COPD burden on sexual well-being

Online survey was assessed from March to May 2019 through the association Santé Respiratoire France (https://​sante-respiratoire.​com). Four pulmonologists, one sexologist, one expert patient and one psychiatrist developed the questionnaire. The following data were collected at inclusion: demographic characteristics (age, gender, marital or relationship status, occupation, living area: urban, suburban, rural), smoking history, breathlessness and exacerbations etc.).

A detailed description of the questionnaires is available in Additional file 1.

We carried out a typological analysis to split the studied population into distinct homogeneous groups depending on their responses to the various questionnaires. We conducted latent class and profile analyses of CODP burden and respiratory assistance (dyspnea, fatigue) and psychological symptoms (depression and anxiety) to identify distinct classes (subgroups) of symptom profiles [20]. Patients were assigned a probability of being in each of the identified classes with the goal of creating a model that uniquely assigned a subject to a given class (e.g. Pr(ClassA) = 1.0; Pr(ClassB) = 0.0), or at minimum, provided a distinctively high probability to a given class versus all others (e.g. 0.95 versus 0.05). Model fit was evaluated using information criteria fit indices (Bayesian Information Criterion, BIC and Akaike’s Information Criterion, AIC); and low values indicate model parsimony [21]. We also used other criteria to identify a meaningful fit of model and data, and these included class interpretability (the extent to which additional classes provided unique information), class prevalence (preferring classes with at least 2% of the sample for improved replicability), and entropy (a measure of classification based on posterior probability values, with higher values representing better classification). We used analysis of variance and Chi-Square statistical analysis, which provides analysis of dependencies between data, establishing a hierarchical relationship, and displaying them in a two-dimensional graphical form. The two main factorial axes identified were the psychological state (ranging from good mental health to worsened mental health) and the need for respiratory assistance (starting from no assistance to respiratory assistance such as oxygen or non-invasive ventilation).

The calculation of the sample was based on the sampling strategy without experimental treatment. The sample size was calculated using the following formula: minimal sample size to obtain significant results for an event and a fixed level of risk: n = 100 X t² × p × (1-p) / m²n, with t: Confidence level, the standard value for the 95% confidence level will be 1.96, p: estimated proportion of the population with the characteristic of interest, m: Margin of error (usually set at 5%). Previous studies indicate that sexual dissatisfaction affects between 67.7% and 81% of COPD patients. Thus, for an event with a 60% probability of occurrence, and taking a 95% confidence level and a 5% margin of error, the sample size should be: n = 100 × 1.96² × 0.4 × 0.6 / 0.05² = 368.79. We should have at least 369 subjects and we included more than 700. Qualitative variables were described in numbers and percentages. Quantitative variables were described by mean and standard deviation or median and inter-quartile range. Missing data are reported in the tables. Differences in clinical characteristics were assessed using chi-square tests or Fisher’s exact tests, as appropriate, for discrete variables, and Wilcoxon or Kruskal–Wallis tests for quantitative variables. Relationships between different variables were examined with the Pearson correlation test. A value of p ≤ 0.05 was considered statistically significant.

The analysis of the results was carried out with Graph Pad Prism® statistical software.

Seven hundred and fifty one patients responded to the survey. Participants were predominantly women (51% of patients) with a mean age of 61 years old and 62% of them lived as a couple. The majority of the participants were retired (70%). Every COPD grade was represented and one third of responders received oxygen. Only a quarter of the responders lived in an urban setting. Clinical characteristics are reported in Table 1.

Forty percent had no sexual activity among which, 19% lived as a couple. From the remaining 60%, 74% had to change their rhythm, 64% their frequency and 64% their position. Two thirds of the participants were dissatisfied with their current and expected sexual function and 13% declared that they used specific medications to improve their sexual performance. Forty three percent used routinely (17%) or occasionally (26%) a short-acting bronchodilator before sexual activity. Expressed reasons of sexual dissatisfaction were breathlessness, fatigue and physical problems. Men were more often dissatisfied with their current and expected sexual function (66% of men vs 53% of women, p = 0.0002), however it did not translate into more medication, especially bronchodilators (which respectively concerned 28% of men and 22% of women). Besides, the level of sexual dissatisfaction (assessed by the following evaluation: “improved, stable, worsened”) was unrelated to age (p = 0.391) or gender (p = 0.92, Table 2).

ASEX score revealed an altered sexual appetite: low for 49% of the subjects and absent for 19%. Sexual desire was compromised in 44% of the subjects and 16% lost all desire. However, the deterioration of sexual appetite or sexual desire, assessed with ASEX questionnaire, is not linked to airflow obstruction (Fig. 1f, g). Subjects with COPD had to make adaptations to compensate for their diminishing activity tolerance, increasing dyspnea, role and sexual self-image, anxiety, and increased need for communication. Self-reported libido loss and reduced sexual performance worsened over time for 67% of the responders.

Based on the cut-off of 22 for VQ11 score, quality of life was altered among responders with a mean score of 34.8. The increase in airflow obstruction severity was inversely proportional to the quality of life, as assessed with VQ11 augmentation (Fig. 1a). Impaired quality of life (defined as VQ11 ≥ 22) is significantly more frequent among responders with a higher airflow obstruction (Fig. 1b). Specific underscores, including psychological, functional and relational components of quality of life, indicate that every aspect of life is impaired and an increasing with airflow obstruction further impairs the quality of life (Fig. 1c-e).

Mental disorders such as anxiety and depression were associated with COPD severity in the responders. A mean score of 4.6 on the FSS scale translates into a depressive tendency in patients and FSS mean score increases with airflow obstruction severity, as shown in Fig. 2a. Similarly, HAD scale found high proportions of patients with suspected anxiety or depression with a mean score of 10 and 8 for HAD-A and HAD-D respectively. HAD mean score for depression underscore increased with airflow severity (Fig. 2c) but the same was not reflected in HAD mean score for anxiety (Fig. 2b). However, both HAD scores for anxiety and depression increased when responders had at least two episodes of bronchitis in the previous year (Fig. 2d, e).

Mental disorders are also associated with sexual dissatisfaction. Higher depression as assessed by HAD underscore for depression was strongly indicative of decreased sexual appetite and to a lesser extent anxiety as assessed with HAD underscore for anxiety (Fig. 2f, g). Inversely, a higher impairment of sexual desire caused an increase in depression and anxiety (Fig. 2h, i).

Discussion of sexual issues remains a cultural taboo. Ninety percent of the responders declared that sexual dysfunction had never been discussed by doctors (general practitioners or pulmonologists) and only 6% had a specialized consultation whereas 36% of them declared that they would have liked a specialized consultation with a sexologist. Twenty one percent would be willing for an online consultation.

We carried out typological analyses to split the population studied into distinct homogeneous groups depending on their responses to the different questionnaires. This allowed us to combine the individuals into distinct homogeneous groups. We were able to isolate five behavioral profiles: Group A Preoccupied (22%), Group B Discouraged (19%), Group C Resigned (23%), Group D Naive (12%) and Group E Familiarized (24%). The map in Fig. 3 provides a pictorial representation of these 5 profiles.

The patients in group A were predominantly female, active, less than 60 years old, with recently diagnosed and non-severe COPD. These patients had a moderate physical impact of COPD, suffering from symptoms of depression and anxiety.

The patients in group B were also predominantly female, but polymorbid, suffering from severe COPD requiring respiratory assistance. They can no longer, or with difficulty, maintain a sexual life, and have important psychological symptoms leading to social isolation. The physical impact of COPD, but above all psychological and relational impact within this subgroup was clearly evident (extreme right-hand side of the graph).

The patients in group C were older than 70, with a long history of severe COPD, requiring daily ventilation. The degradation of their emotional and sexual life could be attributed to inactivity, breathlessness and substantial fatigue. They suffered mainly from physical symptoms, which explain their position at the top of the graph. However, they displayed only minor psychological symptoms.

The patients in group D were young and active, with recently diagnosed, non-severe COPD not requiring any respiratory assistance. Their sexual life was unaffected and for the disease did not require any life modification. Their social and emotional lives were normal. This group is situated at the lower left quadrant of the graph, displaying low limitations on the two axes.

Finally, the patients in group E were elderly, suffered from a longstanding moderate COPD. They remained active despite breathlessness. They were not worried or anxious, and tried to adapt according to their respiratory capacity. These were patients with physical symptoms largely, albeit of a moderate nature.