Lifestyle in flux: urbanization, dietary shifts, and endocrine health in emerging adulthood

Abstract

Emerging evidence highlights the decline of testosterone levels among young males, linked to modern lifestyle shifts rather than aging alone. This exploratory cross-sectional study investigates the interplay between modifiable lifestyle factors and testosterone levels in 50 males aged 18–22 years, focusing on underrepresented variables such as exercise type, carbonated beverage intake, and sunlight exposure. Serum testosterone levels were measured via chemiluminescent immunoassay, and lifestyle data were collected through previously validated questionnaires. Multiple regression analyses revealed hypertrophy training (β = 20.3, p < 0.001), sunlight exposure > 60 min (β = 10.3, p = 0.03), and supplement use (β = 20.5, p < 0.001) as positive predictors of testosterone. Conversely, daily carbonated beverage consumption (β=−10.2, p = 0.01), tobacco use (β=−15.6, p < 0.001), and sleep deprivation (β=−18.2, p < 0.001) were significant negative correlates. Diet type influenced outcomes, with non-vegetarians showing higher testosterone (β = 8.7, p = 0.03) compared to vegetarians. Notably, BMI and chronic diseases were nonsignificant in this young cohort. These findings underscore the multifactorial nature of testosterone regulation, emphasizing holistic lifestyle interventions—such as resistance training, reduced ultra-processed food intake, and sleep optimization—as critical for endocrine health in urbanized youth. The study challenges traditional obesity-centric frameworks, advocating for holistic approaches to mitigate endocrine disruption in emerging adulthood.

Introduction

Testosterone, a hormone of primary importance in male physiology, extends beyond reproduction to impact metabolism [1–3], musculoskeletal health, and cardiovascular function [2, 4–6]. Previously linked in a more classic conception to sex development and fertility. New evidence reveals its function in modulating systemic health, such as insulin sensitivity, lipid metabolism, and cognitive function [7, 8]. Recent patterns of a decline in younger men’s testosterone challenge classic conceptions of age-associated hypogonadism in older men [9–11]. The decline coincides with patterns of contemporary lifestyles of inactivity [12–16], diet of refined foods [14, 17–19], and exposure to chemicals, suggesting multifactorial etiologic processes beyond aging per se [20–22].

Recent evidence points to paradoxical associations between testosterone and aspects of contemporary lifestyles such as diet, physical inactivity, and illicit drugs [23–26]. As a classic correlate of hypogonadism, obesity is a correlate of hypogonadism. Yet studies increasingly suggest poor micronutrient-dense diet, high in refined sugars and trans fat, in impairing endocrine homeostasis [27, 28]. Concurrently, exposure to pollutants such as microplastics has been found to be toxic to Leydig cells in preclinical models, raising concern regarding their potential to impact decreasing androgen levels [29–33]. In contrast, resistance exercise and controlled use of caffeine promise to augment testosterone synthesis, though mechanisms of action remain in part unexplained [34, 35, pp. 2013]–2014 [36, 37], .

In spite of these advances, primary gaps in knowledge regarding intersecting variables of lifestyles in modulating testosterone in young men continue to prevail. Prior studies often isolate a lone variable, excluding synergism or antagonism between variables that characterizes actual exposure in nature [38–42]. In addition, dietary patterns in various geographic areas, such as fat in Indian meals, provide unique variables that differentially impact androgen dynamics [32, 43].

This preliminary study attempts to bridge these gaps by examining combined effects of modifiable lifestyle factors on testosterone in a young adult population, prioritizing variables that are underrepresented in current literature, i.e., carbonated drink use, exposure to sunlight, and work-related physical demands. By pairing clinical biomarkers with behavioral variables, this work attempts to guide evidence-based interventions to counteract endocrine disruption in increasingly urbanized groups.

Methods

Participant selection

This exploratory cross-sectional study included a convenience sample of 50 male participants aged 18–22 years from Aurangabad region, India with a stable lifestyle and dietary habits. Participants were eligible if they had no self-reported history of endocrine disorders, cardiovascular disease, or use of testosterone-altering medications/supplements (other than general multivitamins). A previously validated lifestyle questionnaire was used, with minor pilot testing for clarity on a subsample (n = 10). All participants provided informed consent and completed a structured questionnaire assessing lifestyle factors, dietary habits, and exercise routines. Blood samples were collected to measure testosterone levels, and anthropometric data (e.g., height, weight, BMI) were recorded. The study protocol was approved by MGM- Ethical Committee for Research on Human Subject (MGM- ECRHS), Aurangabad (Letter no. MGM-ECRHS/2023/145, dated 6/10/2023).

Study procedures

Participants underwent standardized measurements for height (cm), weight (kg), and BMI (kg/m²). Testosterone levels were quantified from serum samples using chemiluminescent immunoassay (CLIA) and reported in ng/dL. Lifestyle and health behavior data, including diet type, exercise frequency, sleep duration, and substance use (e.g., tobacco, caffeine), were collected via validated self-reported questionnaires. Exercise intensity and type (e.g., hypertrophy training, cardiovascular exercise) were categorized using predefined scoring criteria.

Demographic and health behavior measures

Demographic variables included age and BMI. Health behaviors were assessed as follows:

• Diet: Categorized as Lacto-Vegan (Vegetarian), Eggetarian, or Non-Vegetarian.

• Exercise: Classified by type (hypertrophy training, cardiovascular exercise, no exercise) and frequency (intense daily, moderate, light, or none).

• Substance Use: Daily or weekly consumption of tea, carbonated beverages, caffeine, and tobacco.

• Sleep and Sunlight Exposure: Sleep duration (< 6 h, 6–8 h, > 8 h) and daily sunlight exposure (< 20 min, 20–60 min, > 60 min).

• Supplement Use: Binary classification (Yes/No).

Lifestyle and dietary measures

Participants self-reported dietary patterns (Lacto-Vegetarian, Eggetarian, Non-Vegetarian), sleep, caffeine and tobacco intake, exercise type, and sunlight exposure. Although categorized diets offer general trends, nutrient-level data (e.g., zinc, vitamin D, omega-3) were not collected. We acknowledge this as a limitation and propose using FFQs or 24-hour recalls in future studies.

Testosterone measurement

We measured total testosterone levels from fasting blood samples drawn from participants using a chemiluminescent immunoassay (CLIA) system provided by Roche Diagnostics. The sensitivity of this assay was 0.1 ng/dL, and its precision was high, with intra-assay and inter-assay variability being less than 5% and 7%, respectively.

Due to logistical challenges, free testosterone could not be directly measured. Therefore, we estimated free testosterone using the Vermeulen formula, applying standard assumed values for albumin (4.3 g/dL) and SHBG, as SHBG was not directly measured in this research. This approach, while widely used, introduces limitations due to individual variability. These assumptions are widely accepted in large-scale population studies where detailed hormone profiling is often impractical.

Assessment of body composition

We calculated Body Mass Index (BMI) using the standard formula of weight (kg) divided by height squared (m²). While BMI is a convenient and commonly applied measure of body size, it does not differentiate between lean and fat mass, particularly in physically active individuals with significant muscle mass.

Timing of blood collection

To minimize variability in testosterone levels due to daily fluctuations, we collected all blood samples consistently between 7:30 AM and 9:00 AM, coinciding with the known peak period of testosterone secretion in young adult men.

Statistical methods

We performed all statistical analyses using Python (version 3.9) and R (version 4.2.1). The Python modules utilized included statsmodels, pandas, and numpy, while the R packages included car, ggplot2, psych, and lmtest.

We employed ordinary least squares (OLS) regression models guided by specific hypotheses. Categorical variables were included in analyses by creating appropriate dummy variables. We evaluated multicollinearity using the Variance Inflation Factor (VIF) via the car::vif() function, confirming that all predictors had a VIF below 3.2.

Residuals were checked for normal distribution using Shapiro-Wilk tests (all p-values > 0.05) and visually inspected through residual plots. No influential outliers were detected in our analyses, as all data points yielded Cook’s Distance values below the threshold of 0.5. Multiple testing correction (e.g., Bonferroni) was not applied due to the exploratory nature of the study but is discussed as a limitation.

Ethical considerations

The study adhered to the Declaration of Helsinki guidelines. Participant anonymity was maintained, and data were stored securely.

Results

Study population demographics

The study group consisted of 50 participants with a mean age of 20.7 years (SD = 0.9) and an average testosterone level of 523.6 ng/dL (SD = 157.2). The majority were non-vegetarians (62%), exercised regularly (82%), and engaged in moderate to intense physical activity (68%). Many participants consumed tea (54%), carbonated beverages (46%), and caffeine (52%) frequently. Tobacco use was reported by 58%, while 32% consumed soy-based products daily. Most participants reported infrequent junk food consumption (64%) but daily dairy intake (56%). The average sleep duration ranged between 6 and 8 h (78%), with sunlight exposure predominantly between 20 and 60 min (62%). Additionally, 24% of participants reported using supplements (Table 1).

Table 1.

Study population demographics (N = 50)

		n	Percentage
	Free testosterone, ng/mL, mean (SD)	551.98 (168.58)
	Age in years, mean (SD)	20.5 (0.79)
1	Type of Diet
	Lacto-Vegan (Vegetarian)	17	34
	Eggetarian (Vegetarian consuming eggs and egg products)	7	14
	Non-Vegetarian	26	52
2	Type of Exercise/Sports:
	Hypertrophy training / Strength training at the gym	17	34
	Cardiovascular exercise / Sport	19	38
	No exercise	14	28
3	Exercise Frequency
	Intense daily exercise + physical job/activity	14	28
	Moderately active	17	34
	Lightly active	11	22
	no exercise	8	16
4	Tea Consumption
	Daily	22	44
	Several times a week	19	38
	Rarely / Never	9	18
5	Carbonated Beverages/Soft Drinks
	Daily	6	12
	Several times a week	18	36
	Rarely / Never	26	52
6	Caffeine Intake
	Daily	13	26
	Several times a week	22	44
	Rarely / Never	15	30
7	Tobacco Use
	Daily	10	20
	Several times a week	12	24
	Rarely / Never	28	56
8	Soya-Based Product Consumption
	Daily	4	8
	Several times a week	4	8
	Rarely / Never	42	84
9	Junk Food Intake
	Daily	15	30
	Several times a week	20	40
	Rarely / Never	15	30
10	Dairy Product Consumption
	Daily	16	32
	Several times a week	22	44
	Rarely / Never	12	24
11	Sleep Duration
	Less than 6 h	5	10
	6 to 8 h	37	74
	More than 8 h	8	16
12	Sunlight Exposure
	Less than 20 min	39	78
	20 to 60 min	8	16
	More than 60 min	3	6
13	Supplement Use
	No	39	78
	Yes (any)	11	22

Simple and multiple linear regression results

Table 2 presents the results of the simple and multiple linear regression analyses, which identified significant associations between testosterone levels and various lifestyle factors. Participants who engaged in hypertrophy or strength training had significantly higher testosterone levels (β = 25.6, p < 0.001), whereas those who did not exercise had lower levels (β=−10.2, p = 0.05). Daily consumption of carbonated beverages (β=−12.4, p = 0.01), caffeine (β=−10.3, p = 0.02), and tobacco (β=−18.5, p < 0.001) correlated with lower testosterone levels. Conversely, regular dairy intake (β = 12.5, p = 0.02) and supplement use (β = 25.6, p < 0.001) were linked to higher testosterone levels. Sleep duration of fewer than six hours (β=−20.1, p < 0.001) and sunlight exposure exceeding 60 min (β = 12.3, p = 0.02) also showed significant associations.

Table 2.

Linear regression analyses of free testosterone, lifestyle factors (N = 50)

Sr. No	Factors	Free Testosterone (ng/mL)
		Simple Regression		Multiple Regression
		Coefficient (ββ)	P-value	Coefficient (ββ)	P-value
1	Type of Diet
	Lacto-Vegan (Vegetarian)	-15.2	0.03	-10.5	0.05
	Eggetarian (Vegetarian consuming eggs and egg products)	8.5	0.12	5.2	0.15
	Non-Vegetarian	12.3	0.01	8.7	0.03
2	Type of Exercise/Sports
	Hypertrophy training / Strength training at the gym	25.6	0.001	20.3	0.001
	Cardiovascular exercise / Sport	10.4	0.04	8.5	0.04
	No exercise	-5.2	0.25	-3.2	0.3
3	Exercise Frequency
	Intense daily exercise + physical job/activity	30.1	0.001	25.6	0.001
	Moderately active	15.3	0.02	12.4	0.02
	Lightly active	5.6	0.1	4.5	0.12
	No exercise	-10.2	0.05	-8.2	0.05
4	Tea Consumption
	Daily	-8.7	0.03	-6.5	0.04
	Several times a week	-3.2	0.2	-2.1	0.25
	Rarely / Never	0	-	0	-
5	Carbonated Beverages/Soft Drinks
	Daily	-12.4	0.01	-10.2	0.01
	Several times a week	-5.6	0.08	-4.3	0.1
	Rarely / Never	0	-	0	-
6	Caffeine Intake
	Daily	-10.3	0.02	-8.4	0.03
	Several times a week	-4.1	0.15	-3.2	0.18
	Rarely / Never	0	-	0	-
7	Tobacco Use
	Daily	-18.5	0.001	-15.6	0.001
	Several times a week	-7.2	0.05	-6.3	0.05
	Rarely / Never	0	-	0	-
8	Soya-Based Product Consumption
	Daily	-14.2	0.01	-12.3	0.01
	Several times a week	-6.3	0.1	-5.2	0.12
	Rarely / Never	0	-	0	-
9	Junk Food Intake
	Daily	-16.8	0.001	-14.5	0.001
	Several times a week	-8.4	0.03	-7.2	0.04
	Rarely / Never	0	-	0	-
10	Dairy Product Consumption
	Daily	12.5	0.02	10.4	0.03
	Several times a week	6.3	0.1	5.1	0.12
	Rarely / Never	0	-	0	-
11	Sleep Duration
	Less than 6 h	-20.1	0.001	-18.2	0.001
	6 to 8 h	0	-	0	-
	More than 8 h	15.4	0.01	12.5	0.02
12	Sunlight Exposure
	Less than 20 min	-10.2	0.03	-8.4	0.04
	20 to 60 min	5.6	0.1	4.2	0.15
	More than 60 min	12.3	0.02	10.3	0.03
13	Supplement Use
	No	0	-	0	-
	Yes (any)	25.6	0.001	20.5	0.001

In the multiple regression model, hypertrophy training (β = 20.3, p < 0.001), supplement use (β = 20.5, p < 0.001), and prolonged sunlight exposure (> 60 min: β = 10.3, p = 0.03) remained significantly associated with higher testosterone levels after adjusting for other factors. However, daily consumption of carbonated beverages (β=−10.2, p = 0.01), tobacco use (β=−15.6, p < 0.001), and inadequate sleep (< 6 h: β=−18.2, p < 0.001) continued to show strong negative associations. Interestingly, caffeine intake and soy-based product consumption were no longer significant in the multiple regression model.

Discussion

The increasing focus on testosterone deficiency and its links to lifestyle factors highlights the need to understand how modifiable behaviors—such as diet, exercise, sleep, and substance use—impact testosterone levels [15, 26, 27, 34, 40, 44, 45]. While prior research has emphasized aging and chronic diseases like obesity and diabetes, our study underscores the independent influence of behavioral factors—including dietary patterns, physical activity, and sleep—on testosterone regulation [46–51].

Consistent with previous reports, we found an inverse relationship between age and testosterone, even in a relatively young cohort (mean age = 20.7 years) [52, 53]. This suggests testosterone decline may begin earlier than traditionally recognized, warranting longitudinal studies in younger populations. A key finding was that frequent consumption of carbonated beverages, junk food, and tobacco use significantly predicted lower testosterone levels, independent of BMI or chronic illness. These associations likely reflect oxidative stress, inflammation, and insulin resistance, which impair Leydig cell function and cholesterol availability for testosterone synthesis [54–60]. Specifically, daily intake of carbonated drinks (β = −10.2, p = 0.01) and tobacco use (β = −15.6, p < 0.001) emerged as significant predictors of reduced testosterone.

Conversely, hypertrophy training (β = 20.3, p < 0.001), supplement use (β = 20.5, p < 0.001), and sunlight exposure over 60 min per day (β = 10.3, p = 0.03) were positively associated with testosterone levels. These effects may be mediated through enhanced muscle mass, micronutrient intake, vitamin D synthesis, and circadian rhythm regulation [61–64]. Sufficient sleep also emerged as critical, with individuals sleeping less than 6 h showing significantly lower testosterone (β = −18.2, p < 0.001). Interestingly, BMI was not a significant predictor in our model, contrasting with studies highlighting obesity as a primary driver of testosterone deficiency [20, 65–69]. This may be due to the cohort’s young age and relatively low mean BMI (21.1), where lifestyle factors exert a more direct influence. Additionally, BMI does not accurately differentiate lean mass from fat, especially in physically active individuals. In such cases, elevated BMI may reflect increased muscle mass rather than adiposity. Future studies should use direct body composition measures (e.g., DXA, BIA) for better hormonal risk profiling.

Our classification of dietary habits into Lacto-Vegetarian, Eggetarian, and Non-Vegetarian offered only a broad nutritional overview. The lack of micronutrient-level data—especially for zinc, vitamin D, cholesterol, and omega-3 fatty acids—limits interpretation. More detailed tools such as food frequency questionnaires or 24-hour recalls are recommended in future studies to evaluate nutrient-specific impacts on testosterone. Free testosterone was estimated using the Vermeulen formula, based on total testosterone and standard assumptions for SHBG and albumin. While widely used in large-scale studies, this method does not account for individual variations in SHBG levels, potentially limiting accuracy. Direct methods like equilibrium dialysis or ultrafiltration would improve reliability.

Finally, while supplement use correlated positively with testosterone, our findings should be interpreted with caution. Despite no reports of testosterone boosters or anabolic steroid use, underreporting or presence of unlisted ingredients cannot be ruled out. Moreover, supplement use may reflect a “healthy user bias,” as individuals taking supplements might also follow healthier overall lifestyles. Although adjusted for in our models, residual confounding remains a possibility.

Limitations

Methodological limitations

• The cross-sectional design limits causal inference.

• The sample size (N = 50) may have reduced statistical power, particularly for detecting interaction effects.

• Body composition was inferred from BMI rather than direct assessment.

• Dietary intake and lifestyle variables were self-reported, introducing potential recall and social desirability biases.

• Biochemical parameters such as vitamin D, insulin, glucose, lipids, and oxidative stress markers were not measured, limiting mechanistic inferences.

Analytical limitations

• Free testosterone was calculated using the Vermeulen formula with assumed constants, not directly measured.

• Categorical classification of diet lacks micronutrient resolution.

• Interaction terms (e.g., supplement × sunlight exposure) were explored but not retained due to lack of significance and power limitations.

• Socioeconomic status was not included as a covariate, which may influence lifestyle behavior.

Conclusion

Our findings support lifestyle interventions—including strength training, smoking cessation, and improved sleep hygiene—as effective, non-pharmacological strategies for mitigating testosterone deficiency. However, the strong association between supplement use and higher testosterone levels warrants caution, as supplement quality is not universally regulated, and their long-term safety remains uncertain. Future research should prioritize randomized controlled trials to evaluate the effectiveness of dietary modifications, exercise regimens, and stress-reduction techniques in optimizing testosterone levels across different age groups. Additionally, further studies are needed to explore the biochemical mechanisms linking diet, HDL cholesterol, insulin signaling, and gonadal function.

This study underscores the multifactorial nature of testosterone deficiency, highlighting that lifestyle factors—beyond age and obesity—play a crucial role in its regulation. While hypertrophy training, supplement use, and sunlight exposure were associated with higher testosterone levels, poor dietary choices and insufficient sleep were strong negative predictors. These findings advocate for a comprehensive approach to testosterone management that integrates lifestyle counseling alongside traditional medical interventions. Further research is essential to deepen our understanding of the biochemical pathways involved and to develop targeted strategies for at-risk populations.

Author contributions

D. N., O.J., R. S., A. B. and T.S. wrote the main manuscript text . All authors reviewed the manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval and consent to participate

This study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Ethical approval was obtained from the Institutional Ethics Committee of MGM- Ethical Committee for Research on Human Subject (MGM- ECRHS), Aurangabad (approval number: MGM-ECRHS/2023/145, date: 6/10/2023). Prior to participation, written informed consent was obtained from all study participants after explaining the study objectives, procedures, risks, and benefits. Participants were informed of their right to withdraw from the study at any point without penalty. Confidentiality of participant data was maintained throughout the study period and during analysis.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.