Chronic pain prevalence and severity in females before & after menopause versus males in a large multi-center database

ABSTRACT

Aims

This observational study aimed to explore the prevalence and severity of chronic pain in menopausal-aged females.

Materials & Methods

We identified participants in the NIH-sponsored All of Us database with a diagnosis of chronic pain. We examined the severity of pain based on numeric rating scale score (NRS). Among those with chronic pain, we generated four sub-cohorts: females ≥ 51 years old who were menopausal, females 18–50 years who were considered pre-menopausal, and age matched male cohorts.

Results

The 53,289 participants were separated into four cohorts by age and sex: females 18–50 (10,269), females ≥ 51 (20,182), males 18–50 (4,989), and males ≥ 51 (17,849). Both older females and older males had lower pain levels than the sex-matched 18–50 group (p < 0.001). Findings were consistent in White, Black, and non-Hispanic cohorts. When comparing older adults, females ≥ 51 had higher pain levels than males ≥ 51 (p < 0.001).

Conclusions

We found that independent of sex, NRS scores were lower in older adults compared to younger cohorts. Older women however had more pain than their male counterparts. These findings have implications for the design of patient-tailored therapeutic treatments and warrant further exploration of physiological and psychosocial causes.

Plain Language Summary

Chronic pain is a major health concern that affects millions of people and interferes with everyday life. Women often report pain more than men and this may be related to hormones as they age, especially around menopause. In this study, we identified participants in a large, diverse national database with a diagnosis of chronic pain to see how age and menopause affect pain levels. We compared 4 groups: older women who were menopausal, younger women who were pre-menopausal, and men in the same age groups. We found that older adults reported less pain than younger adults. However, older women had higher pain scores than men of the same age. This was true across racial and ethnic groups. This may suggest that hormone changes in women may play a role in how pain levels. Understanding this difference can help create more personalized treatment plans.

1. Introduction

Chronic pain is a major worldwide health concern and is characterized by a wide range of clinical presentations due to the complex interplay between nociceptive and neuropathic characteristics [1]. There is evidence that chronic pain disproportionately affects females more than males [2–4]. The prevalence of chronic pain in adults in the United States is estimated to be 20%, with 8% suffering from extreme levels of chronic pain, with effects on day-to-day and work activities [5]. Due to these extreme levels, the health of a population is impacted beyond the emotional, physical, and sociological toll it places on individuals. It is estimated that chronic pain costs the United States $560–635 billion per year when accounting for direct medical costs, lost productivity, and disability programs [6]. Although the prevalence and impact of chronic pain is well documented, how the burden of this condition and related comorbidities are distributed across the population has not been well defined [7–10].

When considering chronic pain in females and the influence of aging, it is important to consider whether menopause has occurred and the effect of the menstrual cycle and its associated changes in the concentrations of reproductive hormones and regulation of body water [11]. Menopause is associated with a decrease in estrogen production and occurs most commonly in the United States between the ages of 48 to 52 [12]. Menopausal symptoms can vary widely; menopause can cause pain to develop, intensify, or improve [13]. Menopausal symptoms that are associated with chronic pain include hot flashes, sleep disorders, depression, chronic fatigue, anxiety, and many others [12]. Moreover, the role of estrogen in pain regulation has been extensively explored, with previous studies suggesting that the decline in estrogen is associated with increased pain sensitivity and predisposition to developing chronic pain such as musculoskeletal pain and fibromyalgia [14–16]. The mechanisms by which estrogen modulates pain are quite complex and involve multiple pain pathways. Previous studies have suggested that estrogen plays a role in nociception and endogenous opioid neurotransmission [17].

Differences exist in the development and experience of pain between sex and age, with menopausal status playing an important role in women and how they experience pain. Thus, it is relevant to investigate how these factors including sex, age, and menopausal status influence the severity of chronic pain in different groups of people. This study leverages the All of Us database, which is a diverse dataset that includes the historically underrepresented in research, such as racial and ethnic minorities, rural populations, and older adults across the country which better reflects the composition of the U.S. population. Over 840,000 people have joined this database. To date, no published study has used this database to examine the prevalence and severity of chronic pain among pre- and post-menopausal-aged females and compared it to age-matched males. By incorporating menopausal status and focusing on sex, age, and race/ethnicity as demographic characteristics, this study may contribute to more personalized approaches to chronic pain management in the future, potentially lowering the burden of chronic pain on individuals and society. These findings can ultimately inform clinical trial design by highlighting the importance of not only age, sex, and ethnicity but also hormonal status and lead to more effective pain management treatment.

2. Materials and methods

The data in this study was derived from the All of Us research program, which originated in May 2018 with the intention to improve human health and accelerate biomedical research through precision medicine. The goal of the All of Us research program is to recruit a diverse group of patients in the United States to ensure representation of all races, ethnicities, and socioeconomic backgrounds [18]. The Florida Atlantic University Health Sciences Institutional Review Board reviewed and determined that this study was exempt from federal regulation under 45 CFR 46.104 and did not meet the definition of human subjects research. All methods were carried out in accordance with relevant guidelines and regulations.

We first identified anyone who had a chronic pain diagnosis in the All of Us database via ICD-10 classifications. Chronic pain diagnoses in the database included chronic pain syndrome, fibromyalgia, persistent pain following procedure, neurogenic claudication, chronic intractable migraine without aura, chronic back pain, complex regional pain syndrome, intermittent claudication, and transformed migraine. To assess pain severity in the groups, we only included patients with chronic pain who also had a numeric rating scale (NRS) score documented. The NRS is a widely used unidimensional tool to provide a reliable and valid assessment of subjective pain levels with 0 representing no pain and 10 representing the worst pain imaginable. In our analysis, we classified those with NRS scores <7 as mild/moderate chronic pain, and ≥ 7 or higher as severe pain [19].

Given that the mean age of menopause in the United States ranges from 48–52 [12], to compare age-matched groups, we included females 18–50 years old and females > 51 years old with no upper age limit imposed compared to males 18–50 and males >51 also with no upper age limit imposed. In females >51, they were only included if menopause was documented and in females 18–50, patients were excluded if they were experiencing menopause. Sex, age, and race (White, Black, Asian, more than one population) and ethnicity (Hispanic, non-Hispanic) were examined as covariates using ANOVA with Tukey post-hoc analysis. Statistical analyses were conducted to assess correlations between demographic factors and chronic pain severity. Statistics were performed using a Jupyter Notebook run in Python with data from All of Us Registered Tier Dataset v7, and in R Studio version 2022.07.2. A p-value of <0.05 was considered significant.

3. Results

We identified 53,289 individuals with ICD-10 codes consistent with chronic pain and documented NRS scores. The most common diagnoses comprising greater than 80% of patients were chronic pain syndrome and fibromyalgia; more males had the former and more females had the latter (Table 1). Of these participants 30,451 were females (57.14%) and 22,838 (42.86%) were males, with most participants reporting mild-to-moderate pain (Table 2). Both older females and males had lower mean pain scores than their younger same sex counterparts (4.38 ± 2.96 v. 4.23 ± 2.93 for women (p < 0.001), 4.48 ± 3.03 v. 4.09 ± 2.93 for men p < 0.001). Notably, males >51 in the total cohort had lower pain scores than similar aged females (4.09 ± 2.93 v. 4.23 ± 2.93 (p < 0.001)). We next compared sex- and age-matched cohorts based on severity of pain. In the severe pain groups, there were no differences in pain scores. In the mild/moderate pain group, women under 50 had higher pain scores than those >51 (2.92 ± 2.06 v. 2.83 ± 1.99 (p < 0.05)). Also in this group, females >51 had higher pain scores than males > 51 (2.83 ± 1.99 v. 2.72 ± 2.00 (p < 0.001)).

Table 1.

Percentage of the participants with specific pain diagnoses in each subset.

	Males (%)			Females (%)
Chronic Condition	18–50	> 51	% out of Total Male	18–50	> 51	% out of Total Female
Chronic Pain syndrome	85.71	80.72	81.60	68.25	74.12	72.39
Fibromyalgia	2.70	2.08	2.19	15.43	12.20	13.15
Persistent pain following procedure	4.52	3.82	3.94	2.90	2.47	2.59
Neurogenic claudication	0.51	4.76	4.01	0.10	2.97	2.13
Chronic intractable migraine without aura	1.77	0.71	0.89	5.62	1.79	2.92
Chronic Back Pain	1.18	1.65	1.57	1.45	1.57	1.54
Complex regional pain syndrome	1.34	1.41	1.40	2.18	1.75	1.87
Intermittent claudication	0.26	3.25	2.73	0.19	1.01	0.76
Transformed migraine	0.07	0.03	0.04	0.58	0.22	0.32
Other	1.89	1.53	1.59	3.27	1.87	2.28

Table 2.

Cohort sizes with mean NRS and standard deviation.

Cohort, Age (years)	Total Participants (Mean NRS + SD)	No. of Participants with NRS < 7 (Mild-Mod) (Mean NRS + SD)	No of Participants with NRS ≥7 (Severe) (Mean NRS + SD)	Prevalence of Severe Chronic Pain
Females, 18–50	10269 (4.38 ± 2.96)	7362 (2.92 ± 2.06)	2907 (8.09 ± 1.09)	28.31%
Females, ≥ 51	20182 (4.23 ± 2.93) #***	14799 (2.83 ± 1.99) #*	5383 (8.07 ± 1.07)	26.67%
Males, 18–50	4989 (4.48 ± 3.03)	3420 (2.83 ± 2.04)	1569 (8.08 ± 1.09)	31.45%
Males, ≥51	17849 (4.09 ± 2.93) #***, &***	13235 (2.72 ± 2.00)&***	4614 (8.01 ± 1.05)	25.85%
Total Sample	53289	38816	14473	27.16%

Comparisons are between groups accounting for sex and age using ANOVA with Tukey’s correction. #: Same Sex, 18–50 v ≥51; &: Same Age, Female vs Male. *:p < 0.05; **:p < 0.01; ***:p < 0.001.

In the total cohort, 63.09% identified as White, 16.89% Black or African American, 1.8% Asian, 1.49% more than one population, and 16.41% belonged to other populations or did not identify. In terms of ethnicity, 82.14% were non-Hispanic, 14.46% were Hispanic, and 1.22% did not identify with either category. We then looked at the effect of race and ethnicity on groups. In both White and Black patients, older males and females had lower mean NRS scores as a group compared to younger patients (Table 3; p < 0.001). There were no significant age-related differences in Asian or more than one population in males and females. Consistent with the total cohort, White males > 51 had lower pain scores than White females > 51 (4.13 ± 2.95 v. 4.23 ± 2.93 (p < 0.05)). A similar pattern was found in the group identifying as more than one population (p < 0.01). Asian males > 51 have a mean NRS score higher than age-matched Black males (Table 3; p < 0.05). In non-Hispanic patients, older males and females had lower mean NRS scores as a group compared to younger patients (Table 4; p < 0.001). However, in non-Hispanic patients, males > 51 reported significantly lower NRS scores than females > 51 (4.11 ± 2.93 v. 4.24 ± 2.93 (p < 0.001)). There were no statistically age-related differences in either sex for Hispanic or Neither groups.

Table 3.

Mean and standard deviation of NRS and pain severity in each group based on race and sex.

Sex	Race	Age	Mean NRS ± SD	NRS < 7	NRS ≥ 7
Female	White	18–50	4.40 ± 2.95	2.95 ± 2.06	8.09 ± 1.09
		> 51	4.23 ± 2.93#*	2.84 ± 2.00	8.08 ± 1.07
	Black	18–50	4.27 ± 3.01	2.81 ± 2.05	8.14 ± 1.12
		> 51	4.25 ± 2.91#***	2.86 ± 1.99	8.05 ± 1.08
	Asian	18–50	4.87 ± 3.08	2.95 ± 1.98	8.30 ± 1.01
		> 51	4.24 ± 2.78	2.92 ± 1.81	8.05 ± 1.01
	MT1P	18–50	4.39 ± 2.78	3.26 ± 2.08	8.01 ± 1.18
		> 51	5.13 ± 2.54	3.70 ± 1.65 WM***, BM***	8.07 ± 1.13
Male	White	18–50	4.50 ± 2.99	2.89 ± 2.02	8.08 ± 1.08
		> 51	4.13 ± 2.95#***, &*	2.72 ± 2.00	8.01 ± 1.05
	Black	18–50	4.76 ± 3.09	2.87 ± 2.08	8.12 ± 1.08
		> 51	4.02 ± 2.91#***	2.73 ± 2.01	8.02 ± 1.04
	Asian	18–50	3.99 ± 2.96	2.70 ± 2.03	8.17 ± 1.13
		> 51	4.60 ± 3.04BA*	2.93 ± 2.05	8.32 ± 1.01
	MT1P	18–50	4.65 ± 3.02	2.34 ± 2.06	7.88 ± 1.08
		> 51	4.07 ± 2.25WM**, BM***, &**	3.39 ± 1.89	7.89 ± 0.39

Comparisons between groups accounting for sex using ANOVA with Tukey’s correction. #: Same race and sex, 18–50 v ≥51. &: Same race and age, Female vs Male. Within same sex and age, WB: White vs Black. WA: White vs Asian. WM: White vs MT1P. BA: Black vs Asian. BM: Black vs MT1P. AM: Asian vs MT1P. *:p < 0.05; **:p < 0.01; ***:p < 0.001.

Table 4.

Mean and standard deviation of NRS and pain severity in each group based on ethnicity and sex.

Sex	Ethnicity	Age	Mean NRS ± SD	NRS < 7	NRS ≥ 7
Female	Hispanic	18–50	4.33 ± 3.06	2.90 ± 2.08	8.02 ± 1.08
		> 51	4.12 ± 2.90	2.80 ± 1.96	8.09 ± 1.08
	Not Hispanic	18–50	4.39 ± 2.94	2.92 ± 2.05	8.10 ± 1.10
		> 51	4.24 ± 2.93#***	2.84 ± 1.99	8.07 ± 1.07
	Neither	18–50	4.34 ± 2.83	2.65 ± 2.13	8.61 ± 1.12
		> 51	4.34 ± 2.65	2.40 ± 1.90	8.08 ± 1.13
Male	Hispanic	18–50	4.41 ± 2.95	2.59 ± 2.04	8.10 ± 1.12
		> 51	4.10 ± 3.00	2.70 ± 1.98	7.98 ± 1.07
	Not Hispanic	18–50	4.48 ± 3.04	2.90 ± 2.03	8.09 ± 1.08
		> 51	4.11 ± 2.93#***, &***	2.73 ± 2.00	8.01 ± 1.05
	Neither	18–50	4.29 ± 3.21	3.45 ± 2.21	7.90 ± 1.06
		> 51	4.40 ± 2.96	2.92 ± 1.98	7.93 ± 0.99

Comparisons are between groups accounting for sex using ANOVA with Tukey’s correction. #: Same Ethnicity 18–50 v ≥51. &: Same Ethnicity at same age Female vs Male. ***:p < 0.001.

4. Discussion

Our analysis demonstrates that mean NRS values for chronic pain are significantly lower in males and females ≥51 as compared to the sex-matched younger cohort. These findings held true in patients that identified as White, Black, or non-Hispanic. Additionally, we found that older females had higher pain scores than similarly aged males both in the total cohort and in the non-Hispanic, White, and more than one race populations specifically. In patients with mild to moderate pain, post-menopausal-aged females had lower scores than younger women but higher scores than similar aged men. Previous reports of chronic pain severity have demonstrated that women experience more intense pain and disability than their male counterparts [3,20]. One of the explanations for this is that females and males may perceive and process pain along different pathways. There is evidence in preclinical models that males perceive pain through the microglial pathway and females perceive pain through the T-cell pathway [9]. The use of the microglial pathway in males allows for the release of peripheral inflammatory biomarkers [21]. This pathway could partially explain differences in pain perception between sexes as females do not use this pathway.

The study of chronic pain in menstruating women is particularly difficult. Fluctuations in several reproductive hormones, particularly estrogen and progesterone, directly influences nociception along different pathways, which can in turn affect pain perception. These cycles likely impact reported pain scores at any one moment in time and can vary widely with individuals. Moreover, this issue is further complicated by the gradual to abrupt decline in these hormone levels women experience during menopause. Taken together, recent research into the effect of the absence of hormones such as estrogen and progesterone on chronic pain in menopausal and post-menopausal women has demonstrated significant correlations between more severe chronic pain, increased opioid use, and menopausal symptoms [2,22–28]. Making the study of chronic pain over the life of a patient even more complicated, there is strong evidence that the types of pain that are commonly reported also changes through a patient’s life as seen in our data. There is conflicting data as to whether menopause worsens or improves pain or whether pain stays the same, depending on the type of pain experienced [13,29]. The older a person, the more likely they will have had an injury or been exposed to something that triggers chronic pain [3]. However, it is widely appreciated that older adults are less likely to report pain to their physicians and to minimize it. The reasons for this are multi-factorial and may in part be due to expectations that pain comes with aging as well as fear that pain may represent a serious condition [30,31]. These cultural and behavioral factors may contribute to the underreporting and potentially skewed pain evaluation in older adults.

Our analysis also compared the effects of ethnicity and race controlled by age and sex on chronic pain severity. Intriguingly, our data when stratified by race demonstrates that Asian males and multiracial females show a trend toward increasing mean NRS pain scores over time. This is in direct opposition to all other race and sex combinations which show a consistent decrease in average pain score with aging. This trend may be due to a limited sample size, as only 1.49% of the cohort identified as belonging to more than one population. However, it is worth noting that previous studies have highlighted differences in pain perception, pain management, and treatment access across different racial and ethnic groups [32,33]. Specifically, individuals with Asian descent tend to report higher levels of clinical pain and experimental pain sensitivity, compared to non-Hispanic White individuals [34].

Despite the significance of these findings, we acknowledge that it is important to recognize the limitations of our study. Nearly 2/3 of the All of Us database population from which to select our sample consisted of females and could introduce self-selection bias. While the present study focused on high-level demographic comparisons, this study lacked potential confounders such as socioeconomic status, comorbidities, and medication use which could provide greater interpretive validity.

Although NRS is a commonly used tool to measure pain intensity, it does not take into account the tridimensional aspect of pain. Incorporating additional pain assessment tools such as Pain Catastrophizing Scale (PCS), the Roland-Morris Disability Questionnaire (RMDQ), and physiological pain measures (Pressure Pain Threshold – PPT, Temporal Summation of Pain – TSP, and Two-Point Discrimination Threshold – TPD) can provide more information about the level of pain severity beyond pain intensity alone.

We also did not account for early-onset menopause in both groups: females aged 18–50 who were menopausal were excluded while females aged 51 and older were not excluded if they had premature menopause. This is an important limitation, as early onset menopause as this may have different effects compared to menopause at or after the typical age.

While we aimed to differentiate between pre- and post-menopausal women based on menopausal status and age, this approach does not allow for precise reproductive staging using validated criteria such as STRAW + 10 [35] Future studies will include this criteria.

The average NRS score for the group with scores < 7 and classified as mild/moderate chronic pain was below 3 in most cohorts, indicating many experienced minimal pain. Using 3 categories (no/low pain, moderate pain and severe pain) could have potentially reflected the spectrum of pain severity more accurately and reduced bias. Additionally, the size of each cohort in this study is not uniform due to the expectation that pain increases with age, which could have skewed our results.

5. Conclusion

In this study, we demonstrate that pain scores decrease with age in both sexes. However, older women have higher rates of pain than age-matched men, suggesting potential biological, hormonal, and psychosocial factors. This difference is not observed in the younger cohort. Further research is warranted to elucidate these differences and potentially develop targeted pain management strategies for menopausal women.

Funding Statement

This paper was not funded.

Article highlights

• Chronic pain is more commonly reported by women than men and this may be influenced by menopause.

• Using the All of Us Research database, we analyzed pain severity in over 53,000 participants with chronic pain and divided these participants into 4 groups: females 51 and older who were menopausal, females under 51 who were pre-menopausal, and age-matched men

• Older adults reported lower average pain levels than younger adults, regardless of sex

• Older women reported high pain levels than older men

• These trends were consistent across racial and ethnic groups

• It is important to consider sex and age when managing chronic pain

• These findings support more personalized pain treatment strategies, especially for older women

Author contributions

Conceptualization – JGP; Data analysis: LG, DB, IT; Visualization: LG, DB; Supervision: JGP; Writing – original draft: AR, UEHA Writing – review & editing: AR, NF, JJM, OK, UEHA, CW, JGP.

Disclosure statement

Dr. Pilitsis receives grant support from Medtronic, Boston Scientific, Abbott, Focused Ultra-sound Foundation, NIH-NeuroBlueprint MedTech 5U54EB033650 and NIH R18EB036591. She is the medical advisor for Aim Medical Robotics and has stock equity. She is a medical advisor for Myndtec and has stock equity. Dr. Telkes has received grant support from NIH R00NS119672, Florida Atlantic University COECS/I-SENSE, and University of Arizona Eighteenth Mile TRIF Funding. Dr. Goudman is a postdoctoral research fellow funded by the Research Foundation Flanders (FWO), Belgium (project number 1211425N). STIMULUS received independent research grants from Medtronic. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Reviewer disclosure

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Ethical declaration

The FAU Health Sciences IRB reviewed and determined this project to be exempt from federal regulation under 45 CFR 46.104. Under 45 CFR 46.104, certain projects can be exempt from the requirements of the Basic HHS Policy for Protection of Human Research Subjects. All methods were carried out in accordance with relevant guidelines and regulations.