Study design and participants

This descriptive-analytical study was carried out in two phases. In the first phase, the conceptual model of the social determinants of participants’ quality of sexual life was designed based on the literature review and using expert opinions (the Nominal Group Technique (NGT)). In a systematic review of the literature, the PubMed/Medline (NLM), Scopus, Web of Sciences, ProQuest, CENTRAL, and Embase databases and the Google Scholar search engine were searched between 2010 and 2022 using relevant keywords. Accordingly, PubMed and Embase were searched using the Mesh (Medical Subject Headings) and Emtree (Embase Subject Headings) search terms, respectively in order to identify the structural and intermediary social determinants of quality of sexual life based on the WHO’s SDH approach [12]. The preliminary conceptual model was then designed after analyzing two relevant systematic reviews [22, 23]. The next step was to assess the validity and reliability of the model using NGT by organizing a meeting with a group of reproductive health specialists and psychologists with expertise in sexuality and SDH (Fig. 1). In the second phase, the designed model was tested in a descriptive-analytical study of marginalized women.

Fig. 1

Conceptual model of social determinants of quality of sexual life in women

The sample was selected among all married women of reproductive age (15–49 years) who had been living in the marginalized areas of Kermanshah for at least one year and were able to talk and communicate with the researcher. The exclusion criteria included being pregnant, going through the first year postpartum, going through menopause, experiencing traumatic life events in the last six months (e.g., death of a loved one, husband’s betrayal, and illness), suffering from a physical or physiological illness affecting sexual function (e.g., diabetes, heart failure, resistant hypertension, thyroid disorders, and chronic kidney disease), suffering from gynecologic cancer or breast cancer, using drugs, stimulants or alcohol, experiencing sexual dysfunction (woman or her husband), having a history of using drugs that impair sexual function, and being unwilling to participate in the study.

Sample size

The sample size was calculated using pmsampsize (a module for determining sample size when developing prediction models) in Stata 14. Based on the literature review and expert opinions, 28 variables associated with women’s quality of sexual life were identified. According to the study of Yuksekol et al., the sample size was calculated to be 487, with Mean = 65.58, Sd = 26.77, and R-squared = 0.378 [24]. The final size of the study sample was then set at 560 by considering a loss to follow-up of 15% (due to the large number of questionnaires).

Procedure

Participants were selected between September 23, 2021 and April 19, 2023. To this end, the researcher first identified the health centers in the marginalized areas of Kermanshah, including 6 comprehensive health centers and 6 health centers. Then, the list of eligible women was obtained by visiting the centers and using Iran’s Integrated Health System (IHS), known as the “SIB System” (https://sib.kums.ac.ir), by applying filters for age group, marital status, history of living in the marginalized area, and history of pregnancy. SIB is an integrated system that electronically records and updates the health records of all Iranian. The total number of eligible marginalized women was 1700, among whom 560 individuals were selected as the sample. The women were selected from each center using proportional allocation and random numbers available on the website “www.random.org”. In the next step, the researcher contacted the selected women by telephone and asked them to participate in the study. Women who agreed to participate in the study were asked to visit the respective health centers at a specific time to complete the questionnaires. Those who were not willing to participate were replaced with randomly selected people from the aforementioned list. Some of the participants were illiterate and had a low level of education; thus, the researcher completed the questionnaires by interviewing them. Participants completed the questionnaires in approximately 45 to 60 min (they were free to take short breaks while completing the questionnaires). Written informed consent was obtained from all participants prior to the interviews.

Data collection toolsThe researcher’s demographic questionnaire

This questionnaire included age, spouse’s age, duration of marriage, frequency of spouse’s marriage (The total number of legal marriages the husband has entered), family planning method, number of children, and frequency of sexual intercourse per week.

Socioeconomic status scale

This 7-item questionnaire measures the socioeconomic status of people. The four domains of this instrument include educational qualifications (2 items), job status and income level (3 items), economic class (1 item), and housing status (1 item). The items are scored on a five-point Likert scale from 1 to 5 (Total score range: 7 to 35). Higher total scores indicate greater socioeconomic status [25]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.71).

Domestic violence questionnaire

This 26-item questionnaire measures the physical (10 items), sexual (5 items), and emotional (11 items) aspects of violence. The items are scored on a five-point Likert scale (Total score range: 0 to 104) [26]. The reliability of the instrument for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0. 88).

Depression, anxiety and stress scale (DASS-21)

Stress (7 items), depression (7 items), and anxiety (7 items) are the three subscales of this 21-item scale. The items are scored on a four-point Likert scale from 0 to 3. Total score is calculated by adding the scores given to all subscales. The minimum and maximum scores for each subscale are 0 and 21, respectively. The total score ranges between 0 and 63 [27]. The reliability of the questionnaire for the target population was confirmed using internal consistency method and Cronbach’s alpha value was 0.78 for the depression subscale, 0.74 for anxiety and 0.77 for stress on marginalized women.

Unhealthy behavior checklist

This checklist was designed by Kiani et al. in Iran. The questionnaire has 24 items and two sections, namely violence and spouse’s addiction (addiction to smoking, alcohol, and drugs). The scores given to the both sections are added and presented as a percentage [28]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.84).

General health

This questionnaire includes 28 questions in four subscales of physical symptoms, anxiety symptoms and sleep disorders, social functioning and depression symptoms. The items are scored on a four-point Likert scale from 0 to 3. In each of the subscales, a score of 7 or higher and a total score of 23 or higher indicates disease symptoms [29]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.76).

Access to health services questionnaire

This 9-item questionnaire was developed by Kiani et al. in Iran to measure the access of people to health services. The items are scored on a four-point Likert scale from 0 to 3. Total score is calculated by adding the scores given to all items (Total score range: 0 to 27) [30]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.74).

Perceived social support

To measure social support, the Multidimensional scale of Perceived Social Support (MPSS), Zimmet et al., was used. This tool has 12 items on a Likert scale of 0 to 7. Total score ranges from 12 to 84 [31]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.87).

Spiritual health

Polotzin and Ellison spiritual health questionnaire with 20 questions was used in this study. This tool has 20 items on a Likert scale of 1 to 6. Total score ranges from 20 to 120 [32]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.92).

Marital satisfaction scale

The short form of Enrich marital satisfaction scale was used to evaluate participants’ satisfaction with their married lives. The 10 items of this questionnaire are scored on a five-point Likert scale from strongly disagree (score 1) to strongly agree (score 5). Total score ranges from 10 to 50 [33]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.93).

Self-esteem

Rosenberg’s self-esteem questionnaire was used to evaluate self-esteem. This questionnaire is a standard measure of 10 positive and negative questions that includes people’s real feelings about each question in a four-point Likert scale from 0 to 3. Total score ranges from 0 to 30 [34]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.72).

Female genital self-image scale

Assessment of women’s satisfaction and beliefs about their genitals was done using FGSIS-7. This tool consists of seven questions with a four-point Likert scale from completely agree to completely disagree. The overall score of the tool ranges from 7 to 28 points [35]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.71).

Sexual Quality of Life-Female (SQOL-F)

The items of this 18-item instrument are scored on a six-point Likert scale. Items 1, 5, 9, 13, and 18 are scored inversely, and total score ranges from 18 to 108. The four domains of SQOL-F include psychosexual feelings (item 2, 3, 7, 8, 10, 16, and 17), sexual and relationship satisfaction (item 1, 5, 9, 13, and 18), self-worthlessness (item 4, 6, and 15), and sexual repression (item 11, 12, and 14). Higher scores indicate greater quality of sexual life [36]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0. 83). The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.83).

Female sexual function index (FSFI)

Rosen et al. designed this 19-item questionnaire to evaluate women’s sexual function. Total score is determined by adding the scores given to the 6 subscales of FSFI including desire, subjective arousal, lubrication, orgasm, sexual satisfaction, and pain. The maximum score for each subscale and the whole instrument is 6 and 36, respectively. Higher scores imply greater sexual function of women [37]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.71).

International physical activity questionnaires (IPAQ)

The 7-item version of IPAQ collects data on the various types of physical activities that people perform on a daily basis (performing activities at home and at work, walking from one place to another, engaging in recreational activities, and doing exercises). The Metabolic Equivalent Task (MET) values of “3.3”, “4”, and “8” were assigned to “walking”, “moderate activities”, and “vigorous activities”, respectively. The MET values were then multiplied by the duration of the activity in minutes and again by the number of days the activity was undertaken [38]. The reliability of the questionnaire for the target population was confirmed using internal consistency method (Cronbach’s alpha value = 0.69).

Data analysis

The data were analyzed in SPSS 24 to determine the demographic characteristics of participants and to examine the correlations among the research variables. In addition, the conceptual model was assessed in LISREL 8 using path analysis. The goodness of fit indices, including the chi-square statistic to the degree of freedom (χ²/df < 3), the norm fit index (NFI > 0.9), the comparative fit index (CFI > 0.9), the goodness of fit index (GFI > 0.9), and the root mean square error of approximation (RMSEA ≤ 0.05) were used to assess goodness of fit of the model. T-values greater than 1.96 were considered significant [39].