Women with endometriosis have decreased health-related quality of life until late fertile age: a population-based cohort analysis at ages 31 and 46 years

Endometriosis is a chronic, oestrogen-dependent gynaecological disorder that is estimated to affect 5–10% of fertile-aged women, causing severe pelvic pain, dysmenorrhoea, dyspareunia, and infertility ( Shafrir et al. , 2018 ; Zondervan et al. , 2020 ; Taylor et al. , 2021 ). Increasing evidence suggests that endometriosis is a systemic disease, as it has been shown to have multifactorial effects throughout the body also leading to increased comorbidity risk ( Rossi et al. , 2021a , 2023 ; Taylor et al. , 2021 ). It is well known that endometriosis impairs health-related quality of life (HRQoL), with several HRQoL dimensions affected by this disease ( Jones et al. , 2002 ; Nnoaham et al. , 2011 ; De Graaff et al. , 2013 ; Facchin et al. , 2015 ; Della Corte et al. , 2020 ). A number of endometriosis-related symptoms, such as chronic pain and dyspareunia, have been found to have an independent negative effect on physical and mental components of HRQoL ( Souza et al. , 2011 ; Tripoli et al. , 2011 ). Previous studies have shown that endometriosis-related pelvic pain in particular, decreases HRQoL, most likely because it has consequences in daily life, including burdening sexual life and relationships ( Jones et al. , 2002 ; Souza et al. , 2011 ; Facchin et al. , 2015 ). Hormonal treatment has beneficial effects by improving endometriosis-associated pelvic pain and dysmenorrhoea ( Carvalho et al. , 2018 ; Taylor et al. , 2021 ). However, many women with endometriosis remain symptomatic and have impaired HRQoL even if properly treated in a tertiary care centre ( De Graaff et al. , 2013 ). Furthermore, according to a 10-year follow-up study, even though HRQoL improved after hysterectomy in women with endometriosis, the women never reached same HRQoL levels than those of non-endometriosis women ( Rahkola-Soisalo et al. , 2020 ). This may relate to the fact that endometriosis associates with several other health conditions further burdening women ( Rossi et al. , 2023 ). Diagnosing endometriosis is challenging, as women meet on average seven physicians before the disease is diagnosed; the diagnostic delay can be as long as 11 years from the onset of the symptoms ( Ballard et al. , 2006 ; Greene et al. , 2009 ; Nnoaham et al. , 2011 ; Taylor et al. , 2021 ). Endometriosis has been thought to alleviate towards late fertile age and after menopause due to the decrease in oestrogen levels. However, the pain-associated symptoms seem to persist in women with a history of endometriosis, as our previous data showed lower pain threshold and tolerance until late fertile age of 46 years ( Vuontisjärvi et al. , 2018 ). Given this, the aim of the present study was to assess HRQoL and associated factors until late fertile age 46 years in a population-based study setting. Furthermore, as the latest European Society of Human Reproduction and Embryology (ESHRE) endometriosis guideline no longer considers a surgical approach an imperative gold-standard procedure for the diagnosis ( Becker et al. , 2022 ), and as it has been reported that self-reported diagnosis is fairly accurate ( Shafrir et al. , 2021 ), our second aim was to investigate HRQoL in two subgroups of women: those self-reporting an endometriosis diagnosis without hospital contact and those with a hospital-set diagnosis and, most likely, effective and targeted management of the disease.

The baseline characteristics are shown in Table 1 . At age 31 years, women with endometriosis were more often underweight and less often obese than their non-endometriosis counterparts. At both ages, women with endometriosis suffered more often from infertility and depression and had fewer children. Surprisingly, in women with endometriosis, hormonal contraceptive use (current and ever) was less frequent when compared with women without endometriosis. At late fertile age, widespread pain and climacteric symptoms were more frequent in women with endometriosis when compared with the reference group. Characteristics of all women with endometriosis (ENDOall) and reference group at ages 31 and 46 years. At age 31 years, three different areas of pain were inquired, whereas at age 46 years, eight areas were inquired. The number of cases in each analysis may differ due to available cases for different variables. P -values in bold indicate statistical significance ( P  < 0.05). At both timepoints, the 15D total score was lower in women with endometriosis compared with women without, however, the differences in this total score were not clinically significant (they were smaller than 0.015) (age 31 years: 0.944 vs 0.951 [difference 0.007], P  = 0.039; age 46 years: 0.916 vs 0.924 [difference 0.008], P  = 0.03). In the single index score analysis, women with endometriosis had lower scores in the following domains at age 31 years: ‘sleeping’ (0.893 vs 0.915 [difference 0.022], P  = 0.03), ‘depression’ (0.871 vs 0.903 [difference 0.032], P  < 0.001), and ‘distress’ (0.887 vs 0.910 [difference 0.023], P  = 0.02); and at age 46 years: ‘sleeping’ (0.798 vs 0.823 [difference 0.025], P  = 0.006) when compared with non-endometriosis women ( Fig. 2 ). 15D health-related quality of life (HRQoL) showed impairment in women with endometriosis. 15D as a profile (single items connected) with single indices (mobility, vision, hearing, breathing, sleeping, eating, speech (communication), excretion, usual activities, mental function, discomfort and symptoms, depression, distress, vitality, and sexual activity) at ages 31 ( A ) and 46 years ( B ) in all women with endometriosis (ENDOall) and in the reference group. Mann–Whitney U test was used to analyse differences between the study populations in 15D total and single index scores. y -Axis shows the 15D mean score and x -axis shows the 15 different items as well as the sum of the items. * P  < 0.05, ** P <0.01, *** P <0.001. Women with endometriosis tended to be more often in the lowest quartile of 15D scores at fertile and late fertile age when compared with women without endometriosis (age 31 years: 29.0% vs 23.4%, P  = 0.06; age 46 years: 28.8% vs 24.1%, P  = 0.08, Fig. 3A ). The unadjusted analysis did not show association between endometriosis and the lowest HRQoL quartile (age 31 years: OR [95% CI] 1.33 [1.00–1.78]; age 46 years: OR [95% CI] 1.28 [0.98–1.66], Fig. 3B ). In the multivariate analysis, confounding analysis including lifestyle and socioeconomic factors (Model 1: BMI, education, and smoking), strengthened the association between low HRQoL and endometriosis at both time points showing increased risk for low HRQoL in endometriosis (age 31 years: OR [95% CI] 1.42 [1.06–1.91]; age 46 years: OR [95% CI] 1.42 [1.06–1.89], Fig. 3B ). The endometriosis-related health outcomes were tested in Model 2 (Model 2: depression, infertility, pain, and poor general health), and were shown to contribute low HRQoL at both time points (31: OR [95% CI] 1.24 [0.90–1.72]; 46: OR [95% CI] 1.10 [0.80–1.52]). The confounding factors shown to improve endometriosis were considered in Model 3 (Model 3: current use of hormonal contraceptives and parity) resulting in strengthening of the association between endometriosis and low HRQoL at age 31 years, but not at age 46 years (age 31 years: OR [95% CI] 1.40 [1.03–1.91]); age 46 years: OR [95% CI] 1.29 [0.94–1.77]). All above mentioned confounding factors as well as climacteric symptoms at age 46 years were considered in Model 4 and they contributed to low HRQoL at both time points (age 31 years: OR [95% CI] 1.31 [0.92–1.85]; age 46 years: OR [95% CI] 1.10 [0.74–1.66]) ( Fig. 3B ). Lowest quartile of health-related quality of life (HRQoL) measure 15D; i.e. worst HRQoL and change in 15D total score from age 31 years to age 46 years in all women with endometriosis (ENDOall) and the reference group. ( A ) The proportion of women with endometriosis (ENDOall) and the referent women belonging to the lowest quartile of 15D at ages 31 and 46 years. ( B ) Lowest quartile of HRQoL adjusted for different confounding factors at ages 31 and 46 years. Model 1: adjusted for BMI, education, and smoking, Model 2: adjusted for depression, infertility, pain, and poor general health, Model 3: adjusted for current use of hormonal contraceptives and parity, Model 4: adjusted for BMI, education, smoking, depression, infertility, pain, poor general health, current use of hormonal contraceptives, parity, and climacteric symptoms at age 46 years. OR, odds ratio. ( C ) Change in 15D total score from age 31 to 46 years in all women with endometriosis (ENDOall) and the reference group. y -Axis shows the 15D mean score and x -axis timepoints of ages 31 and 46 years. The change in 15D total score between the two time points for each study population was calculated using the paired-samples t -test; P -value <0.05 was considered statistically significant, and all tests were two-tailed. In women with endometriosis, as well as in the reference group, there was a significant impairment in HRQoL from age 31 years to age 46 years (endometriosis: from 0.944 to 0.915 [difference 0.029], P  < 0.001; non-endometriosis: from 0.951 to 0.924 [difference 0.027], P  < 0.001, respectively). The impairment was similar in both groups ( Fig. 3C ). The characteristics of the subgroup populations (ENDOsr and ENDOreg) are shown in Supplementary Table S1 . At age 31 years BMI, women in ENDOreg were less often overweight or obese when compared with the reference group. At both ages, women in ENDOsr and in ENDOreg suffered more often from infertility and depression when compared with women without endometriosis. Women in the ENDOreg group had fewer children and currently used hormonal contraceptives less frequently at both ages, when compared with women without endometriosis. At late fertile age, widespread pain and climacteric symptoms were more frequent in the ENDOsr and ENDOreg groups when compared with the reference group. ENDOsr and ENDOreg group reported poorer general health at age 31 years when compared with the non-endometriosis women, but not at age 46 years. In the ENDOsr group, the 15D total score was lower at both ages 31 and 46 years compared with the reference group (age 31 years: 0.922 vs 0.951 [difference 0.029], P  < 0.001; age 46 years: 0.902 vs 0.924 [difference 0.022], P =0.001) with significant impairment of ‘excretion’, ‘mental function’, ‘discomfort and symptoms’, ‘depression’, ‘distress’, ‘vitality’, and ‘sexual activity’ at age 31 years. At age 46 years, ‘hearing’, ‘breathing’, ‘sleeping’, ‘excretion’, ‘discomfort and symptoms’, and ‘vitality’ were affected in the ENDOsr group ( Supplementary Fig. S1A and B ). Women in the ENDOsr group also had more often a 15D total score in the lowest quartile compared with the reference group at both ages (48.6% vs 23.4%, P  < 0.001 and 36.7% vs 24.1%, P  = 0.006, respectively) ( Supplementary Fig. S2A ). At age 31 years, adjustment for confounding factors did not affect this finding ( Supplementary Fig. S2C ), whereas at age 46 years endometriosis-related factors (Model 2, depression, infertility, pain, and poor general health) appeared to have a contributing effect ( Supplementary Fig. S2C ). The ENDOreg group had 15D total scores similar to those of women without endometriosis at age 31 years (0.953 vs 0.951 [difference 0.002], P  = 0.72) and at age 46 years (0.958 vs 0.924 [difference 0.034], P  = 0.62). However, they had a significantly lower index score for ‘depression’ at age 31 years compared with the reference group ( Supplementary Fig. S1C ). There was no difference between the ENDOreg and the reference group at age 46 years in the 15D questionnaire, nor were the women in the ENDOreg group at risk for being in the lowest HRQoL quartile (at age 31 years: 21.1% vs 23.4%, P  = 0.52; at age 46 years: 25.0% vs 24.1%, P  = 0.80) ( Supplementary Fig. S2B ). Adjusting for the confounding factors did not affect the results ( Supplementary Fig. S2D ). Both ENDOsr and ENDOreg 15D total scores decreased from age 31 years to age 46 years ( Supplementary Fig. S3A ); however, this decrease was similar to that of reference group ( Supplementary Fig. S3B ).

This is a cross-sectional analysis of the prospective, population-based Northern Finland Birth Cohort 1966 (NFBC1966, University of Oulu 2020 ), which is a unique, population-based dataset comprising all expected births in 1966 in the Northern Finland area (live-born females n = 5889). Enrolment for this database began at the 24th gestational week, and data (postal questionnaire and/or clinical measurements) have been collected at ages 1, 14, 31, and 46 years ( Nordström et al. , 2022 ). This study utilized the two latter collections. The study population was formed by two approaches: (i) linking the birth cohort individuals to the Finnish Institute for Health and Welfare Care Register for Health Care (CRHC) to identify women with a hospital-set International Classification of Diseases (ICD) code for endometriosis (ICD-10 N80.1–N80.9, ICD-9 617.1–617.9, or ICD-8 625.30–625.39) from 1972 until 2020, n = 298 (first endometriosis diagnosis was set in 1986) and (ii) self-reported history of endometriosis diagnosis at age 46 years with the question ‘Have you ever been diagnosed with endometriosis by a physician?’, n = 283. Both self-reported and register-based endometriosis diagnosis was found for 162 women; hence, the total endometriosis population was n = 419 (ENDOall). A total of 121 had only self-reported a diagnosis without a CRHC ICD code (ENDOsr) and thus no tertiary care contact due to endometriosis. The 298 women who had an endometriosis diagnosis in the CRHC register formed the other subgroup (ENDOreg). The population has been validated earlier by our previous study ( Vuontisjärvi et al. , 2018 ). Subgroups enabled a closer evaluation of the effect of tertiary care contact on endometriosis symptoms as well as an evaluation of women who self-reported a former diagnosis of endometriosis. The women who did not have an ICD coding for endometriosis or who did not report a former diagnosis of endometriosis were considered the reference group, n = 3279. The study population has been described in more detail in our previous publications ( Vuontisjärvi et al. , 2018 ; Rossi et al. , 2021a , b , 2023 ). A flowchart for the study population selection is shown in Fig. 1 . Flowchart of data collection and study population forming cases and references. The final study population, i.e. the total endometriosis population (ENDOall), the subgroups, including self-reported endometriosis cases (ENDOsr) and cases with endometriosis diagnosis registered in the Care Register for Health Care (ENDOreg), and the reference group are marked in bold typing. HRQoL was assessed with the 15D instrument ( Sintonen and Pekurinen, 1993 ) that was included in the postal questionnaire sent to the cohort participants at ages 31 and 46 years. The response rates were 76% (n = 4523) and 65% (n = 3848), respectively. The 15D instrument is a generic, 15-dimensional, self-administered, standardized, and well-validated measure of HRQoL that can be used both as a profile and as a single-index score measure composed of the following dimensions: mobility, vision, hearing, breathing, sleeping, eating, speech (communication), excretion, usual activities, mental function, discomfort, and symptoms (e.g. pain, ache, nausea, itching, etc), depression, distress, vitality, and sexual activity. The questionnaire usually takes 5–10 min to complete ( Sintonen, 2001 ; Alanne et al. , 2015 ; Bourdel et al. , 2019 ). For each dimension, the respondent chooses one of five levels best describing their current state of health. The 15D score and the dimensional level values range from 0 to 1, score 1 indicating full health and 0 indicating death. These values are calculated from the health state descriptive system by using a set of population-based reference or utility weights ( Sintonen, 2001 ; Alanne et al. , 2015 ). For improvement or deterioration, the minimally important change, i.e. the clinically significant difference in HRQoL total score, is 0.015 in 15D. A change of 0.035 in the total score can be considered a large improvement or a deterioration in HRQoL ( Alanne et al. , 2015 ). The validity of 15D has been shown against well-known HRQoL tool SF-36 ( Richardson et al. , 2016 ). All available and relevant health and socioeconomic factors relating to endometriosis (tested and/or relevant according to the existing literature) were considered in the analyses. The questionnaires can be found online ( University of Oulu, 2020 ). The participants’ weight was measured with a digital scale that was calibrated regularly. Height was measured twice using a standard calibrated stadiometer. If measured data were not available (in 30.6% of the cohort population at age 31 years and 15.0% at age 46 years), self-reported values were used. There was no difference between the self-reported and clinically measured BMI values ( Ollila et al. , 2016 ). Women were categorized into BMI groups according to WHO criteria for normal weight (18.5–24.9 kg/m 2 ), overweight (25.0–29.9 kg/m 2 ), and obese (BMI ≥ 30.0 kg/m 2 ) (Report of a WHO consultation; World Health Organization, 2000 ). Data on education were derived from the questionnaires and stratified into three categories according to number of years of education: 9 years (primary), 10–12 years (secondary), and more than 12 years (tertiary). Smoking history and present smoking were inquired by questionnaires: ‘Have you ever smoked (yes/no)?’ and ‘Are you currently smoking (yes/no)?’ Smoking was then categorized as follows: never, former or occasional, and active. Alcohol use was inquired in the postal questionnaire and was categorized as abstinence, low-risk drinking (≤20 g/day), and high-risk drinking (>20 g/day). Depression was inquired at both time points as: ‘Have you ever been diagnosed with depression (yes/no)?’ Infertility was inquired at both time points as: ‘Have you ever suffered from infertility (yes/no)?’ At age 31 years, three pain sites (neck, shoulder, lower back) were inquired in the questionnaire. At age 46 years, eight pain sites (neck, shoulders, arms, wrists/hands/fingers, lower back, hip, knees, and ankles/feet) were inquired. The questionnaire included a question ‘How do you rate your current health status?’ at both time points, and the answers were dichotomized into good (very good/good) and poor (moderate/poor/very poor) which has been shown to be a valid measure to define a person’s general health ( Idler and Benyamini, 1997 ; Halford et al. , 2012 ). Current use of hormonal contraceptives was inquired at both ages: ‘Are you currently using any hormonal contraception?’ and at age 46 years, past use of hormonal contraceptives was inquired: ‘Have you ever used any hormonal contraception (yes/no)?’ Parity was inquired at ages 31 and 46 years as follows: ‘If you have been pregnant, how many deliveries have you had?’ According to the number reported, parity was stratified into three categories: 0, 1–2, and 3 or more deliveries. Climacteric symptoms were inquired at age 46 years: ‘Do you have climacteric symptoms (yes/no)?’ Statistical analyses were performed using IBM SPSS Statistics for Windows, version 29 (IBM SPSS Statistics for Windows, Version 29.0. Armonk, NY: IBM Corp. Released 2022, USA). The Pearson chi-square test was used to analyse the characteristics of the study population. The 15D was analysed as constant total score, and the different dimensions of the HRQoL questionnaire were analysed as single index scores. The Mann–Whitney U test was used to analyse differences between the study populations in 15D total and single index scores. The Pearson chi-square test was used to analyse the percent proportion of women in the lowest 15D quartile. The odds for the women with endometriosis being in the lowest 15D quartile was assessed with binary logistic regression analysis. The results are presented as odds ratio (OR) and 95% CI. First, we conducted univariate analysis for each confounding factor and then moved on to a multivariate approach. To further analyse possible contributing factors, four models were built. Model 1 (health-related factors) included BMI, education, and smoking; Model 2 (endometriosis-related factors) included depression, infertility, pain, and poor general health; Model 3 (treatment) included current use of hormonal contraceptives and parity and Model 4 (all factors) included BMI, education, smoking, depression, infertility, pain, poor general health, current use of hormonal contraceptives, parity, and at age 46 years climacteric symptoms. Finally, the change in 15D total score from age 31 to 46 years was calculated using the paired-samples t -test; P -value  < 0.05 was considered statistically significant, and all tests were two-tailed. This study follows the principles of the Declaration of Helsinki. All participants provided written informed consent to use NFBC1966 study data for research purposes and to combine the data with register data. The Ethics Committee of the Northern Ostrobothnia Hospital District approved the research (latest registration number: 94/2011) on 14 December 2011.

Previous studies have reported women with endometriosis having lower HRQoL at fertile age when compared with women without the disease ( Jones et al. , 2002 ; Sepulcri and Amaral, 2009 ; Tripoli et al. , 2011 ; Simoens et al. , 2012 ; De Graaff et al. , 2013 , 2016 ; Kalfas et al. , 2022 ). However, the literature is missing population-based studies, especially those including women at late fertile age, and thus the present data contribute significantly to the existing literature. The present results indicate that there is a risk for affected women to score on the lowest HRQoL quartile showing impairment in HRQoL items like ‘sleeping’, ‘depression’, and ‘distress’, yet the total score not exceeding the 0.015 difference that has been considered a clinically significant change ( Alanne et al. , 2015 ). Depression, infertility, pain, and poor general health contributed to lower HRQoL, whereas BMI, smoking, education, parity, hormonal treatments, and tertiary care contact may improve HRQoL in women suffering from endometriosis. The finding of impaired sleeping in women with endometriosis is in line with previous literature. Similar findings were reported in a study including 257 fertile-aged women with endometriosis and in two case-control studies conducted with women aged 32–34 years ( Facchin et al. , 2015 ; Nunes et al. , 2015 ; Leone Roberti Maggiore et al. , 2017 ). To our knowledge, our study is the first to report impaired sleeping in women with endometriosis in a population-based data setting reaching until late fertile age. This impairment seemed not to be related to climacteric symptoms, as the overall impaired HRQoL in endometriosis was not associated with climacteric status in our analysis. As sleep impairment was already present at age 31 years, the finding might relate to endometriosis-related symptoms like pain. Indeed, previous studies have linked pain with poor sleep quality, regardless of the underlying disease ( Nunes et al. , 2015 ; Irwin, 2019 ; de Souza et al. , 2023 ). On the other hand, non-restorative sleep has been shown to be predictive for the development of pain and exacerbation of existing pain ( Stroemel-Scheder et al. , 2020 ). Moreover, given that chronic inflammation is one of the main underlying pathophysiologic mechanisms that drive endometriosis, sleep disturbances resulting in abnormal inflammatory responses may also explain our results ( Bonavina and Taylor, 2022 ). Hence, impaired sleeping, pain, and inflammation can create a vicious cycle in women with endometriosis where one condition maintains and intensifies the other. Previous literature suggests that chronic pelvic pain (CPP) and mental health could have a greater effect on HRQoL than endometriosis per se ( Jones et al. , 2002 ; Delanerolle et al. , 2021 ). ENDOall and the subgroups had more often intensive pain compared with the reference group. On the other hand, HRQoL in the ENDOsr subgroup was impaired by endometriosis even after adjusting for pain. In contrast, at late fertile age, impaired HRQoL was contributed by depression, infertility, pain, and poor general health. This is a novel finding which has not been investigated before in a population-based setting. The finding is, in line with our previous study showing that endometriosis alters pain perception as well as pain intensity and troublesomeness still at age 46 years ( Vuontisjärvi et al. , 2018 ). In the present study, ‘depression’ and ‘distress’ were increased in women with endometriosis at age 31 years, but not at age 46 years. This finding is also in line with previous literature as far as fertile age is concerned. Endometriosis has been associated with mental distress ( De Graaff et al. , 2016 ; Delanerolle et al. , 2021 ; O’Hara et al. , 2021 ), although causality between endometriosis and psychological distress remains unsolved, as other factors, such as CPP, contribute to this finding ( Delanerolle et al. , 2021 ). Interestingly, ‘depression’ and ‘distress’ components were not increased at age 46 years. This is somewhat conflicting, as the women with endometriosis reported being diagnosed as having depression by a physician more often than their non-endometriosis counterparts at both ages 31 and 46 years. Moreover, our previous study on the same data found almost 2-fold odds for having an ICD code for mood disorders in women with endometriosis until age 50 years ( Rossi et al. , 2023 ). However, it must be noted that the 15D is not an instrument designed to specifically assess depression, which may explain the discrepancy. In contrast with previous literature, we did not find significant impairment of ‘sexual activity’ in women with endometriosis at ages 31 or 46 years. Endometriosis has been previously found to negatively impact sexual function ( Fairbanks et al. , 2017 ; Della Corte et al. , 2020 ; Norinho et al. , 2020 ). Indeed, pelvic pain and dyspareunia have been found to impair sexual function in women with endometriosis. It is possible that women at fertile age do not report lower sexual activity if they wish to conceive; thus, more targeted questionnaires should be used to assess their sex life satisfaction ( Elmerstig et al. , 2008 ; Fritzer et al. , 2013 ). However, this would not explain the results at age 46 years. The finding concerning late fertile age is also novel, as earlier studies have focused on fertile-aged women only ( Ferrero et al. , 2005 ; Fritzer et al. , 2013 ; Di Donato et al. , 2014 ; De Graaff et al. , 2016 ). This result brings a positive message for women with endometriosis but also raises the need to include other, more specific instruments to assess this area of life. Interestingly, in ENDOsr subgroup analysis, both ‘vitality’ and ‘discomfort and symptoms’ were significantly impaired. The association between endometriosis and vitality is still controversial. A large, prospective, multicentre study did not find vitality to be affected by endometriosis ( Nnoaham et al. , 2011 ), but in a smaller study, endometriosis was found to decrease vitality ( Friedl et al. , 2015 ). Discomfort and overall symptoms in women with endometriosis are scarcely mentioned in existing literature. Three Finnish studies reported endometriosis surgery or other treatment improving this domain, but one of the studies by Setälä et al. did not include subjects at their late fertile age ( Taipale et al. , 2009 ; Setälä et al. , 2012 ; Pynnä et al. , 2021 ). Hormonal treatment is the first-line treatment modality for endometriosis to relieve symptoms and to slow disease progression ( Becker et al. , 2022 ). In the current data, the usage of hormonal contraception was associated with better HRQoL, and the users were less often in the lowest HRQoL quartile. Surprisingly, in our data, women with endometriosis reported less frequent current use of hormonal contraception at both ages 31 and 46 years when compared with the reference women. There are various possible reasons for this surprising finding. The women might be trying to conceive, as the time to pregnancy has been shown to be longer for women with endometriosis ( Evans and Decherney, 2017 ; Bonavina and Taylor, 2022 ), or they might have been treated inadequately for endometriosis at the time of the questionnaires (in 1997 and 2012). Awareness of endometriosis has increased since, and today the common consensus underlines early and effective management of the disease. Supporting this hypothesis, at age 46 years, the women with endometriosis had an ‘ever use’ history for use of contraception similar to that of the women without the disease. Unfortunately, the question ‘ever use’ was not included in the 31-year postal questionnaire, and we were unable to evaluate this data before age 46 years. Our results suggest that self-report of endometriosis captures a population of women with more bothersome symptoms, as they present significantly lower 15D total scores (and especially in the domains of ‘excretion’, ‘mental function’, ‘discomfort and symptoms’, ‘depression’, ‘distress’, ‘vitality’, ‘hearing’, ‘breathing’, and ‘sleeping’) when compared with reference women. By contrast, the ENDOreg women seemed to have a better HRQoL, as they had 15D total scores similar to those of the reference group, except in the single index of ‘depression’, which was significantly lower in the ENDOreg women. This finding is interesting and may suggest that women in the ENDOreg group had incidental endometriosis diagnoses or that this group consisted of women with endometriosis who were treated comprehensively in tertiary care and benefited from the treatment. Four Finnish studies have shown that endometriosis surgery/hysterectomy improves HRQoL in women with endometriosis. One of the studies included fertile-aged women, other three studies included female population with mean age of 51.1, 45.3, and 49.6 years ( Taipale et al. , 2009 ; Setälä et al. , 2012 ; Rahkola-Soisalo et al. , 2020 ; Pynnä et al. , 2021 ). In contrast to our results, De Graaff et al. (2013) found that treatment in a tertiary care centre is ineffective in improving HRQoL in the long-run for fertile-aged women with endometriosis. There are several strengths in our study. The high novelty of the data is the population-based approach and assessments both at mid and late fertile age. Our vast cohort data also enabled consideration of various confounding factors related to endometriosis and HRQoL. The possibility of deriving endometriosis cases from hospital records and postal questionnaires enabled us to have a large population of women with endometriosis with a highly specific diagnosis. The population-based cohort setting minimized selection bias, as the cohort was not specifically established to study endometriosis. It should be noted that even though endometriosis-specific questionnaires are good for the evaluation of women with endometriosis, for example the Endometriosis Health Profile (EHP-30), they are not appropriate for use in the general population ( Bourdel et al. , 2019 ). A systematic review of QoL measures in patients with endometriosis found 15D to be accurately validated and reliable ( Bourdel et al. , 2019 ), and we have previously shown the instrument to be usable in the same study cohort targeting women with PCOS ( Karjula et al. , 2020 ). 15D questionnaire has been found to correlate well with the SF-36 questionnaire for assessing quality of life ( Richardson et al. , 2016 ). Moreover, the different items in the 15D instrument fit well with endometriosis symptomology as the instrument includes components such as ‘sexual activity’ and ‘excretion’. It must be considered, however, that our estimates may be underestimates given that an Australian study showed 15D giving smaller differences between public and patient utilities when compared with other QoL measures ( Richardson et al. , 2016 ). Self-report of endometriosis diagnosis may be considered a limitation. The diagnosis was inquired only at age 46 years, and the time of diagnosis assessment was not known. However, self-report of previous endometriosis diagnosis was verified in our previous study, which showed that almost 80% cases had been diagnosed in a hospital ( Vuontisjärvi et al. , 2018 ). Self-reported diagnosis has been used in previous studies and is at least moderately accurate, but accuracy improves if additional information is available ( Saha et al. , 2017 ; Shafrir et al. , 2021 ). In the sub-analyses, the size of endometriosis populations was somewhat limited. Moreover, in the ENDOreg group, we were not able to assess the type of intervention in the tertiary care centre and hence cannot draw conclusions on whether HRQoL improvement was due to the treatment that they might have received. Also, it is possible that after self-report of endometriosis and CRHC codes, there were still some women with endometriosis in the reference group. Such cases would, however, be diluted among the large reference group. This cohort population is mainly of Caucasian ethnic background, hence, the results might not be globally generalizable. The cohort setting allowed us to highlight associations between endometriosis and impaired HRQoL but did not allow us to show causalities. Here, we show that endometriosis impairs HRQoL, especially items like ‘sleeping’, ‘depression’, and ‘distress’ in an unselected population-based study warranting evaluation of these symptoms. Women with endometriosis should be offered care with holistic approach to increase their HRQoL. Overall, it is crucial to increase awareness of endometriosis among healthcare professionals as well as in society to shorten the time from symptom onset to diagnosis and to offer comprehensive treatment modalities.