Participant characteristics
Of the 1432 participants, 85% were male (n = 1213) and 54.7% were in active service at the time of data collection (n = 783; see Table 1). The majority of participants had been in regular service rather than reserve (81.5% vs 18.5%). A moderate proportion of participants met criteria for probable mental health problems on the psychometric measures, with 8.7% meeting criteria for likely PTSD (n = 124), 18.6% meeting criteria for alcohol misuse (n = 266), and 19.3% meeting criteria for CMD (n = 276). It should be noted that these are not prevalence rates or representative of mental disorder prevalence in the AF as participants were included if they self-reported mental health, stress or emotional problems.
Perceived stigma and barriers to care
Individuals meeting criteria for current probable PTSD and CMD were consistently more likely to endorse internalised stigma of mental illness, perceived stigma of mental health care/providers and difficulties with access to care compared to those who did not meet case criteria (Table 2, Additional file 1: Table S2). Odds ratios found for individuals with probable PTSD were greater than the odds ratios of those meeting criteria for likely CMD or alcohol misuse across all three scales; however, overlapping confidence intervals indicate these differences may not be statistically significant. Although, differences in effects sizes between those meeting probable PTSD criteria and individuals meeting likely alcohol misuse criteria are approaching significance for the internalised stigma of mental illness and perceived stigma of mental health care/provider scales.
The relationship between probable mental disorders and endorsement of the access to mental health services as well as internalised and perceived stigma items is described below.
Access to mental health services
In terms of practical barriers to accessing treatment, compared to those who did not meet criteria for a probable mental disorder, those meeting criteria for PTSD were significantly more likely to report access problems (AOR 1.82; 95% CI 1.38; 2.40; Table 2). In particular, individuals meeting case criteria for probable PTSD significantly more commonly endorsed concerns that ‘it would be difficult to schedule an appointment’ (32.5% agreement in cases versus 12.5% non-cases, Additional file 1: Table S1) and it would be difficult to get time off work for treatment compared to those not meeting probable PTSD criteria (29.2% PTSD cases agree/strongly agree vs 15.8% of non-cases).
Access-related concerns were also significantly more commonly reported by those meeting criteria for probable CMD compared to those who did not meet case criteria (AOR 1.56; 95% CI 1.26; 1.94) (Table 2). Similar to PTSD cases, those meeting criteria for probable CMD were more likely to report that ‘it is difficult to schedule an appointment’ (25.9% of CMD cases agree/strongly agree vs 11.6% of non-cases) and ‘it would be difficult to get time off work for treatment’ (26.7% of CMD cases agree/strongly agree vs 14.7% of non-cases) (Additional file 1: Table S1).
Those meeting criteria for probable alcohol misuse were not significantly more likely to report access problems compared to individuals not meeting case criteria (AOR 1.17; 95% CI 0.92; 1.49) (Table 2). Those meeting criteria for probable alcohol misuse did not noticeably endorse any particular access-related items compared to those not meeting criteria.
Internalised stigma of mental illness
Internalised mental health-related stigma was significantly more likely to be reported in those meeting criteria for probable PTSD compared to those not meeting criteria (AOR 1.33; 95% CI 1.20, 1.47) (Table 2). A number of items were strongly endorsed by individuals with probable PTSD, including concerns that ‘my unit bosses might treat me differently’ (73.8% PTSD cases agree/strongly agree vs 46.0% of non-cases) and ‘members of my unit might have less confidence in me’ (62.9% PTSD cases agree/strongly agree vs 41.1% of non-cases) (Additional file 1: Table S1).
Compared to those not meeting criteria, individuals meeting criteria for probable CMD were significantly more likely to report concerns relating to the stigmatisation of mental illness (AOR 1.19; 95% CI 1.11, 1.28); including ‘I would be seen as weak by those who are important to me’ (56.9% CMD cases agree/strongly agree vs 36.4% of non-cases) and ‘my boss would blame me for the problem’ (33.6% CMD cases agree/strongly agree vs 13.1% of non-cases).
Individuals meeting criteria for probable alcohol misuse were significantly more likely than those not meeting criteria to report mental health related stigma (AOR 1.12; 95% CI 1.04; 1.21) (Table 2). In particular, ‘concerns about what my friends and family might think’ was a more commonly endorsed item by those meeting probable alcohol misuse case criteria (45.4% cases agree/strongly agree vs 29.9% of non-cases) (Additional file 1: Table S1).
Perceived stigma of mental health care/providers
Compared to those not meeting criteria, individuals meeting criteria for probable PTSD were significantly more likely to report issues related to mental health treatment and service providers (AOR 1.68; 95% CI 1.36; 2.07) (Table 2). Items, including ‘mental health care doesn’t work’ (20.1% PTSD cases agree/strongly agree vs 5.9% of non-cases) and ‘my bosses discourage the use of mental health services’ (13.4% PTSD cases agree/strongly agree vs 3.3% of non-cases) were particularly commonly endorsed by those meeting probable case criteria for PTSD (Additional file 1: Table S1).
In comparison to those not meeting criteria, participants meeting probable CMD criteria were also significantly more likely to report concerns relating to mental health treatment and service providers (AOR 1.30; 95% CI 1.09; 1.55) (Table 2). Those with probable CMD strongly endorsed several items, including ‘not wanting a mental health problem to be on my medical records’ (59.3% CMD cases agree/strongly agree vs 44.2% of non-cases) and ‘I’ve had bad experiences with mental health professionals’ (23.5% CMD cases agree/strongly agree vs 10.1% of non-cases) (Additional file 1: Table S1).
Those meeting criteria for probable alcohol misuse were not significantly more likely to report concerns relating mental health treatment and service providers compared to those who did not meet diagnostic criteria (AOR 1.18; 95% CI 0.99; 1.40), although this effect is approaching statistical significance (Table 2). Nonetheless, compared to those who did not meet criteria, individuals meeting criteria for probable alcohol misuse particularly endorsed ‘wanting to solve the problem on my own’ (75.3% cases agree/strongly agree vs 58.9% of non-cases). This high-level of self-reliance was also more frequently endorsed by those meeting probable CMD and PTSD criteria compared to non-cases. Moreover, in comparison to those not meeting criteria, individuals with probable alcohol misuse were more likely to report concerns that ‘my visit would not remain confidential’ (15.4% cases agree/strongly agree vs 9.6% of non-cases) (Additional file 1: Table S1).