This study included four fisheries in Alaska (salmon gillnet), Oregon (Dungeness crab), and Massachusetts (scallop and lobster). Participating fishermen signed a consent form, under Mary Imogene Bassett Hospital Institutional Review Board protocol #1481898. Fishermen in the salmon, crab, and lobster industries were randomly selected from permit lists in specific geographic locations to the travel distance for inperson health assessments. Scallopers were recruited through settlement houses, as pre-existing permit lists were not available. Further recruitment efforts were conducted dockside, via social media outlets, through partner promotions, and local radio.

Survey instruments captured demographics, fishery/boat information, sleep, caffeine/substance use, diet, physical, emotional and financial health. Health assessments captured data on height, weight, vision, respiratory health, blood pressure, diabetes risk (A1C and glucose), cholesterol, and risk for heart disease (Framingham scores¹²). A screener for probable PTSD, the Primary Care PTSD Screen for DSM-5 (PC-PTSD-5)¹³ was self-administered. This five-question screener assesses whether an individual has experienced a traumatic event (i.e. has trauma history), and, if so, whether and how much that event impacted them over the past month. Scores of 3 or higher denote probable but not diagnosed PTSD. A clinical interview/assessment (not available for this study) is required for a definitive diagnosis. Henceforth, “PTSD rates/diagnoses” in this study is referring to probable rates/diagnoses obtained from the screening tool.

All data were recorded on paper, cleaned, and entered into the project REDCap database system.¹⁴ Two general sets of analyses were performed. The first involved a dichotomous variable denoting whether the subject screened positive for PTSD (Probable PTSD Outcome). The second set of analyses was a sub-analysis of the frequency of self-reported PTSD symptoms among those who experienced a traumatic event (Symptom Severity).

To identify probable PTSD, subjects were given the Primary Care PTSD Screen for DSM-5, where they were asked a series of questions (see Table 1).

Subjects who answered “yes” to three or more of these questions were classified as probable PTSD positive,¹³ resulting in 17 subjects being identified as PTSD positive and 125 as PTSD negative (negative subjects included 85 responding “no” to the original question, plus 40 who responded “yes” but had less than three positive responses to the five follow-up questions). Responses to categorical questions were compared between PTSD positive and negative subjects using chi-square, or Fisher’s Exact Test in cases where the assumptions for chi-square were not satisfied. Continuous variables were compared using the Wilcoxon Rank Sum Test.

Prior studies indicate the PC-PTSD-5 screener has high levels of diagnostic accuracy.^15,16 However, it is important to note that several studies suggest that a cut-point of four “yes” answers versus three would maximize sensitivity (few false negatives) while maintaining acceptable specificity (few false positives),¹⁵ and maximize efficiency (overall accuracy of test).¹⁶ For this study, researchers used three as the cut point to be compliant with current screener instructions, but note that validation of the screener tool is an important focus for future research.

For all 57 subjects who reported a trauma history, the sum of the “yes” responses to the five follow-up questions was calculated, resulting in a score of 0 to 5. The relationship of this variable to other demographic and sleep quality variables, was measured using Spearman’s Rank-Order Correlation Coefficient.

Male fishermen reported higher screening rates of PTSD (12.1%) than female fishermen (10.0%); however, there were few females in our sample (n = 10). On average, those screening positive for PTSD had been in the industry longer than those without PTSD. PTSD outcomes did not appear to affect sleep patterns while fishing, with all subjects getting consistently less than five hours per night (see Table 2).

Initial analyses identified factors associated with PTSD outcomes. Roughly, one-fifth (17.7%) of fishermen screening positive for PTSD also reported having depression, while 1.6% of fishermen screening negative for PTSD reported having depression (self-reported diagnosis) (p = .0118). Fishermen reporting financial struggles (p = .0157), or that they were unable to control important things in their lives over the last month (p = .0023), were significantly more likely to screen positive for PTSD.

Several sleep related variables were also significantly correlated with PTSD outcomes. Fishermen reporting difficulty falling asleep (p = .0278), staying asleep (p = .038), and having poor sleep quality (p = .0635, marginally significant) were more likely to screen positive for PTSD. Additionally, fishermen were asked how different factors impacted their sleep while they were on the boat. Negative working relationships on the boat (p = .0126), pain (p = .0104), concerns with family/relationships (p < .001), and unreliability of crew/captain (p = .0201) were all significantly correlated with positive PTSD screenings.

Using the cut-point of three for probable PTSD, 12.0% of fishermen (male and female) in the study screened positive. Referring back to the prior discussion on cut-point, if a cut-point of four is utilized, 6.3% screen positive, above the national average of 5% in any given year.¹⁷

Higher levels of financial insecurity were positively correlated with higher scores (i.e., the sum of “yes” responses) (p = .0067). Fishermen who reported that their sleep was impacted by: working relationships on the boat (p = .0065), pain (p = .0014), stress (0.0088), concerns with family/relationships (p < .0001), unreliability of crew/captain (p = .0273), or safety concerns on the boat (p = .0088), were more likely to have a higher score. Reporting current depression (p = .0207) or feeling unable to control the important things in life (p = .0153) were also positively correlated with higher scores.

Average hours of sleep per night while fishing (p = .5998), and least hours of sleep in a 24-hour period during the fishing season (p = .7264) were not significantly correlated with higher scores. Sleep quality questions, including getting enough sleep (p = .3857), having trouble falling asleep (p = .3964), having trouble staying asleep (p = .4004), and having great sleep quality (p = .3285) were also not correlated with the sum of “yes” responses.