Culture and COVID-19 Related Impacts on Alcohol-Exposed Pregnancy Risk among Urban American Indian and Alaska Native Young Adults: A Path Analysis

Abstract

Purpose:

We investigated the influence of COVID-19 pandemic-related factors and cultural identification on indicators of alcohol-exposed pregnancy risk among urban American Indian and Alaska Native (AIAN) youth ages 16–20.

Methods:

Path analysis was used to evaluate associations among measures collected at baseline of the Native WYSE CHOICES, an intervention study to reduce alcohol-exposed pregnancy risk among urban AIAN youth. Data were collected nationally from 439 female-at-birth AIAN youth (mean age 18.1 years) residing in urban areas.

Results:

Identification with Native American culture and heritage was significantly and positively associated with self-efficacy to abstain from alcohol (β=0.178; p<0.001), and, indirectly, significantly and negatively associated with the risk of alcohol-exposed pregnancy (β=−0.046; p=0.001). Alternatively, distress about the pandemic’s consequences for Native American communities and experiencing economic impacts exacerbated by the COVID-19 pandemic were significantly and negatively associated with self-efficacy to abstain from alcohol (respectively: β=−0.165, p<0.01; β=−0.126, p<0.01) and, indirectly, significantly and positively associated with risk of alcohol-exposed pregnancy (respectively: β=0.043, p<0.01; β=0.033, p<0.05).

Discussion:

Results underscore the protective role that connectedness to Native American culture and heritage plays directly and indirectly in alcohol-exposed pregnancy risk potential. Conversely, the experience of economic barriers and distress over loss of Native American population and culture are direct and indirect risk factors for potential alcohol-exposed pregnancy risk. Development of future alcohol-exposed pregnancy interventions for urban AIAN young adults should consider Native American cultural identification, historical trauma, and economic impacts in program design and implementation.

Introduction

The COVID-19 pandemic resulted in social, economic, and educational devastation for most youth; for American Indian and Alaska Native (AIAN) youth, this devastation was compounded by the historical and contemporary trauma experienced by their communities, increased risk of illness for themselves and their families, overburdened and under-resourced healthcare, and in some cases, cultural isolation.^1,2 A small but growing body of research has identified mental and physical health sequelae including anxiety, depression, and substance use.^1,3,4 Among the concerning pandemic-related outcomes was an elevated risk for alcohol-exposed pregnancy,⁵ brought about by increased levels of alcohol use and constrained access to contraception.^6,7 Alcohol-exposed pregnancy can cause Fetal Alcohol Spectrum Disorder among children and lifelong economic, social, and health consequences.⁸ The prevalence of alcohol-exposed pregnancy among AIAN young adults often go unreported in national studies due to small sample sizes.^9,10 For AIAN youth and young adults, alcohol-exposed pregnancy risk is frequently assessed through component behaviors: alcohol use, sexual activity, and contraceptive use. Ample research has documented concerning levels of each of these,^4,11–13 suggesting similarly concerning levels of risk for alcohol-exposed pregnancy. A few studies have also documented the role of protective factors in substance use and sexual risk-taking for AIAN youth, including community and culture as buffers to those risk behaviors,^1,14,15 although most of this work has been based on small, mainly rural samples.

To date, little is known about urban AIAN youth’s experience of the pandemic and its impact on alcohol-exposed pregnancy, even though urban AIANs represent over 70% of those identifying as AIAN.¹⁶ Since they comprise only a small fraction (often less than 1%) of most urban communities, sample sizes are commonly too small for meaningful analysis in most national and local urban surveys. AIAN alcohol-exposed pregnancy risk is shaped by intrapersonal, behavioral, community, and cultural context;^17,18 experiencing the seismic shocks of the pandemic likely shifted those patterns and profiles in important ways for urban AIAN communities. The context of urban AIAN life has been shaped by a long history of trauma, genocide, and forced removal from ancestral lands, including mid-20^th century federal policies of relocation of AIAN people from rural reservation communities to urban areas.^3,19 Some who migrated maintained close connection to tribal communities; others endured severed or limited cultural connections. Thus varied levels of cultural engagement – both strong and limited – have had continued impact through generations.²⁰ In urban centers, where many cultural groups are represented, inter-tribal connection has also been important.²⁰ While cultural identification and connection is highly varied in urban AIAN communities, research has found a positive correlation between cultural identification and urban AIAN youth reports of good health,^21,22 although results have been mixed, requiring contextual interpretation.²³ For urban AIAN young people, culture, community, and connection may be a source of resilience in many situations. Each of these, in addition to knowledge, has been hypothesized to strengthen self-efficacy and confidence in navigating substance use and sexual activity.^24,25 The pandemic likely disrupted many of these pathways. Urban AIANs experienced significantly elevated rates of COVID-19 compared to non-Hispanic Whites in both pre-vaccination and post-vaccination periods,²⁶ subjecting many to high levels of familial and cultural loss, in addition to pervasive social and economic upheaval.² Thus, urban AIAN young adults, while diverse in ancestry and connection with tribal communities, were likely to have experienced unique cultural, socioeconomic, and health impacts of the pandemic relative to their tribal-land-based counterparts,^27,28 including on alcohol-exposed pregnancy risk-taking. For example, Dickerson and colleagues found that during the pandemic, social networks were associated with emotional well-being among AIAN young adults, depending on Native community experience and residence.²⁰ Haskins and colleagues examined explicit associations of pandemic experiences and found that among urban AIAN adults, pandemic-related threats, such as sickness and loss of culture, were inversely related to emotional well-being, but not substance use.²⁹

Using a unique data source of AIAN youth from urban communities nationally we build on this prior work to investigate the relationship between pandemic experiences and alcohol-exposed pregnancy-related risk behaviors. Identifying relationships among pandemic conditions, cultural identification, knowledge and self-efficacy with alcohol-exposed pregnancy risk may assist in developing appropriate ongoing support for these AIAN youth and inform preventive strategies and programs for other traumatic events.

Methods

We used baseline data collected as a part of a randomized controlled trial of Native WYSE CHOICES,³⁰ a culturally tailored mobile app intervention to prevent alcohol-exposed pregnancy, in a national sample of urban AIAN young adults. The study was conducted entirely virtual, including consent, data collection, and intervention administration. The study was approved by the University of Colorado Multiple Institutional Review Board (#20-3122) and is registered with ClinicalTrials.gov (#NCT04376346).

Participants

Participants were eligible for the study if they self-identified as AIAN alone or in combination with other races or ethnicities, were 16–20 years-old, assigned female at birth, lived in an urban area (population 50,000+) outside of tribal lands, were not pregnant or breast-feeding, and were willing to provide an email address. Alaskan residents, representing about 2% of the total AIAN population, were excluded because we did not receive the Alaska-specific approval in time for the study of interest.

In total, 3,280 potential participants accessed the study screening protocol through social media posts, e-newsletters, listservs, or in-person events as possible during the pandemic.⁴¹ All those found eligible (n=534) were invited to enroll in the study. Of those, 48 declined to participate, and 47 never started the baseline survey, for a total of 439 enrolled participants across 42 states.⁴¹ For each eligible participant, we obtained informed consent and sent a unique survey link via email (parental consent was waived for minors). Participants were asked to complete the survey within four weeks of receiving the link and received up to four reminders via email or text. Baseline data were collected from August 4, 2021 (roughly coinciding with the Omicron surge) through January 4, 2023. Participants were paid $25 for completing the survey.

Measures

Pandemic-related measures.

With the abrupt onset of COVID-19 pandemic, few validated and reliable measures specific to pandemic experiences for youth existed. Several measures included in the study were based on questions found in Native-focused surveys and in the PhenX toolkit;³¹ others were developed by the project team. Measures were piloted from May to August 2021. We performed psychometric analysis of data to assess scale reliability and made appropriate modifications.

Economic impacts.

We asked participants if, in the past 30 days, they or anyone they currently lived with lost a job, had an employer decrease their work hours, had trouble getting enough food, or had trouble paying for housing, transportation, electricity, heat or other utilities, internet services, or a personal phone. Participants who endorsed any economic impact received a score of 1; 0 otherwise.

Pandemic-related historical trauma.

We used a five-item scale to measure pandemic-related historical trauma.^29,32 Participants were asked to rate how much they thought about whether the following would happen as a result of the pandemic: loss of Native languages, loss of elders and cultural keepers, loss of cultural practices, a reduction in their tribal population, and being reminded of how ancestors were “wiped out” from similar diseases. Responses were recorded on a 4-point scale ranging from 1 = “I never think about this” to 4 = “I think about this all the time.” Items were summed to create a total score and found to be internally consistent (α =.92).

COVID-19 mortality.

Participants were asked “has anyone you know died from COVID-19?” Positive responses were coded 1; 0 otherwise.

Cultural identification.

We used a cultural adaptation of the Multigroup Ethnic Identity Measure Revised (MEIM-R) to measure cultural identification.³³ Items included: I have spent time trying to find out more about my tribe(s), such as its history, traditions, and customs; I have a strong sense of belonging to my tribe(s); I understand pretty well what my tribal affiliation means to me; I have often done things that will help me understand my tribal background better; I often talk to other people in order to learn more about my tribe(s); and I feel a strong attachment towards my tribe(s). Participants were asked to rate the degree to which they agreed or disagreed with each item on a 5-point scale. The MEIM-R demonstrated acceptable internal consistency (α =.79).

Alcohol-exposed pregnancy knowledge.

Participants were asked 7 true-false questions about alcohol-exposed pregnancy.³⁴ The number of correct answers was summed; missing and “I don’t know” responses were counted as incorrect.

Contraception effectiveness knowledge.

Participants were asked about the percentage of women who would get pregnant within a year using each of 14 different contraceptives including natural family planning, withdrawal, and abstinence.³⁵ The number of correct answers were summed; missing and “I don’t know” responses were counted as incorrect.

Alcohol use self-efficacy.

Self-efficacy to abstain from alcohol was assessed using the Drinking Self-Efficacy Questionnaire (DSEQ).³⁶ The DSEQ includes two sub-scales – one measures likelihood of alcohol abstention in the face of social pressure to drink, the other measures likelihood of abstention in the face of negative emotions. Items are rated on a 6-point scale from 1 = “I am very sure I would drink” to 6 = “I am very sure I would not drink.” Items on the “social pressure” subscale include likelihood of drinking if friends are drinking, if at a party, wanting to feel more confident, someone offers a drink, and seeing others drinking. Items on the “emotional relief” subscale include likelihood of drinking if upset, down, worried, sad, angry, stressed, or nervous. Scores for each subscale were summed and both subscales demonstrated high internal consistency (α = .91 and α = .96, respectively).

Sexual health practices self-efficacy.

We adapted Koch’s scale³⁷ of sexual health practices to create a measure of sexual health efficacy that included 5 items. Participants were asked to rate on a 5-point Likert scale how confident they are to establish a sexually satisfying relationship, talk with a partner about sexual histories, talk with a partner about sexual health issues (e.g., an STI), and deal with a sexual functioning difficulty. Scores were summed and the scale was found to be internally consistent (α = .84).

Contraceptive use self-efficacy.

Participants were asked to state how true each of the following statements were for them: if my partner and I were getting intimate and moving towards having sex and I didn’t have protection, I could easily ask if he had protection; when I think about having sex, I know it’s my decision; and if my partner and I were getting intimate and moving towards having sex and I didn’t have protection, and I didn’t feel comfortable bringing up the subject, I could stop things before having sex.³⁸ Response options ranged from 1 = “not at all true for me” to 5 = “completely true for me.” Scores were summed and the scale demonstrated acceptable internal consistency (3 items; α = .73).

Alcohol-exposed pregnancy risk index.

Following Green,¹⁰ we assessed the accumulation of behaviors contributing to alcohol-exposed pregnancy. We created a scale where higher scores reflected greater alcohol-exposed pregnancy risk based on four risk behavior variables: 1) alcohol use in the past 30 days, 2) vaginal sex with a male in the past 30 days, 3) sexually active and not using effective contraception, and 4) using alcohol with sexual activity in the past 30 days. The alcohol-exposed pregnancy risk index ranged from 0 = no behaviors endorsed to 4 = all behaviors endorsed.

Analysis

Data analysis was conducted using SPSS³⁹ and Mplus⁴⁰ software. Response distributions were inspected for all variables to assess variability and detect floor and ceiling effects. Missing data were minimal since logical skips were coded into the survey. Descriptive statistics, either counts and percentages or means and standard deviations, were reported for model variables.

Path analysis, a form of structural equation modeling that permits assessing both direct and indirect effects of predictor variables on an outcome variable, was used to test whether identification with Native American culture and heritage, losing someone to COVID-19, historical trauma due to COVID-19, and economic impacts during the COVID-19 pandemic were associated with the alcohol-exposed pregnancy risk index through their relationship with alcohol-exposed pregnancy knowledge, contraception effectiveness knowledge, contraceptive use self-efficacy, sexual health practices self-efficacy, social pressure alcohol self-efficacy, or emotional relief alcohol self-efficacy, see Figure 1.

Figure 1.

Hypothesized conceptual model exploring influence of cultural identification and COVID-19 impacts on the alcohol-exposed pregnancy (AEP) risk index through AEP-related knowledge and self-efficacy indicators. Straight arrows indicate directional effects from one variable on another variable, curved arrows show covariances between variables, and circular arrows represent variance.⁴² “D” stands for disturbance and represents residual (unexplained) variation.⁴²

Path models were estimated using maximum likelihood with robust standard errors (MLR) due to non-normality of variables. Path models were assessed using four goodness-of-fit indices. A model chi-square (χ²) value with a ρ-value and the Standardized Root Mean Square Residual (SRMR) were evaluated for prediction, the Bentler’s Comparative Fit Index (CFI) was evaluated for comparative fit, and the Steiger-Lind Root Mean Square Error of Approximation (RMSEA) was evaluated for parsimony.

Results

The average age of participants was 18.1, (SD = 1.4), with roughly equal numbers reporting less than a high school diploma, a high school diploma, or some college. Univariate statistics for the variables in our hypothesized conceptual model (Figure 1) are shown in Table 1. Most participants (67%) reported experiencing one or more economic barriers in the past 30 days (N=419) and 57% reported knowing someone who died from COVID-19 (N=380). The average cultural identification score was 14.33 (SD=3.79, range=5–20) and the average historical trauma score was 2.92 (SD=0.79, range=1–4). Average scores for knowledge measures and average total sum scores were reported for self-efficacy measures. Less than half (43.2%) of participants (N=433) did not endorse any alcohol-exposed pregnancy risk behaviors, 31.4% endorsed one, 14.1% endorsed two, 8.8% endorsed three, and 2.5% endorsed all four.

Table 1.

Descriptive statistics for variables in the hypothesized conceptual model, including cultural identification and COVID-19 impacts, knowledge and self-efficacy indicators, and the alcohol-exposed pregnancy (AEP) risk index.

	No. or mean	% or SD
Economic impacts (N=419)
Experienced one or more economic barriers in the past 30 days.	281	67.06
COVID-19 mortality (N=380)
Yes, someone I know died from COVID-19.	215	56.57
Cultural Identification (6 items)
- I have spent time trying to find out more about my tribe(s), such as its history, traditions, and customs.	4.30	0.79
- I have a strong sense of belonging to my tribe(s).	3.72	0.98
- I understand pretty well what my tribal affiliation means to me.	3.93	0.90
- I have often done things that will help me understand my tribal background better.	4.01	0.86
- I often talk to other people in order to learn more about my tribe(s).	3.67	0.94
- I feel a strong attachment towards my tribe(s).	3.97	0.94
Total average sum cultural identification score (N=439)	14.33	3.79
Historical trauma (5 items)
- How often think: we will lose our Native languages.	2.65	0.97
- How often think: we will lose our elders and cultural keepers.	3.13	0.89
- How often think: we will lose our cultural practices.	2.83	0.97
- How often think: it reminds me of how our ancestors were wiped out from diseases.	2.94	1.01
- How often think: my tribal population will decrease.	3.07	1.01
Average historical trauma score (N=417)	2.92	0.79
Knowledge Variables
Alcohol-exposed pregnancy knowledge score (N=434)	5.43	1.26
Contraception effectiveness knowledge score (N=434)	4.01	2.02
Self-efficacy Variables
Emotional relief alcohol self-efficacy score (N=428)	34.12	9.35
Social pressure alcohol self-efficacy score (N=429)	20.28	7.26
Sexual health practices self-efficacy score (N=419)	17.45	4.73
Contraceptive use self-efficacy score (N=423)	12.99	2.48
AEP risk index (N=433)
No risk of AEP risk	187	43.19
Level of risk of AEP risk
- 1	136	31.41
- 2	61	14.09
- 3	38	8.78
- 4	11	2.54

SD=standard deviation. Cultural identification scores range from 1 to 5. The total average sum cultural identification score ranges from 6 to 20. Higher scores reflect higher identification with Native American culture and heritage. Historical trauma scores range from 1 to 4. Higher scores reflect higher distress over potential impact of COVID-19 pandemic on Native American community. For knowledge variables, higher scores indicate greater knowledge of alcohol exposed pregnancy risk (range: 0 to 7) and greater knowledge of contraceptive use effectiveness (range: 0 to 10). For self-efficacy variables, higher scores indicate greater efficacy not to drink in the face of negative emotions (range: 7 to 42), greater efficacy not to drink in the face of social pressure (range: 5 to 30), greater sexual health practices efficacy (range: 5 to 25), and greater contraceptive volition (range: 3 to 15).

Model fit results.

Information criteria indicated that the hypothesized conceptual model, shown in Figure 1, did not have good model fit. We reviewed empirical information and explored an alternative model, see Figure 2. The model fit indices for the alternative model showed a non-significant χ² value (χ² = 2.877, ρ ≥ 0.05) and the model’s CFI score was equal to 1, indicating that the model had good fit compared to the null model. The model produced a SRMR of 0.017, indicating good predictive fit. The model’s RMSEA was equal to 0 with a 90% confidence interval of (0.000, 0.062), indicating the model had an appropriate amount of parsimony.

Figure 2.

Alternative path model showing pathways from cultural identification and COVID-19 impacts to the alcohol-exposed pregnancy (AEP) risk index through emotional relief alcohol self-efficacy. Unstandardized B coefficients or path coefficients (bolded), covariances between variables, and variances of each variable are presented. Standard errors are in parentheses. Asterisks indicate the relationship was statistically significant at ρ=.05. The straight arrows indicate directional effects from one variable on another variable and the curved arrows show covariances between variables.⁴²

Path analysis results.

The alternative path model is displayed in Figure 2 with unstandardized (path) coefficients showing pathways between cultural identification and COVID-19 impacts and the alcohol-exposed pregnancy risk index through emotional relief alcohol self-efficacy. Direct and indirect effects of path analysis results for the alternative path model are presented as unstandardized and standardized coefficients in Table 2. Standardized estimates and p-values were assessed for statistical significance. In the alternative path model, cultural identification (β = 0.18, ρ < 0.0001), economic impacts (β = −0.13, ρ = 0.007) and COVID-19-related historical trauma (β = −0.17, ρ = 0.002) had significant direct effects on self-efficacy to abstain from alcohol in the face of negative emotions (i.e., emotional relief alcohol self-efficacy). The pathway from COVID-19 mortality was not statistically significant, though it was approaching significance (ρ < .10) both directly and indirectly with the AEP risk index. Self-efficacy to abstain from alcohol in the face of negative emotions subsequently showed a significant direct association with the alcohol-exposed pregnancy risk index (β = −0.26, ρ < 0.0001). A one-point increase in self-efficacy to abstain from alcohol in the face of negative emotions was associated with a 0.03-point decrease in the alcohol-exposed pregnancy risk index (ρ < 0.0001).

Table 2.

Path model direct effects of cultural identification and COVID-19 impacts on emotional relief alcohol self-efficacy, direct effects of emotional relief alcohol selfefficacy on alcohol-exposed pregnancy (AEP) risk index, and indirect effects of cultural identification and COVID-19 impacts on AEP risk index through emotional relief alcohol self-efficacy.

			Two-tailed
	B (S.E.)	β	p-value
Direct effects
Cultural identification → Emotional relief alcohol self-efficacy	0.439 (0.125)	0.178	0.000
COVID-19 mortality → Emotional relief alcohol self-efficacy	1.781 (0.951)	0.094	0.061
Economic impacts → Emotional relief alcohol self-efficacy	−2.504 (0.932)	−0.126	0.007
Historical trauma → Emotional relief alcohol self-efficacy	−1.946 (0.637)	−0.165	0.002
Emotional relief alcohol self-efficacy → AEP risk index	−0.03 (0.006)	−0.260	0.000
Indirect effects
Cultural identification → Emotional relief alcohol self-efficacy → AEP risk index	−0.013 (0.004)	−0.046	0.001
COVID-19 mortality → Emotional relief alcohol self-efficacy → AEP risk index	−0.053 (0.031)	−0.025	0.087
Economic impacts → Emotional relief alcohol self-efficacy → AEP risk index	0.075 (0.032)	0.033	0.019
Historical trauma → Emotional relief alcohol self-efficacy → AEP risk index	0.058 (0.022)	0.043	0.008

S.E. stands for standard error. B represents unstandardized path coefficients. Beta (β) represents standardized estimates. Two-tailed ρ-values are reported for standardized estimates (β coefficients).

Historical trauma (β = 0.04, ρ = 0.008) and economic impacts (β = 0.03, ρ = 0.019) had significant indirect positive effects on the alcohol-exposed pregnancy risk index through emotional relief alcohol self-efficacy. Through this indirect relationship, historical trauma from the COVID-19 pandemic was associated with a 0.06-point increase in potential risk of alcohol-exposed pregnancy risk (ρ = 0.008). Similarly, experiencing one or more economic barriers in the past 30 days had a significant indirect effect reflecting a 0.08-point increase in potential risk of alcohol-exposed pregnancy risk (ρ = 0.017). In contrast, cultural identification exerted significant indirect negative effects on the alcohol-exposed pregnancy risk index, through emotional relief alcohol self-efficacy (β = −0.05, ρ = 0.001). Thus, through its relationship with self-efficacy to abstain from alcohol in the face of negative emotions, identification with Native American culture and heritage was associated with a 0.01-point decrease in potential risk of alcohol-exposed pregnancy risk (ρ = 0.001). The model did not show a direct association between the exogenous variables; cultural identification, historical trauma, cultural identification, and COVID-19 mortality, and the outcome variable, alcohol-exposed pregnancy risk index. The R² value indicated the variables in the model could predict 6.7% of variation in the alcohol-exposed pregnancy risk index.

Discussion

To our knowledge, this is the first study to explore COVID-19 pandemic-related impacts on alcohol-exposed pregnancy risk among urban AIAN young adults. Our study revealed strong and significant direct and indirect associations of Native American culture and heritage, economic impacts, COVID-19-related historical trauma with the alcohol-exposed pregnancy risk index, through emotional relief alcohol self-efficacy.

We did not find direct or indirect pathways of alcohol-exposed pregnancy or contraception effectiveness knowledge with the alcohol-exposed pregnancy risk index. This perhaps is not surprising since the onset of the pandemic and subsequent social upheaval would be unlikely to change knowledge levels. Similarly, the lack of findings for contraceptive use and sexual health practices self-efficacy may also be understandable: If one has skills and confidence in these areas, the change in social and economic conditions wrought by the pandemic may not have an effect.

Evidence shows that adolescent alcohol use behaviors changed during the pandemic.⁴ Results here suggest that self-efficacy to abstain from alcohol in the face of negative emotions rather than in the face of social pressure was strongly associated with alcohol-exposed pregnancy related risk behaviors. Since social settings where pressure to drink alcohol generally did not occur during much of the pandemic, this finding also makes sense.

Previous work has demonstrated that, due to structural factors such as the ongoing harm and oppression due to colonial legacies, AIAN communities suffer disproportionately from poverty and are vulnerable to short- and long-term adverse socioeconomic and health impacts of the COVID-19 pandemic.^1,21 Similarly, we found that economic impacts and increased distress over the impact of the COVID-19 pandemic on the Native American community were associated with an increase in alcohol-exposed pregnancy risk index among our survey population. Such results suggest the importance of considering the economic setbacks and generational trauma brought on by the public health crisis when identifying strategies to support AIAN young adults on their journey to heal and strengthen protective factors associated with alcohol-exposed pregnancy risk.

Importantly, we found that identification with Native American culture and heritage was directly and indirectly a protective factor against indicators for alcohol-exposed pregnancy risk. This result strongly supports findings from multiple studies about the protective value of cultural connection and practice among AIAN youth during the pandemic,^1,28 and more generally,^14,15,25 suggests the need for increased social connection for urban AIAN youth.²⁸ While mechanisms supporting this association of culture with alcohol self-efficacy have been less explored, it is likely that strong cultural identification brings about strong sense of self, even in the face of negative emotions. Interestingly, compared to the other cultural identification items, participants were least likely to report talking with other people to learn more about their tribe. This finding may be an indication of social or cultural isolation among urban AIAN young adults and highlights the need to foster cultural ties and promote connection with cultural identity to cultivate strength and resiliency.

The findings of our study must be interpreted with caution. Though our study was a unique national sample of urban AIAN young adults assigned female at birth, we did not have a representative sample of this population. Study findings cannot be generalized beyond this sample. Additionally, we used cross-sectional baseline data from a randomized controlled trial, and thus we cannot infer causality. Finally, the quickly evolving epidemiological and social context of the COVID-19 pandemic meant using new and untested measures, though they were piloted prior to baseline. This study also holds many strengths. It provides empirical findings that contribute to the limited research on urban AIAN youth and alcohol-exposed pregnancy risk. Alcohol-exposed pregnancy research for the AIAN population is typically targeted (i.e., only those meeting specific behavioral criteria); this study was designed for universal, national reach in urban areas. Finally, these data are rich with insight into COVID-19 dynamics as experienced by this population. Our results are not just about the impacts of a time-limited event, they also provide insights into persistent repercussions long after the declared ending of the COVID-19 public health emergency. Native youth will continue to need support.

Conclusions

Our study demonstrated connections of cultural identification and COVID-19 impacts with self-efficacy to abstain from alcohol in the face of negative emotions, a factor associated with alcohol-exposed pregnancy risk. Our findings, while situated within the context of the COVID-19 pandemic, remain relevant. Though the pandemic is no longer a global or national public health emergency, the virus persists in communities while mental, social, and economic recovery may take years. Our study findings offer practical strategies for new and existing public health programs to reinforce protective factors and reduce risk factors associated with alcohol-exposed pregnancy risk among urban AIAN young adults. Strategies include enhancing community support networks for urban Native youth and investing in outreach programs that (1) establish cultural connectedness as a key program component, (2) acknowledge and provide support for experiences of historical trauma, and (3) offer resources that alleviate economic barriers to help urban AIAN young adults recover from the COVID-19 pandemic and prepare for future societal events and hardships.

Implications and Contributions:

This study demonstrated significant associations of cultural identification and COVID-19 impacts with alcohol use self-efficacy and potential alcohol-exposed pregnancy risk among urban AIAN young adults. Findings support alcohol-exposed pregnancy prevention strategies that prioritize cultural connectedness, incorporate support for historical trauma, and alleviate economic impacts for this population.

List of abbreviations:

AEP

Alcohol-exposed pregnancy

AIAN

American Indian and Alaska Native

CFI

Comparative Fit Index

DESQ

Drinking Self-Efficacy Questionnaire

MEIM-R

Multigroup Ethnic Identity Measure Revised

MLR

Maximum likelihood with robust standard errors

RMSEA

Root Mean Square Error of Approximation

SRMR

Standardized Root Mean Square Residual

STI

Sexually Transmitted Infection