In the current study, we measured serum levels of ECBs, cortisol, and progesterone in 700 pooled samples stratified by age across the lifespan and sex. Confirming that the experiment was robust, serum progesterone showed the expected lifespan variability in females. Similarly, ECB and hormones across the age and sex groups visualised in the PCA plot showed expected variability. This is the first large-scale study to our knowledge of ECBs and related hormones across the lifespan to have reported age and sexual dimorphism, suggesting that these compounds correlate with known periods of major physiological changes in life. Therefore, the current study’s focus on ECBs and related hormones and their age and sex differences may have implications on neurological changes throughout the lifespan, and perhaps how these compounds could be further investigated to improve cognitive function and inform future pathological treatments.

The peak in female cortisol levels during the adolescent and early adult years may be due to pubertal development³³. Additionally, the increase in male cortisol levels following the 30–45 age group occurred simultaneous to the general increases observed in the ECBs, with the exception of AA. Both cortisol and ECBs are influenced by the hypothalamic–pituitary–adrenal (HPA) axis³⁴, with an increase in cortisol associated with increased ECB circulating levels³⁵. The HPA axis alters and becomes less efficient with age³⁶, so it is possible that the similar cortisol and ECB levels are confounded by HPA axis activation.

However, the changes in female cortisol levels are less straightforward, peaking in the 15–30 age group and slightly declining with age until it peaks again in the 85+ age group. Previous studies on the age-related trajectories of cortisol have reported conflicting results, with some showing increases over the lifespan^37,38, and some showing decreases³⁹. A recent study has even shown that cortisol levels over the lifespan follow a non-linear, U-shaped function⁴⁰. Seeman et al.⁴¹ found similar cortisol responses reflected in the current study, with males aged 67–88 showing higher levels of cortisol than males aged 22–26. Though, the current study has utilised the broadest age range in available literature, with discernibility between males and females. As the secretion of cortisol and level of available cortisol activate the HPA axis, these results may have implications on the regulation of stress and neuroprotection in older age.

Oettel and Mukhopadhyay⁴² found that male progesterone levels remained stable and low throughout their lifetime, with no age-related changes reported. This was consistent with our data. However, in the current study, female progesterone levels greatly increased and peaked in the 30–45 age group, then decreased and plateaued starting from the 45–60 age group. The pubertal increase in female progesterone levels occurred simultaneous to the increase in female AA levels. Wilson et al.⁴³ found evidence that progesterone modulates the release of AA, with the inhibition of progesterone increasing the release of AA. We observed that female AA levels increased greatly from the 15–30 age group to the 30–45 age group and continued to increase slowly but insignificantly from then onward. Female progesterone levels increased greatly until the 30–45 age group, and significantly decreased afterward. It is possible that our data reflects a direct relationship between female progesterone and AA, with loss of progesterone reducing the inhibition of AA release.

Serum progesterone in females showed the expected lifespan variability encountered with puberty onset (15–30), middle-end reproductive life and peri-menopause (30–45), and menopause (45–60)^44–46. It may be possible that the use of hormonal contraceptives or hormone replacement therapy (HRT) impacted the detection of progesterone and cortisol, particularly in the younger females regarding the former. Furthermore, this may have altered HPA axis functioning. Combined hormonal contraceptives in particular inhibit the natural production of estrogen and progesterone, thereby eliminating menstrual-cycle-based variability⁴⁷. This alters the HPA and the hypothalamic-pituitary-ovarian axis, preventing the maturation and release of an oocyte⁴⁸ and inhibiting the rise in estrogen that usually occurs in the follicular phase⁴⁹. Anovulation results in static progesterone synthesis in the ovary, which is then maintained at baseline levels⁴⁷. Consequently, females that take hormonal contraceptives have lower serum progesterone than naturally cycling females⁵⁰. Prior studies have also demonstrated that serum cortisol levels are typically elevated in people currently using oral contraceptives⁵¹.

Similarly, progesterone levels have decreased slightly in women taking menopausal progesterone-based HRT in comparison to menopausal women not taking HRT⁵². At least 29% of Australians choose to use an oral contraceptive pill⁵³ and 38% of menopausal Australian women report having ever used HRT before⁵⁴, which may suggest that these results were impacted by some samples that had elevated serum cortisol or lower serum progesterone levels due to the contraceptive pill or HRT.

There is more limited research on the effect of ECBs on the female reproductive system. Though, the fact that marijuana has severe consequences on pregnancy suggests that endogenous ligands of the receptors are involved in the modulation of pregnancy, menopause, and the menstrual cycle⁵⁵. AEA has been quantified in human follicular fluid⁵⁶, in which its levels strongly correlated to oocyte maturation and quality in women undergoing infertility treatment⁵⁷. When quantifying levels of AEA across pregnancy, it was also found that AEA levels were high in the first trimester – similar to levels measured in the luteal menstrual phase—and increase almost two- to four-times more with the onset of labour^58,59. Therefore, ECB levels are most likely to fluctuate during a time of pregnancy, which may partially explain the slight increase of ECB levels in females in the 15–30 age group. The highest number of pregnancies occur in women’s twenties, with an increasing number of pregnancies occurring in women in their thirties⁶⁰. Due to the unidentifiable nature of the pathology test data collection, the current study is unable to discern pregnant females within the study population. Though, these results may have implications on female reproduction and fertility, and how ECB levels may affect these. Future research should monitor ECB levels over the course of pregnancy to further investigate this.

Additionally, female menopause and post-menopause are partially affected by ECB levels. The activity of menopause, and the end of the menstrual cycle fluctuations, were significantly associated with increased 2-AG levels in a metabolic parametric study⁶¹. Postmenopausal AEA levels were also found to be similar to those in the first trimester of pregnancy and the luteal menstrual phase, which were significantly higher than AEA levels in the later trimesters⁵⁸. Similarly, the current study found that females in the 40–60 age range had higher ECB levels compared to younger age groups. Additionally, females in the 60–75 age group, where post-menopause is most experienced, recorded a peak increase in AEA and 2-AG levels. Therefore, as the activities of the menstrual cycle cease in mid-life, ECB levels tend to increase to a peak. This may have potential implications on how ECBs are neuroprotective in later life⁶² and, more broadly, have a role in neurological decline.

AA levels trended differently to the other ECBs; as age increased, AA levels were higher in females compared to males. In contrast, as age increased, AEA, 2-AG, OEA, and PEA were higher in males compared to females. Similar results were found by Lohner et al.,²⁰ in which AA was significantly lower in males compared to females, across the age groups. Additionally, AA was significantly higher in females aged 51 years and older, with an insignificant increase in females aged 13 to 50 years old, compared to males. This may be because AA, in comparison to the other ECBs, is also considered a long-chain polyunsaturated fatty acid. Recently, there has been growing evidence of a complex interplay between the ECB system and the long-chain polyunsaturated fatty acids⁶². Notably, both AEA and 2-AG are derived from AA⁶³. If there is a high amount of AA available in one’s body, it can potentially be converted into comparatively more AEA and 2-AG. Similarly, when these ECBs are no longer needed, they are rapidly degraded back into AA⁶². Therefore, it may be suggested that if AA is high, AEA and 2-AG may be lower. This can be further evidenced by findings of inversely proportioned 2-AG and AA in mice⁶⁴. For the current study, this may explain that while AA was higher in females than males, AEA and 2-AG were observed to be comparatively lower. Additionally, female progesterone levels were negatively correlated with 2-AG levels. As female progesterone levels spiked during puberty and pre-menopause, 2-AG levels remained low. In older years, progesterone decreased while 2-AG levels increased. This may indicate progesterone regulation of 2-AG (or vice versa), or may simply be a coincidence given that 2-AG also increased in older age in men. Potentially, the rate of metabolism of AA into its metabolites 2-AG and AEA may change over the lifespan, and it is possible that this is influenced by changes in sex hormones with increased age. Further research into this question may have implications in neurogenesis and brain development across the lifespan⁶⁵.

A study by Fanelli et al.⁶⁶ may suggest an alternative confounding explanation of our ECB results. By measuring body mass index (BMI) and the metabolic parameters of normal, overweight, and obese adults, Fanelli et al.⁶⁶ suggested that N-acylethanolamines play specialised roles in the body’s energy balance. AEA influenced energy intake and storage and became a biomarker for abdominal fat, while OEA mediated anorectic signals and lipid oxidation. Additionally, increased AEA in men and low OEA in men and menopausal women were also associated with low HDL-cholesterol. Abdominal fat and low HDL are highly characteristic of obesity⁶⁷. The likelihood of obesity peaks between 40 to 75 years old⁶⁸ and is more prevalent in older males than older females⁶⁹. Additionally, ECBs increase and become overactive in obese individuals⁷⁰. This was reflected in our results, as AEA, 2-AG, and OEA peaked in older males 60–75 years old. These ECBs also remained higher in males across almost all age ranges, compared to females. Therefore, obesity was a potential confounding factor that influenced the age and sex-related effects on ECBs.

Male testosterone levels across the lifespan potentially relates to the ECB levels recorded in the current study. Previous animal studies have provided direct evidence of the ECB system in the testis, thereby implicating sex-related differences and an effect on testosterone production^71–73. In marijuana users with a highly expressed ECB system, it was found that plasma testosterone was reduced in comparison to non-marijuana users⁷⁴. Therefore, when ECB levels are elevated, testosterone levels are decreased. Additionally, AEA was found to downregulate luteinising hormone in the male anterior pituitary^75,76. Luteinising hormone upregulates testosterone production⁷⁷, which suggests that AEA activity leads to a decrease in testosterone production. Testosterone levels are found to peak in late teen years, and thereby steadily and slowly decrease over the lifespan, and more readily decrease after 80 years old^78–80. In the current study, male ECB levels trend upwards, with most ECBs peaking in the 60–75 age group and dropping in the 75–85 age group before slowly increasing again. Interestingly, AEA and OEA levels were also found to be significantly positively associated with cortisol, only in male serum. The addition of testosterone as a measurable variable may have allowed better understanding of the relationship between male cortisol levels and ECB levels. It is possible that as male testosterone levels slowly decrease across the lifespan, ECB levels slowly increase in association. Future research should look at broader sex-related hormones and ECBs across the lifespan to specifically investigate male sex-related changes, and a potential implication on male sex-related diseases such as prostate cancer and testicular cancer.

To date, the current study is the largest on this topic, with a broad range of under-researched ECBs and related hormones. The results from this study could have implications on how ECBs affect age-related diseases and declines, especially as ECBs are mostly produced in the brain. Therefore, the progression of ECB levels across the lifespan could inform understanding of neurological diseases, such as dementia, brain cancer, and Parkinson’s disease. ECBs are also produced in various muscles, fatty tissues, and immune cells, and thus the current results have implications on broader affected pathways. As ECBs are related to sex hormones, further research should be conducted on how the expression of ECBs and a more exhaustive list of sex hormones develop over the lifespan. Specifically, further investigating testosterone could inform further understanding of the changes in male ECB levels.

As one of the first investigations of this size, it is important for future studies to expand on the results of the current study. Pooling the data in the current study enabled a large sample size of highly generalisable data from the general population, but due to the deidentifying nature of pathology test data collection, no demographics or possibly confounding health conditions were recorded. Therefore, it cannot be truly concluded that these ECB, cortisol, and progesterone levels reflect actual changes across one’s lifetime. Additionally, 2-AG and PEA levels were significantly positively associated in the pooled analysis of male and female sera, but no longer reached significance once stratified by sex. It may be possible that separate lifespan changes for males and females may have affected these results, such as neurological decline, female menopause, or obesity, though the experimental design of the current study has limited the ability to discuss these implications. Future studies should be encouraged to build upon these results to further understand the sex- and age-related affects on 2-AG and PEA, amongst related ECBs. Ideally, the sample should be followed throughout their lifetime with their analyte levels recorded consistently. Though, the sampling of the current study utilises one of the most readily available and convenient methods. Additionally, due to the wide scope and unanimity of pathology sampling in Australia, a considerable portion of blood tests are taken by healthy individuals for other routine non-pathological reasons, such as immunology, nutrition, pregnancy, or medication. Any effect that undisclosed health conditions may have had on the results were most likely insignificant.

It is also noted that, due to sample pooling, diurnal variations of cortisol and cyclic variation of progesterone in females could not be analysed. There is evidence of changes of ECB, progesterone, and cortisol levels across age strata mirroring lifespan milestones such as puberty and female menopause, though not across different times of the day or ovarian cycle. Peak cortisol levels are recorded in the morning and decrease to its lowest level late at night⁸¹, while progesterone is decreased during the follicular phase and increased in the luteal phase⁸². Though, research on cortisol diurnal intensity and progesterone cyclic intensity across the lifespan is limited. To compare the activity and adaptations across the lifespan, it is recommended that these patterns are further investigated across age and sex in the future.

The current study finds evidence for changes in ECB, cortisol, and progesterone serum levels across age that were substantially affected by participant sex. These findings add to our knowledge of the relationship between ECBs and sex hormones as well as life changes, and support the theory that ECBs show sex-dimorphism across the lifespan.