Sex differences in alcohol and tobacco use disorders among individuals with panic disorder: A cross-sectional analysis from the genomic psychiatry cohort
Anxiety disorders and substance use frequently co-occur, yet the moderating effects of sex and ancestry on these relationships remain understudied. This investigation examined associations between presumed panic disorder (pPD) and both presumed alcohol use disorder (pAUD) and tobacco use disorder (pTUD) in 10,953 individuals from the Genomic Psychiatry Cohort screened against severe mental illness. Our sample was demographically diverse (56% female; 55% European Ancestry, 45% African Ancestry), allowing robust comparison across these groups. Individuals with pPD (n = 342) demonstrated significantly higher mean severity scores for both pAUD (1.26 vs. 0.33, p < 0.05) and pTUD (1.65 vs. 0.93, p < 0.05) compared with those without pPD. While female sex was associated with decreased risk for pAUD (B: −0.351, p < 0.05) compared with males, we observed no significant ancestry-based differences in substance use patterns among those with pPD. Two-way interaction analyses revealed that sex significantly moderated the relationship between pPD and pAUD (B: −0.97, p < 0.001), with the association being stronger among males than females. Additionally, comorbid presumed posttraumatic stress disorder was significantly associated with increased risk for both pAUD (B: 0.650, p < 0.05) and pTUD (B: 0.825, p < 0.05) but did not interact significantly with pPD. These findings advance our understanding of how biological sex influences the manifestation of comorbid panic and substance use disorders, offering clinical implications for assessment and treatment strategies that acknowledge sex-specific vulnerability patterns while highlighting the consistent relationship between these conditions across ancestral groups.
Introduction
Panic disorder (PD) is a debilitating condition characterized by repeated episodes of fear that last for several minutes or longer. Physical symptoms may include heart palpitations, sweating, chills, trembling, breathing problems, weakness, dizziness, and/or feeling of being out of control or impending doom. The etiology of PD is multifactorial. Current research suggests that biological, psychological, and/or environmental factors may contribute to why some individuals develop this disorder compared with others. From a genomic perspective, within certain families, some suffer more from PD than others. According to the World Mental Health Surveys (WMHS), the prevalence of PD across various countries is estimated at approximately 1.7%. Typically, the age of onset is 32, with most cases emerging between the ages of 20 and 47 (1). PD afflicts about 4.7% of U.S. adults at some time in their lives and based on diagnostic interview data from the National Comorbidity Survey Replication, an estimated 2.7% of U.S. adults were diagnosed with PD in 2017 (2). In terms of sex differences, PD tends to afflict more women than men. The prevalence is higher in females (3.8%) than in males (1.6%) (2). Further, the National Survey of American Life (NSAL) and the National Comorbidity Study (NCS) showed that non-Hispanic White participants reported higher rates of PD than non-Hispanic Black participants (4.8% vs. 3.5%) (3). WMHS also found that 80.4% of lifetime PD cases had comorbidity with other psychiatric illnesses (1). Lifetime comorbidity with other mental disorders includes other anxiety disorders, affective, substance use disorders (SUD), and impulse-control disorders (1, 4–6). PD is also associated with physical and emotional distress, detriments in social and romantic relationships, suicide attempts, financial difficulties, and SUD (7–10). Several theories for the comorbid relationships between anxiety disorders such as PD and SUD have been posited. For example, “self-medication” or “tension-reduction” theory states that patients with anxious disorders use substances to relieve their panic symptoms temporarily, thus developing SUD (11). Physiologic and/or environmental stressors from chronic substance use create situations or circumstances in which anxiety symptoms are more likely to emerge or worsen (12). As noted, the self-medication hypothesis highlights a key mechanism by which individuals with PD may use substances like alcohol and tobacco to manage their symptoms, leading to higher rates of AUD (alcohol use disorder) and TUD (tobacco use disorder). While there is a robust relationship between PD with TUD and AUD (13, 14). there are few studies that examine sex differences in PD and SUD comorbidity (15) and even fewer reporting on ancestry/race (or ancestry-by-sex) differences in the comorbidity of PD with common SUD (AUD and TUD) in large, diverse samples. Sex and ancestry/race may influence comorbidity patterns in PD and SUD. For example, sociocultural factors may shape substance use behaviors, potentially leading to varying prevalence rates of comorbid PD and SUD across these groups.
Individuals with PD are more likely to smoke and misuse alcohol compared to individuals with any other anxiety disorders (16–20). Multiple population-based studies have demonstrated the strong association between PD with TUD and AUD, including the Epidemiologic Catchment Area (ECA) survey, NCS, and the National Epidemiological Survey of Alcohol Related Conditions (NESARC). With regard to sex differences observed in recent studies on PD comorbidity with DSM-5 AUD and TUD, data derived from the NESARC found that females with PD were 4.5 times more likely to have TUD compared with males with PD (21). Regarding comorbidity, there are three main hypotheses of the mechanism underlying the association between smoking and PD, though none compare by ancestry/race as we will. (1) Smoking may increase the risk of development of PD through impaired respiration and increased vulnerability for panic attacks (22, 23) and through the expected calming effects of nicotine (24, 25); (2) Stimulant properties of nicotine may also reduce the threshold for experiencing panic attacks, especially as nicotine is metabolized in the bloodstream, and the nicotine levels drop (23, 26–28); (3) The association may be moderated by other variables such as anxiety sensitivity, a risk factor for developing PD (11).
Using data from the NCS, Kessler described that even though males had a higher incidence of AUD, females still had higher incidences of comorbid PD and AUD than males. Females were more likely to have a diagnosis of PD before AUD, whereas males were more likely to have a diagnosis of AUD before PD, and this may exacerbate alcohol problems in females with PD since it may be a form of self-medication (29). In another study that hypothesized there were sex differences between PD individuals, 169 patients with PD and established alcohol dependence (89 males and 80 females) in the NHIRD [Taiwan National Health Insurance Research Database] were nested with other 9480 patients with only PD, the standardized incident ratio (SIR) for AUD was significantly higher for females with PD (SIR = 6.29) as compared with males with PD (SIR = 3.36) (15).
Previous studies have consistently demonstrated that PD is more prevalent among White Americans in comparison to their African Ancestry (AA) counterparts (30–33). Surprisingly, panic disorder with comorbid AUD and TUD with a focus on ancestry/race has been sparsely researched. In a study pooling data from the NESARC, Smith et al. analyzed racial differences in the 12-month prevalence and co-occurrence between DSM-IV AUD and PD. In the study, AA had higher rates of comorbid AUD and PD with and without agoraphobia (OR = 4.2 and OR = 3.5, respectively) compared with EA counterparts (OR = 2.7 and 2.1, respectively) (30). This finding was consistent with Grant's previous NESARC publication (21). In a separate study, Huang et al. compared the prevalence and co-occurrence of DSM-IV AUD along with PD among European Caucasian Ancestry (EA) and AA in a large representative sample of the U.S. population between 2001 and 2002. Consistent with past studies, Blacks (AA) (OR = 2.5) were more likely than their white counterparts (OR = 1.9) to have co-occurring PD and AUD (34).
In this study, we examined rates of “presumed” alcohol and tobacco use disorders (pAUD, pTUD) and their association with “presumed” panic disorder (pPD) based on screening items for each disorder and the moderating influence of sex and ancestry/race. As mentioned earlier, PD has many psychiatric comorbidities, including PTSD (posttraumatic stress disorder). Berens conducted a further analysis of the NESARC study, finding that approximately 27.4% of individuals with primary PD also have lifetime PTSD (35). These findings align closely with those of other epidemiological studies, such as the NCS, which similarly reported a comorbidity rate of 27% (36). Therefore, we also examined the role of pPTSD (presumed PTSD) in the associations of pPD with pAUD and pTUD and differences by sex and ancestry/race. Thus, our research questions for this analysis of the GPC cohort, based on recent findings, are:
Does sex and ancestry/race have an association in the prevalence of individuals who experience comorbid pPD and pAUD and pTUD compared with those without PD (no pPD)?
How do sex and ancestry/race work separately and in combination with one another to influence the prevalence rates of comorbid pPD and pAUD and/or pTUD?
What are the effect sizes of pAUD/pTUD in pPD alone versus with comorbid pPD + pPTSD? Is there an interaction between pPD and comorbid pPTSD and/or pAUD and pTUD?
Results
Among all participants in the analytic sample (n = 10,953), 342 screened positively for pPD, and 10,611 did not (no pPD). As shown in Table 1, a higher proportion of participants were female; however, rates of pPD as compared with no pPD were similar among females (pPD: 57.9%, no pPD: 56.0%) and males (pPD: 42.1%, no pPD: 44.0%). Similarly, while there was a greater number of participants of EA than of AA, rates of pPD were not statistically significantly different (EA: 55.6%; AA: 44.4%) (Table 1).
| Presumed panic disorder (pPD) N = 342 (%) |
No presumed panic disorder (no pPD) N = 10,611 (%) |
Significance (p-value) | Overall N = 10,953 (%) |
|
|---|---|---|---|---|
| Female Sex | 198 (57.9) | 5942 (56.0) | 0.488 | 6140 (56.1) |
| Male Sex | 144 (42.1) | 4668 (44.0) | (reference group) | 4813 (43.9) |
| Age (mean, SD) | 43.75 (14.789) | 42.22 (15.368) | 0.069 | |
| African Ancestry (AA) | 152 (44.4) | 4763 (44.9) | 0.871 | 4915 (44.9) |
| European Ancestry (EA) | 190 (55.6) | 5848 (55.1) | (reference group) | 6038 (55.1) |
| PTSD screen positive | 103 (30.1) | 556 (5.2) | < 0.05 | 659 (6.0) |
As shown in Table 2 and Figure 2, mean alcohol problem scores were significantly higher among those who screened positive for pPD when compared with those with no pPD (pPD: M = 1.26, SD = 1.923; no pPD: M = 0.33, SD = 1.940; p < 0.05). Similarly, tobacco use scores were significantly higher among those who screened positive for pPD when compared with those with no pPD (pPD: M = 1.65, SD = 1.675; no pPD: M = 0.93, SD = 1.435, p < 0.05). Female gender was associated with a decreased risk for AUD (B: −0.351, p < 0.05) and TUD (B: −0.520, p < 0.05) as compared with males. Screening positive for pPTSD was associated with increased risk for pAUD (B: 0.650, p < 0.05) and pTUD (B: 0.825, p < 0.05) (Figure 3).
| Alcohol Use Disorder | Tobacco Use Disorder | |||
|---|---|---|---|---|
| Effect size (beta) | p-value | Effect size (beta) | p-value | |
| Main Effects | ||||
| pPD** | 1.411 | <0.05 | 0.556 | <0.05 |
| Interactions | ||||
| pPD x female sex** | −0.974 | <0.05 | −0.355 | 0.083 |
| pPD x AA | 0.017 | 0.918 | 0.125 | 0.594 |
| pPD x sex x AA | −0.202 | 0.339 | 0.347 | 0.257 |
| pPD x PTSD | −0.101 | 0.402 | −0.323 | 0.064 |
**Significance p < 0.05.
Two-way interactions (pPD x ancestry/race; pPD x sex; pPD x pPTSD), and three-way interactions (pPD x sex x ancestry/race) were also examined (Table 2). These interaction effects were included to explore whether the relationships among pPD, pAUD, and pTUD varied by demographic factors such as sex, ancestry/race, and pPTSD status. Significant interaction terms indicate that the association between pPD and substance use outcomes was modified by these factors, suggesting differences in comorbidity patterns across demographic groups. The only significant moderating influences of sex on the association of pPD and alcohol use problems were observed (B: −0.97, p < 0.001). That is, the influence of pPD on alcohol use problems is greater among males (see Figure 3) as compared with females (Table 2; Figure 3).
Discussion
This study stands out as one of the few that investigates the intersection of pAUD and pTUD within individuals with pPD. Moreover, it aims to delineate these groups more precisely by considering factors such as sex and ancestry/race, adding depth to our understanding of these complex relationships. To our knowledge, this is the first paper that looks at comorbid pPTSD in patients with pPD along with pAUD and pTUD. As aforementioned, the prevalence of people with PD in the United States is about 2.7%, which is similar to our sample's prevalence rate of 3.22%. Females are typically more afflicted with PD than males, which is confirmed in our data as well since in our overall case sample, about 58% of our cases with pPD were female, and 42% of our cases with pPD were male. We have also found, like others, that individuals with pPD have significantly more comorbidities, including pTUD, pAUD (Figures 2 and 3), and pPTSD (Tables 1 and 2).
Prior studies report that individuals with PD are two to four times more likely to smoke and drink compared with the general population (37). While our study does not measure prevalence directly, we found that among those who engage in these behaviors, individuals with pPD had a higher mean severity score for pTUD (1.65/4.0) compared with those with no pPD (0.93/4.0). Similarly, individuals with pPD had a higher mean severity score for pAUD (1.26/6.0) compared with those with no pPD (0.33/6.0). This suggests that individuals with pPD may experience more problematic use rather than just higher rates of use. Our analysis confirms the current literature that individuals with PD do smoke more tobacco and drink more alcohol (Figures 2 and 3) (16–19, 20, 23, 38–40).
Ancestry Differences in pPD and Comorbid pTUD or pAUD
Although past studies have reported ancestry/race differences in individuals with comorbid pPD + pAUD with AA individuals having higher prevalence than EA individuals, we found a lower prevalence in our AA sample (21, 30, 34, 41). Initially, it appeared that ancestry/race had an effect when looking at several covariates at once; however, this was not the case when we performed two-way and three-way interactions (Table 2). One possible explanation is that alcohol misuse and AUDs are less prevalent among AA individuals than those of EA, and therefore this may be driving these results. Ransome et al. (2017) (41) demonstrated that Black women with AUD had different health outcomes and drinking patterns than their White counterparts, potentially influenced by sociocultural factors and systemic barriers to care. While this study did not specifically examine PD or PTSD, its findings on racial disparities in alcohol use align with our observation of lower pAUD prevalence in AA individuals.
An alternative hypothesis could relate to the inherent nature of structural racism in the United States. For example, a recent study has highlighted the role of discrimination in influencing mental health and substance use outcomes. Vu et al. (2019) (42) demonstrated that intersecting identities, such as ancestry/race and sexual orientation, and experiences of discrimination significantly impact mental health outcomes, including depressive symptoms and substance use. This could further explain differences in comorbid substance use and PD across racial groups (42).
Additionally, AA children are more likely socialized to expect hostility, irrational restrictions, insults, and unfair treatment based on the color of their skin. Thus, to protect against this, AA children are taught to develop high levels of tolerance for unfair acts (43). This type of socialization has been shown to be protective of mental health symptoms (44). Although this socialization was not explicitly tied with PD, it could be postulated as a factor for the lower prevalence of PD. While this hypothesis is speculative, it warrants the need for further investigation.
Another possible explanation could be that limited access to mental health care and poor quality care when accessing care may be a factor for the lower prevalence of PD diagnoses (45, 46). Third, stigma and judgment also play a role in the lower rates of AA individuals receiving treatment. In a qualitative study by Alvidrez, one-third of black individuals regarded receiving treatment for mild depression and anxiety as “crazy” (47). In regards to comorbid substance use, studies have shown that even in precarious living situations comparably to their white counterparts, Black individuals with PD engaged in fewer unhealthy behaviors, such as tobacco, alcohol, and substance use (48, 49). Thus, the reasons for a decreased prevalence of PD and TUD or AUD in AA are multifactorial and are likely underreported in our sample (Table 1).
Sex Differences in pPD and Comorbid pTUD or pAUD
In our analysis, pPD is associated with increased pAUD and pTUD when looking at the main effect, adjusting for age, sex, ancestry/race, and pPTSD. When examining interactions between sex, ancestry/race, and pPTSD (two-way interactions), and sex and ancestry/race (three-way interactions) for pAUD and pTUD, only the interaction of sex with pAUD was significant, with it being greater among males compared with females (Table 2; Figure 3). This is in agreement with decades of research demonstrating a greater risk for alcohol use problems among males compared with females (50–52). General explanations for this male bias toward alcohol use problems include trends of male drinkers consuming alcohol more often and in larger quantities than female drinkers and, historically, limited opportunities for females to drink heavily due to societal norms, pregnancy, and child-rearing (53).
However, we should note that this contrasts with some recent studies that have shown that women with anxiety disorders, including but not limited to PD, drink more alcohol compared to men. For example, in a study in Germany, women with anxiety disorders tended to have an earlier onset of drinking, regular drinking, and occurrence of withdrawal symptoms when compared with women without anxiety disorders (54). They also noted there may be more severe physical symptoms among these women (54). Another study also found that women have more severe respiration-related symptoms and more severe agoraphobic avoidance symptomology compared with men (55). With both studies in mind, (54, 55) it could be hypothesized that females with anxiety disorders are already susceptible to significantly distressing withdrawal symptoms that could potentiate the desire to self-medicate with alcohol as a temporary measure to dampen panic disorder symptomology. However, Schneider's (54) study does not explicitly differentiate female individuals with PD from female individuals with other anxiety disorders. Therefore, the difference we found, wherein women with pPD were drinking less than men with pPD, is in agreement with the general literature on increased alcohol use problems in males compared with females and may not be in contradiction to these previous studies of women with anxiety including PD, since they had not separated women specifically with PD from other more common anxiety disorders (e.g., generalized anxiety disorder) (54, 55).
While studies like Schneider et al. (54) and Sheikh et al. (55) explored broader anxiety disorders rather than PD specifically, our findings align with their observations regarding differences in drinking behaviors. However, it is important to note that these studies utilized more comprehensive data collection methods. For instance, White et al. (2015) (53) captured longitudinal drinking patterns and detailed alcohol use history, whereas our study relied on a lifetime screener to assess drinking behaviors across both cases and controls. This difference in methodology could contribute to the observed discrepancies between studies.
With regard to tobacco use, there has been much research dedicated to understanding how substances may be used to alleviate feelings of anxiety, such as those that arise in PD. In an integrative model of smoking-anxiety comorbidity, smokers with anxiety disorders tend to be fearful of anxiety-related symptoms and bodily sensations, and they cope by using nicotine to achieve the “positive” sensations of nicotine in hopes of relaxing themselves. Alternatively, individuals may also use nicotine to escape and avoid emotionally distressing events (56, 57). This creates a feed-forward cycle that is difficult to break because the individual will inevitably feel nicotine withdrawal-related aversive interoceptive cues that occur routinely and cope by smoking some more (58, 59).
Historically, in the United States, males have always smoked more than females. Thus, at first glance, our finding of less smoking among women with pPD than men might seem supported by the literature. However, while smoking rates have decreased throughout most age groups, rates of smoking cessation in women have not. From a physiological standpoint, this may be due to the differences in nicotine metabolism observed between females and males. In a study by Benowitz, intravenous infusions of nicotine and cotinine showed higher clearance rates in females than in males, and oral contraceptives further accelerated this clearance (60). Thus, one explanation of the notable higher smoking rates in females with PD versus women without PD could be a combination of this feed-forward cycle and the more rapid metabolic clearance of nicotine. Goodwin substantiates the hypothesis that female smokers likely smoke due to underlying anxiety spectrum diagnoses. In a sample of 4149 individuals, Goodwin examined the prevalence of mood and anxiety disorders among male and female current smokers in the United States using data from NCS and NCS-R and found that PD was significantly more common among female current smokers than in males in 2001 compared with 1990 (61).
We did not see this TUD interaction in our female sample (Table 2) and wonder whether this is simply due to the power limits of our sample size since, as Table 1 indicates, the overall female sample was only 198 pPD versus 5942 without pPD, and for males. A total of 144 with pPD versus 4668 without pPD. Another limitation includes the diagnoses for PD, PTSD, AUD, and TUD, which are based on screener/self-report items and not on a full diagnostic assessment for a DSM diagnosis. For this reason, AUD and TUD were treated as sum scores, representing the sum of the number of screening items endorsed. Although it may be considered a limitation, in our study, only 342 participants out of 10,611 answered yes to items 11 and 11a. The use of fewer questions could be implemented in clinical settings, where sometimes these questions are not asked, to screen out participants and direct them to appropriate services briefly.
Limitations
Findings should be considered within the context of this study's limitations. The diagnosis for PD and PTSD were based on screener items and not on a full DSM diagnosis. Further, screeners were self-reported and, therefore, subject to bias. However, even with the limited number of questions, only a very small portion met the criteria for PD or PTSD, and rates were comparable to those in the general population. In terms of pAUD and pTUD diagnoses in our sample, to measure severity, we used sum scores to represent the sum of the number of screening items endorsed. Future analysis should determine whether these findings extend to DSM diagnoses of AUD and TUD (e.g., DSM-V AUD, TUD). A larger sample size with PD in future studies would also strengthen our findings. Another limitation is that the cross-sectional study design for a lifetime of illness did not allow us to determine whether AUD or TUD preceded or followed PD or PTSD. Future studies of a more longitudinal nature could examine SUD diagnoses that precede or follow a diagnosis of PD. In terms of missing data, the overall dataset had minimal missing entries. Participants with incomplete data were excluded from analyses involving those variables. Given the low volume of missing data, imputation techniques were not necessary, and the results were not significantly impacted.
Conclusion
In our study, we found significantly more alcohol and tobacco use in those with pPD. By sex, females were more likely than males to have pPD (Table 1), but were less likely to have alcohol use problems compared with males (Figure 3; Table 2). We did not find that ancestry/race (AA vs. EA) had a moderating effect on alcohol and tobacco use problems among individuals with pPD. We did not find any differences in alcohol or tobacco use problems when we looked at interactions with comorbid pPTSD (Table 2). However, we did find significant comorbidity of pPTSD in the pPD sample compared with controls (Table 1). Continuing to document differences in risk factors for mental health conditions by ancestry/race and sex is important, as this can impact and guide national healthcare policy in the United States. Future research should look at specific genetic or environmental factors that reinforce the comorbidities of PD-AUD and PD-TUD. This will help guide future management and treatments. In summary, this study advances our understanding of the interplay between anxiety, SUD, and PD, particularly in the context of sex and ancestry/race differences, offering a foundation for future research and tailored clinical approaches.
Materials and Methods
Sample
The analytic sample is derived from the Genomic Psychiatry Cohort (GPC), a large and diverse sample of individuals screened for neuropsychiatric disorders. The primary focus of the GPC is on severe mental illnesses such as schizophrenia and bipolar disorder. However, it also includes a large cohort of participants who are unaffected by these disorders (N = 10,953), defined as individuals not diagnosed with schizophrenia or bipolar disorder and with no first- or second-degree relatives with these disorders. Controls also have no history of traumatic brain injury and no history of multiple depressive episodes (>4) (62). Further, individuals were identified as male and female based on genome, XX or XY, and self-identified as African or European Ancestry (AA, EA, respectively) were included in this analysis given the limited number of participants of Latino or mixed ancestry. This analytic sample is 56% female, 55% of EA, and 45% of AA (Table 1). One unique feature of the sample is a close to ∼1:1 ratio of individuals of AA and EA in cases and control samples, making it a valuable dataset for the study of ancestry/race (and sex) differences in the comorbidity of these disorders. Participants are drawn from across the United States, Canada, Mexico, Portugal, and other regions of South America and Europe; however, only those who identified as AA or EA were included in the analytic sample.
Measures
Individuals who endorsed both PD screening items (i.e., responded “Yes” to items 11 and 11a) (Figure 1) are described as affected with pPD in this study, and individuals who endorsed 0–1 items, were described as unaffected with panic disorder (no pPD). Using these parameters, we identified 342 individuals with a pPD and compared these individuals with those unaffected noPD (Table 1). The screening questionnaire also includes six items (items 12–17) regarding alcohol use problems and four items (items 18–21) regarding tobacco use problems adapted from the “Cutting down, Annoyance by criticism, Guilty feeling, and Eye-openers”(CAGE) questionnaire (63). For both alcohol and tobacco use problems, screening items, sum scores for severity were computed, AUD ranged from 0 to 6 and TUD from 0 to 4. These specific cut-off points were chosen based on SAMHSA's Treatment Improvement Protocol recommendations, which aim to identify individuals at risk for SUD by casting a wider net. The graded scoring system allowed us to capture a spectrum of severity, where higher scores reflected a greater cumulative burden of alcohol or tobacco use. Of note, while PTSD screening items were included on the screening questionnaire, other anxiety disorders, including generalized anxiety, social phobia, separation anxiety, and specific phobia were not assessed. pPTSD was defined as responding yes to item 32 endorsing a traumatic event and then yes to at least two of the three screener questions 32a, 32b, 32c (Figure 1).
Citation: Genomic Psychiatry 2025; 10.61373/gp025a.0020
Citation: Genomic Psychiatry 2025; 10.61373/gp025a.0020
Citation: Genomic Psychiatry 2025; 10.61373/gp025a.0020
Statistical Analysis
First, Chi-squared and independent t test were used in Table 1 to assess the associations between pPD and demographic variables (Table 1). Significance was defined as a p < 0.05. Then, linear regression models were used to examine the association of pPD with pAUD, pTUD, and pPTSD. These models were computed separately for alcohol and tobacco use problems and included the following covariates: age, sex, self-reported ancestry/race, and pPTSD (Table 2). Further, all individuals identified as male and female and of AA and EA were included throughout this analysis, given the limited number of participants in other groups. All analyses were done using SPSS software (version 24).
Study Approval
The GPC study is approved by the Rutgers Institutional Review Board – Health Sciences, New Brunswick/Piscataway, located at 335 George St., Liberty Plaza Ste. 3100, New Brunswick, NJ 08901. This study involves human subject research, and written informed consent was obtained following a thorough explanation of the study's nature and potential outcomes.
Acknowledgment
We are grateful to the Genomic Psychiatry Cohort Consortium.
Author Contributions
M.J. Chung: conceptualization, writing – original draft. P. Georgakopoulos: formal analysis, project administration. J. Meyers: formal analysis, writing – original draft. S. Sharma: writing, project administration. C.N. Pato MD: conceptualization, funding acquisition, investigation, methodology. M.T. Pato: conceptualization, funding acquisition, investigation, methodology, writing – original draft.
Funding Sources
This work was supported by National Institute of Mental Health (NIMH).
Author Disclosures
The authors declare no conflict of interest.
GPC screener questions. This figure illustrates the screening questionnaire items used in the GPC study to identify pPD, pAUD, pTUD, and pPTSD. The screener includes specific questions for each condition with yes/no response options. For pPD, participants were classified as positive if they answered “yes” to both items 11 and 11a, indicating experience of sudden anxiety attacks with physical symptoms and subsequent worry or behavior changes. The alcohol section includes six questions (items 12–17) addressing risky drinking patterns and consequences, while the tobacco section includes four items (18–21) addressing use patterns and quit attempts. PTSD screening (items 32–32c) focuses on traumatic experiences and subsequent symptoms.
Mean sum scores for alcohol and tobacco risk use for participants with presumed panic disorder (pPD) and without presumed panic disorder (no pPD). This bar graph compares the severity of alcohol and tobacco use problems between individuals with and without pPD. Participants with pPD demonstrated significantly higher mean severity scores for both alcohol use (1.26 vs. 0.33, p < 0.05) and tobacco use (1.65 vs. 0.93, p < 0.05) compared with those without pPD. The blue bars represent individuals with pPD, while the orange bars represent those without pPD. Error bars indicate SDs.
Mean sum scores for risky alcohol and tobacco use among male and female participants with presumed panic disorder (pPD) and without presumed panic disorder (no pPD). This figure illustrates the sex differences in alcohol and tobacco use problems among individuals with and without pPD. The data are stratified by both sex and pPD status, revealing significant interactions. Among males with pPD, mean alcohol risk scores (2.10) were substantially higher than among males without pPD (0.53). Similarly, males with pPD showed higher tobacco risk scores (2.06) compared with males without pPD (1.23). For females, those with pPD also demonstrated higher alcohol risk scores (0.64 vs. 0.18) and tobacco risk scores (1.34 vs. 0.69) compared with females without pPD. All comparisons were statistically significant (p < 0.05). Notably, males with pPD reported higher alcohol risk scores than females with pPD (2.10 vs. 0.64).
Contributor Notes
Publisher's note: Genomic Press maintains a position of impartiality and neutrality regarding territorial assertions represented in published materials and affiliations of institutional nature. As such, we will use the affiliations provided by the authors, without editing them. Such use simply reflects what the authors submitted to us and it does not indicate that Genomic Press supports any type of territorial assertions.
