4 levels could indicate early risk for PCa, as evidenced by a significant association with the risk of highly-probable PCa defined by variations in PSA indices.

In our study, it was observed that the prevalence of PCa is significantly higher in the elderly population compared to younger individuals. Previous studies have shown that PCa is highly associated with aging, with a markedly increased prevalence in older adults (2). PCa, being an age-related tumor, demonstrated a doubling trend in incidence among individuals aged 70 years and above in Asian countries (23). Similarly, in the United States, the high-risk population was predominantly found in the age group of 50 years and above, particularly in the 70 to 74 years age range (24). A recent nationwide study based on the Norwegian Cancer Registry reported that with increasing age, both the proportion and absolute incidence of clinically significant PCa are on the rise, and as age advances, the disease grade also demonstrates a gradual increase (25). One possible biological explanation for the increased risk of PCa with age could be the alteration in testosterone balance as individuals age (26). Meanwhile, thyroid hormone levels also undergo stable changes with age (27). Therefore, age is a potential factor influencing the association between thyroid function and PCa. More importantly, our findings reveal age-specific associations between different thyroid hormones and PCa risk, which holds significant implications for the precise prediction of PCa. Therefore, age subgroup analyses were conducted in subsequent analyses to consider the stratification effect of age. Our results not only show that elevated levels of thyroid hormones TT3 and FT3 in high-risk elderly populations can indicate an increased risk of PCa, but also that even in younger populations, elevated TT4 levels can similarly suggest an increased early risk of PCa as defined by PSA variations. These findings provide evidence-based medical support for monitoring PCa risk across different age groups. In the study by Díez et al., the incidence of PCa was significantly reduced among hypothyroid patients over 60 years old (28). Although our study did not involve thyroid diseases, the trend was similar to our findings, as we observed that elevated TT3 and FT3 were significantly associated with increased PCa risk, whereas low TT3 and FT3 (possibly indicative of hypothyroidism) were significantly associated with reduced PCa risk. Similarly, other studies have suggested that lower TSH and higher FT4 are associated with an increased risk of PCa (18,29). A systematic review and meta-analysis combining nine studies found no significant association between thyroid diseases and the risk of developing PCa (8), which is not entirely consistent with our findings. This discrepancy may be attributed to substantial differences in the racial composition of the two study populations. Previous studies have reported significant differences between White and Black populations in both thyroid function levels and PCa risk (30,31). Moreover, our research focuses on thyroid function–related hormone levels, which are more likely to capture potential associations with PCa compared with clinical hyperthyroidism and hypothyroidism, which have relatively low prevalence in the general population. This is because previous studies have shown that even within the normal range of thyroid function, variations in thyroid hormone levels are associated with many diseases (32-34). In addition, our study used PSA-defined early PCa risk as the outcome, which better reflects the association between thyroid function and the preclinical early risk of PCa, offering greater clinical warning value.

Most epidemiological data link low T4 and high TSH levels (defined as hypothyroidism) to a reduced risk of developing PCa (18,29), which is consistent with our findings. Our study results show that high TSH levels are associated with a higher PSA ratio, while a low PSA ratio is considered a high-risk indicator for PCa. Currently, epidemiological data on the effects of high T3 levels on PCa are limited, with most studies focusing on in vitro experiments or mouse models, yielding inconsistent results. Previous studies have reported that thyroid hormones (T3/T4) can bind to nuclear receptors (such as TRα1) to regulate the expression of cell cycle–related genes (such as cyclin D1), thereby promoting the proliferation of prostate epithelial cells. In addition, thyroid hormones can activate ERK1/2 via integrin αvβ3-mediated non-genomic signaling pathways, inducing VEGF expression and promoting tumor angiogenesis. Moreover, the occurrence and progression of PCa are androgen-dependent. Some studies suggest that thyroid hormones may indirectly influence the development of PCa by regulating the expression level or transcriptional activity of the AR and the expression of androgen transport-related proteins. For example, in vitro studies have shown that T3 can upregulate AR expression in lymph node carcinoma of the prostate (LNCaP) cells, suggesting a potential synergistic role in the early stages of androgen-dependent PCa (8,13,14). Furthermore, interventional studies have shown that T3 treatment can slow the progression of LNCaP tumors in a mouse xenograft model (35), but the underlying mechanism remains unclear. A recent in vitro study found that T3 actively regulates endoplasmic reticulum-associated degradation (ERAD) and promotes androgen signaling in androgen-dependent PCa cells, linking it to the risk of PCa development (36). Although the role of high T3 levels in PCa remains contradictory, most studies indicate an association between high T4 levels and the occurrence and development of PCa (37,38). This aligns with the conclusions drawn from our study. One well-known biological mechanism that may occur is the binding of T4 and T3 to the membrane receptor integrin avb3, activating various oncogenic pathways, including PI3K and MAPK/ERK1/2, and promoting cell proliferation and angiogenesis (10). Future research should combine transcriptomics, single-cell sequencing, or animal models to further explore the regulatory mechanisms of thyroid hormones on the androgen pathway and their differential roles at various stages of PCa.

Of particular significance, our study identified a positive correlation between elevated TT4 or FT4 levels and a higher risk of highly-probable PCa, as defined by PSA indices, in individuals without a history of PCa. This was especially notable in the younger group, where elevated TT4 levels could more effectively indicate an increased early risk of PCa. PSA testing is becoming more prevalent, especially in men over 60 years of age. The European Randomized Study of Screening for PCa reported that regular PSA testing every 2 to 4 years can reduce PCa mortality rates (39). A secondary analysis of a randomized clinical trial in England and Wales also found that a single PSA screening reduced PCa mortality during a 15-year follow-up period compared to standard practice without routine screening (40). Regarding the impact of thyroid function indicators on PSA levels, early in vitro experiments indicated that T3 regulates the expression of the PSA gene at the transcription level, leading to increased expression of androgen-dependent PSA (41). However, current epidemiological studies on the relationship between T3 and T4 levels and PSA levels have not been conclusive. A study conducted at the Semmelweis University Laboratory Medicine Institute found a significant negative correlation between TSH and PSA levels in men aged 40–75 years, but the specific mechanism remains unclear (42). Conversely, the Rotterdam Study and the Health in Men Study did not observe any associations between TSH and PCa (43). The inconsistency in results across different cohorts remains unexplained. Our study found a positive correlation between TSH levels and PSA ratio levels in the younger age group. As the most sensitive indicator of thyroid function, TSH exhibits significant variability, which may serve as an earlier warning sign of PCa risk in young individuals. This has important implications for the early detection and prevention of PCa in younger populations.

Our study suggests that monitoring thyroid function indicators may help identify individuals at higher risk of PCa across different age groups. Our study has several strengths. Firstly, the database involved a nationally representative sample and applied survey weights to ensure the generalizability of our findings to adult men in the USA. Secondly, we included PSA indices to assess the correlation between thyroid function indicators and PCa early risk, a component that is lacking in most studies. Third, a relatively detailed analysis was conducted across different age groups to explore the differential associations between thyroid hormones and PCa. Nevertheless, our study also has some limitations. First of all, as a cross-sectional analysis, the ability to establish a definitive causal relationship is inherently limited. Secondly, we utilized self-reported physician-diagnosed cases of PCa to identify individuals with PCa, which may underestimate the actual prevalence of PCa and introduce potential recall bias. Therefore, caution is advised when analyzing and interpreting the data. Objective imaging or pathological examinations would be more accurate assessment methods. Due to the lack of such objective examinations in NHANES, our study compensated for this limitation by additionally including objective PSA serological test data, aiming to capture earlier changes in PCa through variations in different PSA indicators. Moreover, the subpopulation undergoing PSA testing was randomly selected from individuals aged 40 years and older, making the likelihood of bias caused by individual subjective intention or healthcare accessibility very small, thus ensuring general representativeness for the middle-aged and elderly population. We acknowledge that including participants under 40 years could provide additional clinical evidence and increase the generalizability of the conclusions. Finally, due to limitations in the NHANES database, we could not access clinical data on the grade, stage, and treatment of PCa, hindering the stratified assessment of potential variations in the association between thyroid function indicators and the risk or progression of PCa. Whether our study’s conclusions can be further validated in terms of PCa metastasis, aggressiveness, or longer-term clinical outcomes still requires further investigation. In the future, prospective randomized controlled trials are essential to provide stronger evidence for our findings.

Conclusions

In summary, our study identified age-specific variations in thyroid function indicators that may suggest an increased risk of PCa. We not only identified T3-related PCa risk in the elderly population but also detected early PSA-indicated PCa risk in younger individuals. These findings offer a new perspective for the timely identification of high-risk PCa populations during the aging process. Our study supports monitoring changes in thyroid hormone levels to inform early PSA screening and reduce the risk of PCa.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.