RAMADAN SERIES

Multi-country analysis of daily Stress Rhythm Score reveals significant gender and age-dependent differences among Ultrahuman Ring AIR users

Sama Dalal, Prejwal Prabhakaran, Ved Asudani, Bhuvan Srinivasan, Aditi Bhattacharya
Summary
Daily Stress Rhythm Scores of male and female Ultrahuman Ring AIR users were tracked in a country- and age-specific manner.
The study cohort included 27,885 Ultrahuman Ring AIR users from 33 countries, using stress rhythm scores based on heart rate variations.
Six countries with a comparable representative sample of women and having the lowest and highest Stress scores were chosen for further analysis. These included: Czechia, India, Switzerland, USA, UAE, and UK (13,149: males and 6,224: females).
Men reported higher stress levels (lower Stress Rhythm Scores) than females across all regions by ~4 points.
Younger adults reported higher stress than older adults by 14% across all countries.
In India, the UK, and Switzerland, users below the age of 25 years had the largest gender difference in stress levels, with a difference of ~5.5 points. In contrast, in the Czechia, the USA, and the UAE, the largest difference was observed in users aged 35-44 years, with a difference of ~5.6 points.
A trend of higher stress was observed in older women (>55 years) across a few countries.
Stress seems to have a distinct gender and age profile across countries which has strong implications for general health at a population level.
Background and Rationale
Biologically, stress is a phenomenon characterised by behavioural and physiological responses to real or perceived threats 1. Stress can be a powerful motivator for peak performance, known as “eustress,” such as when preparing for athletic competitions. However, stress that causes lasting damage is termed as “distress,” exemplified by conditions like post-traumatic stress disorder (PTSD). The World Health Organization defines clinical stress as a 'state of worry or mental tension caused by a difficult situation,’ specifically referring to distress 2. A 2021 Gallup survey reported that over 41% of adults globally experience high levels of distress 3, highlighting widespread challenges in managing these negative stressors originating from work, family and daily life 4.
Gender differences in stress responses have been well-documented, particularly relating to the hypothalamic-pituitary-adrenal (HPA) axis, which secretes hormones like Adrenocorticotropic hormone (ACTH) and Corticotropin-releasing hormone (CRH) that regulate the release of cortisol and adrenaline 1. For instance, adult men show a greater hormonal response to social tasks such as public speaking compared to adult women 4. In contrast, women exhibit a more prolonged stress response to negative facial perception and social relationship events, which has been linked to an increased risk of developing mood disorders 5.  Menstrual cycling, involving changes in estrogen and progesterone levels, also affects the female brain and alters threat perception. While consistent findings are reported in animal models, observational studies in humans have yielded variable results 1. Additionally, women undergoing menopause transition experience differing levels of hormonal and age-related stress, making it challenging to generalise findings across all females 6.
Observational studies, which are cohort-based and finite, offer snapshot-like stress information. To correlate stress and physiology better, a more longitudinal and personalised data framework is needed for men and women alike. Continuous monitoring devices, such as wearables with tailored algorithms, can provide this personalised information by interpreting non-specific data streams like heart rate in the context of lifestyle activities. The Ultrahuman Ring AIR's diurnal stress rhythm score is an example of such a technological feature. As a result, this study aimed to investigate the relationship between stress levels, gender, and age among Ultrahuman Ring AIR users across different countries, using daily stress rhythm score features to provide deeper insights into stress management profiles.
Methods
Data was collected from 19,607 male and 8,278 female Ultrahuman Ring AIR users across 33 countries from May 1st to June 5th, 2024  (previously described here). A secondary exclusion filter was employed to retain only countries with a minimum of 100 users per gender, resulting in 13 countries which were taken forward for analyses. Next, we focussed on six countries that exhibited the highest and lowest stress levels. This resulted in a total of 19,373 users (13,149 males, 6,224 females) from these six countries.
Stress levels in this study were based on the stress rhythm scores of these users. This score accounts for variations in heart rate (HR) in alignment with the phases of the circadian rhythm. A higher score suggests effective stress regulation (lower stress levels), while a lower score indicates dysregulated stress (higher stress levels).
To assess the differences in stress scores between genders across different age groups, users were categorised into six age brackets: below 25 years old (<25), 25 to 34 years old (25-34), 35 to 44 years old (35-44), 45 to 54 years old (45-54), 55 to 64 years old (55-64), and above 65 years old (65+).
Statistical analysis included a two-way ANOVA on ranks to assess gender and age interactions on stress scores. Subsequently, we performed pairwise comparisons using Dunn's test, with the Bonferroni correction applied to account for multiple comparisons. To compare stress scores between genders across regions, we employed the Mann-Whitney U test. The tests were performed using the statsmodels and scipy packages available in Python.
The analysis, conducted in compliance with the Ultrahuman Ring AIR application's terms of use, involved de-identified data to ensure user privacy. Data analysts had no direct contact with any users.
Result
Males have higher stress levels (lower stress scores) than females across countries
In our previous study, we found that the global median stress score for Ring AIR users was 76.95, with notable regional variability. As a follow-up, we segmented the data between genders and compared their stress scores to the global average (Figure 1). We observed that males had significantly lower stress scores compared to females across all regions (Mann-Whitney U test, p < 0.0001).
Specifically, in Asia, both males and females had the lowest stress scores, with males reporting 73.15 ± 15.10 and females at 77.12 ± 13.72 values. For both genders, stress scores were similar across Europe (Males: 77.35 ± 13.67, Females: 81.40 ± 12.29) and the GCC (Males: 77.30 ± 14.59, Females: 81.03 ± 13.31). The Americas showed the highest stress scores (lowest stress), with males at 78.40 ± 14.25 and females at 82.08 ± 13.37.
Figure 1: Boxplots denoting the difference in stress scores from the global median (76.95) for males and females across Asia (N: M = 6634, F = 1695), Americas (N: M = 5534, F = 4100), Europe (N: M = 6392, F = 1932), and GCC (N: M = 886, F = 353). Statistical analysis was performed using the Mann-Whitney U test. **** denotes p-value < 0.0001. Error bars denote the 95% confidence interval.
Stress score patterns across countries vary with age and gender
For a more detailed analysis of how stress scores change with age across genders, we selected six countries with the highest and lowest reported stress scores. Consistent with our earlier findings, we found that stress scores increased with age for both genders, but the patterns of this change differed between countries (Figure 2).
Using a two-way ANOVA to analyse the effects of gender and age on stress scores, we found significant differences between genders (p < 0.0001) and then across age (p < 0.0001) in terms of daily stress scores.
Across all countries, males under 25 years appeared to have the highest stress levels, except in India, where males aged between 25 and 34 years had the least Stress Rhythm Scores. In India (4.95), the UK (5.85), and Switzerland (6.22), the largest gap in gender stress scores was found in the under-25 years age group. Conversely, in the USA (5.25), Czechia (6.42), and the UAE (5.25), the most significant gender difference was seen in the 35-44 years age group.
Additionally, the increase in stress scores with age for females was less steep in India, the USA, and the UAE as compared to Czechia, Switzerland, and the UK. Finally, in the UK, for those over 65 years, females had lower stress scores than males. This trend was also observed in India and Switzerland, but for the 55-64 age group.
Figure 2: Stress Scores by age group for males and females across a) India (N: M = 4019, F = 930), b) Czechia (N: M = 1844, F = 479), c) the United Arab Emirates (N: M = 755, F = 308), d) the United Kingdom (N: M = 1019, F = 384), e) the United States of America (N: = 4912, F = 3908), and f) Switzerland (N: M = 600, F = 215). Statistical analysis was performed using Dunn's test. *, **, ****, and ns denote p-value <0.5, <0.01, <0.0001, and not significant, respectively. The error bars denote the 95% confidence interval.
Conclusions, Limitations and Future Directions
Stress significantly impacts healthspan and lifespan by increasing the risk of non-communicable diseases (NCDs). This study aimed to investigate the effects of gender and age on stress levels among Ultrahuman Ring AIR users. Contrary to much of the existing literature, our findings revealed that, on average, males tend to experience higher levels of stress compared to females. A potential reason for this observation could be that men engage in poorer stress management with respect to the circadian rhythm or that they face a higher volume of stressors throughout the day.
Our study also indicated that, on average, younger adults (both males and females) report higher stress levels compared to older adults. These observations align with some existing research suggesting that older adults may have developed more effective coping mechanisms over time, which help mitigate the impact of stress on their well-being 7. However, it is essential to appreciate that the nature of stressors faced by different age groups differs qualitatively. For example, younger adults might be dealing with academic pressures, early career challenges, and financial instability, whereas older adults might face health issues, uncertain care, and concerns about retirement.
Given that Ultrahuman Ring AIR users typically have higher fitness levels than the general population, our findings may not be broadly generalizable. Additionally, the stress rhythm score is based on a single primary data stream of heart rate, which can provide relatively unidimensional data.
In summary, our findings reveal significant gender and age-dependent variability in stress responses across different countries. The observed differences underscore the need for nuanced approaches to stress management that consider both gender and age. Addressing the limitations of this study, such as potential bias in the user base and the unidimensional nature of our stress measure, as well as expanding the scope of future research, will enhance our understanding of these gender and age-related stress patterns. This, in turn, will improve strategies for managing stress effectively across diverse populations and age groups.
Reach out to support@ultrahuman.com for commercial queries and science@ultrahuman.com for scientific queries.
© Ultrahuman 2024