Sexual health status and vitamin D in primary care patients from the Middle East and Sweden

Low vitamin D is more common in Middle Eastern-born patients (Middle Easterners) than in Swedish-born patients of northern European descent (Swedes), and the condition might affect sexual health. The present study (i) investigated primary care (PC) patients (ii) compared self-reported sexual health in Swedes and Middle Easterners; (iii) analysed differences within and between the groups and (iv) analysed differences in vitamin D levels between the groups.

Sexuality is an essential part of being human and is important for quality of life (1). Impaired sexual function is a clinical condition (1)(2)(3). Loss of sexual desire is a primary disorder (1)(2)(3), and is not caused by other problems such as erectile dysfunction, dyspareunia or other diseases. Hormonal changes in females and males due to aging (2) and psychosocial factors (3) may play a role in loss of sexual desire.
Sexual disorders occur as a consequence of obesity in males of all ages (4).
Studies of sexual dysfunction in the UK reported a higher prevalence of the disorder in males of south Asian descent than in males of European descent in the UK (13). Similarly, another study in London concluded that those, who sought care for dysfunction, came mainly from Asia, the Middle East, or North Africa (14). A Swedish study of patients with type-2 diabetes found no differences in sexual dysfunction prevalence. That said, the study reported higher prevalence of decreased sexual desire (self-reported) in Assyrians than in Swedes (15).
Some evidence suggests that vitamin D can affect sexual health. A recent clinical trial found that vitamin D treatment in males with vitamin D de ciency could increase testosterone levels and improve erectile function (16). But the study included males aged 35 to 64 with normal testosterone levels at baseline. A pilot study of young females found a correlation between vitamin D de ciency and lower self-rated sexual health as evaluated by the Female Sexual Function Index -particularly in the areas of sexual desire, orgasm dysfunction, and sexual satisfaction (17).
Vitamin D de ciency prevalence varies between immigrants to Europe and native-born Europeans. A German study reported that 74% of Turkish immigrants (male and female) had a vitamin D de ciencycompared to 33% of native Germans (de ned as 25(OH)D of < 30 nmol/L). De ciency prevalence was somewhat higher in Turkish females, who wore ethnic skin-covering clothing, than females who wore western clothing (86% vs. 69%) (18).
While vitamin D variation between native Europeans and Middle Eastern immigrants was found, no data or analyses indicate whether this variation is associated with differences in sexual health. If this association exists, it might imply that immigrants seeking care for sexual dysfunction should be tested, and when appropriate, treated for vitamin D de ciency. So the aims of this study were to (i) determine if a difference in sexual health exists between Swedes and Middle Easterners (self-reporting, primary care patients) and (ii) determine if any existing differences are associated with vitamin D variation, age, sex, body mass index (BMI), and waist circumference.

Study population
Data used in the present study came from: (i) "Screening and treatment of prediabetes in primary health care: a pilot study" and (ii) "Health problems and vitamin D status in primary health care patients from different ethnic groups". The former (4-D diabetes study; 2013 -2015) was done at the Flemingsberg and Jakobsberg Primary Health Care Centres in Stockholm (PHCC). The latter (vitamin D study; 2015 -2016) was done at the Flemingsberg PHCC.
Inclusion criteria for both studies were (i) being born in Sweden (persons born in Sweden whose parents were also born in Sweden and are of northern European descent) and (ii) being born in the Middle East, Africa, or Asia (an immigrant in Sweden). The 4-D study excluded patients with a diabetes diagnosis.
All patients, who consecutively sought primary care during the relevant periods and met the inclusion criteria, received a verbal invitation to participate in the studies. In total, 929 patients were recruited to both studies at the Flemingsberg PHCC (n=394 in the 4-D study and n=99 in the vitamin D study) and at the Jakobsberg PHCC (n=436 in 4-D study). Patients (n=23), who participated in both studies at the Flemingsberg PHCC, were only counted once. So in total, 903 patients participated in both studies. Firstly, the patients responded to a questionnaire about their health, and then they provided blood samples drawn at the Karolinska University Hospital.
The present study included patients of northern European descent who were born in Sweden and whose parents also were born in Sweden (n=297) and patients of Middle Eastern descent born in Turkey, Iran, or Iraq (n=225), namely, 522 patients in total.

Independent variable
Origin categorized as Swedes and Middle Easterners; see above.

Dependent variables
Overall self-reported sexual health and: sexual function based on "Can you complete the act of intercourse?" -item de ned as ability to have intercourse (yes/no); sexual desire based on "Do you have any sexual desire?" -item dichotomised as either present or absent (yes/no); and sex life dissatisfaction: "How satis ed are you with your sex life?" -item responses on a ve-point scale: very satis ed, somewhat satis ed, neutral, somewhat dissatis ed, very dissatis ed.
The present study selected 25(OH)D because (i) it is the metabolite of vitamin D that is used for determining a person's vitamin D status; (ii) it has a long half-life, and (iii) its production by the liver is not signi cantly regulated -versus production of 1,25(OH 2 )D by the kidneys (19, 20).

Statistical methods
Differences in normally distributed variables between Swedes and Middle Easterners were determined with a Student's t-test. Differences in variables that were not normally distributed were determined with a Mann-Whitney U-test. Categorical variables were gender, sexual function, sexual desire, and sex life satisfaction. Statistical signi cance -of differences in these variables between Swedes and Middle Easterners -was determined with a chi-square test. Logistic regression revealed the odds ratios between Swedes and Middle Easterners and separately, between females and males in each group. Stepwise adjustment for each explanatory variable was presented in models; each model was added stepwise to the previous model. A p-value of less than 0.05 was considered statistically signi cant. STATA 14 software (StataCorp LP, Texas) was used for all statistical analyses.

Demographic characteristics
The present study had 522 patients 297 Swedes (56.9%) and 225 Middle Easterners (43.1%); see Table 1. Females were in the majority among Swedes (58.3%) and Middle Easterners (59.6%). Middle Easterners were signi cantly younger with than Swedes with (mean ages 45.3 and 55.0, respectively). Statistically signi cant differences between the two groups were found in median 25(OH)D levels of 67.4 nmol/L in the Swedes and 37.0 nmol/L in the Middle Easterners. Reporting of sexual dysfunction was signi cantly higher among Swedes.
No statistically signi cant differences occurred regarding sexual satisfaction and sexual desire between the groups. The most striking difference was in vitamin D level: 227 Swedes (81.6%) and 66 Middle Easterners (35.5%) had a normal 25(OH)D level of ≥ 50 nmol/L (p<0.0001). Female Middle Easterners reported signi cantly more sex life dissatisfaction than male Middle Easterners (67.0% versus 33.0%, respectively). Table 3 displays difference in sexual dysfunction in Swedes (with Middle Easterners as reference). The crude OR of reporting sexual dysfunction in Swedes patients was 1.70 (95% CI 1.15-2.50). The adjustment for age and sex increased the odds ratios to 2.14. Stepwise inclusion in the logistic regressions models of waist circumference and sexual desire slightly change the odds ratios (2.17 and 2.11). In the nal model with inclusion of the vitamin D level covariate, the odds ratios changed to nonsigni cant, OR=1.50 (95% CI 0.92-3.23). Age, waist circumference, and vitamin D levels were signi cant covariates. Patients who had adequate vitamin D levels had 19 times higher odds of reporting sexual dysfunction than those with vitamin D levels under 25 nmol/L (OR=19.35, 95% CI 8.50-43.90).

Logistic regression
Because gender differences occurred regarding insu cient sexual desire, more statistical analyses were done to explore possible covariant in uences. Table 4 displays the crude odds ratio for female Swedes (with male Swedes as reference). Female Swedes had a crude odds ratio of 1.71 for insu cient sexual desire; it rose to 1.87 after adjustment for age. So females had the probability of reporting insu cient sexual desire 87% more times than male Swedes. Adjustment for differences in age and waist circumference and sexual function increased the odds ratios from 1.87 to 2.40. In the nal model, with inclusion of vitamin D levels, the odds ratios was almost the same (2.17) and still signi cant.
Results of logistic regression analyses of differences in reporting sexual dysfunction between female Swedes with male Swedes showed that in the crude model, male Swedes had 57% higher odds for having sexual desire -compared to female Swedes. In the age-adjusted model, the odds ratios decreased to nonsigni cant levels (not shown in table). Table 5 displays results of the logistic regression analyses for reporting sex life dissatisfaction among female Middle Easterners -compared to male Middle Easterners. The odds ratio for females was more than twice that for males in the crude model (OR=2.07). When differences in vitamin D levels were accounted for in the nal model, the odds ratio for reporting sex life dissatisfaction in females decreased and was no longer signi cant. So differences in vitamin D levels could explain gender differences in reporting sex life dissatisfaction. The sexual desire covariate was the only signi cant one: regardless of gender, those who reported insu cient sexual desire had 5 times higher odds for reporting sex life dissatisfaction than those who did not report insu cient sexual desire.

Discussion
The present study investigated primary care patients (Swedes and Middle Easterners). It analysed differences within and between the groups regarding (i) sexual health factors and (ii) differences in vitamin D levels.
One-third of all patients reported sexual dysfunction. Compared to Swedes, Middle Easterners had signi cantly lower sexual dysfunction prevalence and odds of having sexual dysfunction. After adjustment for all covariates (including vitamin D), differences between the groups changed to nonsigni cant. This suggests that vitamin D status might help explain the observed difference in probability regarding sexual dysfunction. No signi cant differences existed between Swedes and Middle Easterners regarding self-reported sexual desire or sex life satisfaction.
Results of the present study were unexpected regarding reporting worsen sexual life in Swedish-born than in Middle Eastern-born patients and not aligned with the results of other studies. Malavige et al reported no signi cant difference in erectile dysfunction (ED) prevalence between south Asians and Europeans. And premature ejaculation was signi cantly more common in south Asian males. That said, the present study and this study may not be comparable due to differences in population and methods (13). Another study of males from the Middle East and south Asia reported that most patients who sought care at a clinic in London were of north African, Middle Eastern, or south Asian descent (14). Both studies involved males only, which is not similar to the present study that included males and females.
But a Swedish study with a sample comparable to the present study reported ndings that were not aligned with ndings in the present study either. That particular Swedish study found a higher prevalence of decreased sexual desire in Assyrians than in Swedes -and no differences in sexual function. Although the study was methodologically similar to the present study, it only included patients with type-2 diabetes (15).
Other studies reported higher prevalence of vitamin D de ciency in Middle Easterners living in Europe than in native Europeans (18, 21), and the present study con rmed this nding. This isn't surprising because evidence in this area is strong. Skin colour is one (but not the only potential cause of this difference).
Moreno-Reyes et al reported that Moroccans and Turks residing in Brussels had higher vitamin D levels than residents from the Congo -even though the Congolese had darker skin colour (21). These researchers suggested that it might be due to factors such as socio-economic status and UVB light exposure rather than differences in diet, because few food items contain signi cant amounts of vitamin D. The present study only included Middle Easterners. So no vitamin D de ciency comparison could occur between Middle Easterners and other immigrant groups.
Another study with 1231 participants reported higher vitamin D de ciency prevalence in Turkish immigrants than in ethnic Germans -and even higher prevalence in Turkish females who totally covered their bodies (18). This nding suggests that limiting the amount of sunlight that reaches the skin (with clothing) can affect vitamin D levels (18). And while the present study did not investigate clothing, the populations are comparable, and the present study yielded similar result.
The present study also found signi cant gender differences regarding insu cient sexual desire and sexual function among Swedes. Age explained differences between genders (Swedish) in those who reported sexual dysfunction. Covariates could not explain gender differences regarding insu cient sexual desire among Swedes nor could they explain the probability of reporting insu cient sexual desire.
The aforementioned discrepancies call for further investigation. Explanations could be found in deeper investigation of hormonal disturbances or mental health. But among Middle Easterners, those with lower vitamin D levels had signi cantly lower risk for reporting insu cient sexual desire than those with normal vitamin D levels. This was unexpected but might be explained by perceptions about sex life and mental health rather the physiological pathology. Further, deeper investigations of the sexual desire among Middle Easterners are also warranted.
A growing number of studies reported age-related increases in sexual dysfunction in males -largely attributable to performance disturbances due to atherosclerosis (8,9,22). Consequently in the present study, age was expected to be a signi cant covariate.
Multimorbidity and polypharmacy in older persons might explain this -as might psychological issues. In addition, the present study's population had risks for developing type-2 diabetes -a disease associated with fatigue, depression, and other health conditions that affect sex life.

Strengths and limitations
The material used in the present study consists of data that was the rst of its kind to be gathered in Sweden. The aforementioned 4-D diabetes and vitamin D studies were the rst to compare sexual health factors among Swedes and Middle Easterners. Covering so many factors can be a strength -if a general overview of health is the aim. But it may also be a weakness, because the questionnaire only brie y dealt with sexual health.
Results from the present study imply that greater prevalence of low vitamin D levels among Middle Easterners could be associated with higher risk for sex life dissatisfaction. A Deleskog et al study suggested that vitamin D levels can trigger sex life dissatisfaction -a common complaint among patients with type-2 diabetes. Not unlike patients in the Deleskog study, patients in the present study had higher risk for developing type-2 diabetes. And this suggests that they should be tested for vitamin D levels -as a proxy for type-2 diabetes within 10 years -and be subscribed vitamin D for type-2 diabetes prevention, which the Deleskog study recommended (23). The Deleskog study included non-immigrant population resettled in Stockholm and the results were that those with prediabetes had lower vitamin D levels and higher diabetes risk as measured after 10 years. However, because this was a cross-sectional study, it was not possible to determine causality, i.e., whether the difference in vitamin D levels caused the difference in reporting declined sexual health, in general, and sex life dissatisfaction, in particular. Instead, causes not included in the statistical analysis, such as differences in diet, health conditions, amount of exercise, or measurement of vitamin D levels during different seasons (winter or summer) may explain the ndings.
The previously published studies of patients on haemodialysis concluded that treatment with vitamin D did not improve sexual function in patients with vitamin D de ciency (24, 25). But this could be because of the patients' other medical condition. One study investigated treatment of 102 healthy middle-aged males with vitamin D de ciency and ED with high dosage of ergocalciferol, a form of vitamin D, over at least 12 months. The men's ED improved, their testosterone increased, and their HbA1c decreased (17).
The researchers also observed a small decrease in the men's BMI (16). Clearly, knowledge of this area is insu cient and further studies on effects of treatment with different forms of vitamin D are highly needed.
Another limitation of the present study was type of questions asked. Because of the nature of the two studies from which the data for this study were collected (they were general overviews of the participants' health), the questionnaire only brie y touched on sexual health with three issues: ability, desire, and satisfaction. For example, impaired sexual function consist of one single yes/no question without detailed categorization into its speci c aetiologies (e.g., ED, dyspareunia). Fuller investigation of the topic would require additional questions.
The present study included patients from the countries of origin of large immigrant groups in Sweden.
This makes results of the study readily applicable to a large part of the Swedish immigrant population. But the study sample may not be representative of immigrants who came to Sweden very recentlymany of whom come from Syria and Afghanistan. Although it may be plausible to assume that results of the present study may be applicable to immigrants from Syria (because of geographic proximity of Syria to countries included in the present study), it is less likely that they are applicable to immigrants from Afghanistan, due to geographic distance and perhaps cultural and religious differences.
The skewed sex distribution in the present study's population (58.8% female) may make the results more applicable to females than males. It also makes it harder to compare the present study's results to those of other studies, because many previous studies investigated ED and vitamin D. That said, gender distribution of the Swedes and the immigrants did not vary signi cantly in present study; consequently, the skewed distribution did not affect comparison between these two groups -as seen when gender was add in the logistic regression.
Some factors may render the present study's population as not representative of the entire population of Sweden because it is based on primary care patients. The present study lacks information regarding patients who refused to participate in the aforementioned 4-D diabetes and vitamin D studies. Only patients seeking care at a primary care centre were included, which means that people who were healthy or for some other reason did not seek primary health care were not included in the studies. So compared to the general population, the present study's population may contain a larger proportion of people with chronic health conditions that can affect sexual health. Recruitment occurred in two areas outside Stockholm with high immigrant populations, which might make the results more di cult to apply to the population of other areas in Sweden. Recruitment at multiple geographic locations -and not limited to patients -is warranted for future studies.
Although the results of the present study may not be applied clinically because the study did not established any kind of causality, it provides ideas for future clinical studies that could investigate possible vitamin D treatment effects. The present study lacked data on whether or not participants in the 4-D diabetes study were treated with vitamin D supplements. If found effective in such studies, the vitamin D supplementation could possibly contribute to treatment for sexual health problems. It might also have positive effects on other aspects of health, because other symptoms are linked to vitamin D de ciency (26, 27).

Conclusions
The present study achieved it objectives to investigate primary care patients (Swedes and Middle Easterners), compare their self-reported sexual health status, analyse differences within and between the groups, and analyse differences in vitamin D levels between the groups. The study revealed that vitamin D variation partly explains differences in sexual dysfunction between the groups and between genders within the groups. But the study cannot establish causality. Investigations of vitamin D treatment should be done to determine whether these results are clinically useful. The present study's ndings might be

Consent for publication
Not applicable.

Availability of data and materials
Either data or materials are publicly available as the sharing of the data with other researchers was not included in patient's written informed consent. However data are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study is a part of big 4D project-a strategic programme at Karolinska Institutet and Stockholm County Council. No grand no is available for the programme. The research project group consisted of senior researchers and senior advisors in Endocrinology and Psychiology was selected by both authorities. Neither authorities had in uence over the design and were not involved in execution of the 4D Diabetes study. The research project group independently designed the 4D Diabetes study and collection, analysis, and interpretation of data and in writing the manuscript.
Author's contribution MT was a major contributor in writing the manuscript and performed statistical analyses, KS participate in designing of the 4D study and participated in data collection, CGÖ participated in design of the study, interpretation of the results and critically editing of the manuscript, HS participated in design of the study and critically edit the manuscript. All authors read and approved the nal version of the manuscript.