A Longitudinal Study of Consumer Preference Drift in Deep-Dish Versus Thin-Crust Paradigms

1. Introduction

The question of why adults develop an apparent appreciation for deep-dish pizza as they age — a phenomenon well-known to pizza marketers but never rigorously documented — has implications for product development, brand strategy, and, at a deeper level, our understanding of how human preference systems evolve across the lifespan. Thin-crust pizza is heavily preferred by younger consumers in cross-sectional survey data (Napolitano et al., 2010; de la Salsa & Umamida, 2013), while deep-dish pizza commands higher satisfaction ratings among consumers over 35 (American Pizza Association, 2021). Whether this reflects a genuine intra-individual shift in preference (a maturational effect), a cohort difference (older adults simply grew up in an era with more deep-dish exposure), or a selection artifact (thin-crust-preferring adults age out of survey participation for unknown reasons) cannot be determined from cross-sectional data.

The Pizza Research Institute launched the Deep-Dish vs. Thin-Crust Longitudinal Study (D2TC-LS) in 2009 to resolve this question. 3,400 participants were enrolled at age 18 and assessed annually for 12 years (through age 30), providing the first longitudinal dataset large enough to disentangle maturational, cohort, and period effects in pizza preference. We hypothesized that a genuine maturational shift exists — the Preference Maturation Hypothesis (PMH) — and that this shift is mediated by changes in satiety sensitivity, texture tolerance, and an as-yet-uncharacterized variable we initially labeled "willingness to commit to a meal."

2. Materials & Methods

Cohort Enrollment. 3,400 participants (mean age 18.2 ± 0.4 years at enrollment; 51% female, 49% male; recruited from university first-year programs at six partner institutions) completed baseline assessment in 2009. Annual retention incentives included a $30 grocery gift card and, beginning in Year 4, complimentary access to PRI pizza tasting events, which substantially improved retention among participants who had initially indicated they were "not that into pizza, honestly."

Annual Assessment Protocol. Each annual visit included: (1) a standardized pizza preference survey (PPS-20, 20 items assessing preference across style, crust type, topping density, and general pizza engagement), (2) a blind taste evaluation of three standardized pizza samples (thin-crust, Neapolitan, deep-dish) prepared by PRI staff to laboratory standards, and (3) a structured interview exploring changes in pizza consumption contexts (eating alone vs. social eating, takeout vs. delivery vs. restaurant, etc.). Beginning in Year 6, a subsample (n = 612) also completed a food neophobia scale and a texture sensitivity battery, based on emerging evidence from the sensory science literature.

Attrition and Missing Data. 12-year retention was 82.1% (n = 2,790 completing all 12 annual assessments), exceptional for a longitudinal study of this duration and arguably a testament to the motivational power of free pizza. Attrition was slightly higher among participants who relocated internationally (n = 147 lost to follow-up) and among one participant who simply stopped responding in Year 7 and later sent a card explaining that they had "become more of a pasta person." Missing data were handled via multiple imputation (m = 20 imputations) under the missing-at-random assumption.

This study was approved by PRI IRB #IRB-2008-PZZ-001 and has been renewed annually for 14 years, making it the longest-running active protocol in PRI history. Participants provided written consent at enrollment and re-consent at Year 6.

3. Results

Multilevel growth curve modeling revealed a statistically significant positive trajectory in deep-dish preference across ages 18–30 (γ_10 = 0.31 per year, 95% CI [0.27, 0.35], p < 0.0001), after controlling for cohort effects (non-significant) and period effects (marginally significant in 2020, attributed to pandemic-related changes in pizza consumption context, discussed below). In contrast, thin-crust preference showed a statistically significant negative trajectory (γ_10 = −0.18 per year, 95% CI [−0.22, −0.14], p < 0.0001).

Importantly, the deep-dish preference increase was not linear: latent class growth analysis identified three trajectory subgroups. "Early Drifters" (18% of cohort) showed rapid deep-dish preference increase beginning at age 20–21. "Standard Maturers" (61%) showed the modal pattern of gradual increase beginning around age 24. "Committed Thin-Crusters" (21%) showed no significant deep-dish preference increase across the 12-year observation period, though their thin-crust scores did decrease slightly (γ_10 = −0.09, p = 0.04), a finding the research team found poignant.

The subsample texture sensitivity analysis (n = 612) revealed that deep-dish preference trajectory was moderated by texture sensitivity (β = −0.18 per SD increase in sensitivity, p = 0.002): participants with higher texture sensitivity showed slower preference drift. Neophobia was not a significant moderator (p = 0.31). The "willingness to commit to a meal" construct, formalized as the Meal Commitment Index (MCI), was a significant predictor of deep-dish preference at all time points and of preference drift rate (β = 0.24, p < 0.0001), suggesting that psychological factors — specifically, one's relationship with portion size and dining duration — mediate the maturational shift.

4. Discussion

The D2TC-LS provides the first longitudinal evidence for genuine intra-individual preference maturation in the deep-dish versus thin-crust domain, supporting the Preference Maturation Hypothesis over purely cohort-based explanations. The three-trajectory solution — Early Drifters, Standard Maturers, and Committed Thin-Crusters — suggests that the PMH does not describe a universal developmental process but rather a modal one, with meaningful individual variation in trajectory timing and magnitude.

The Meal Commitment Index as a mediator of preference drift is a theoretically novel finding. We interpret this as evidence that deep-dish pizza preference is partly a function of one's psychological orientation toward abundance, duration, and commitment — qualities that developmental psychology associates with adult identity formation. We recognize this is a strong claim and invite independent replication, ideally from a research group that does not have "Pizza" in its institutional name.

The 2020 period effect (a temporary increase in deep-dish preference across all subgroups, attributable to pandemic-related shifts toward comfort food and home consumption) is consistent with contextual theories of food preference and suggests that preference scores are not fixed traits but respond dynamically to environmental circumstances. Whether this effect persisted beyond our observation window is beyond the scope of the current data.

The "Committed Thin-Crusters" warrant particular attention from a consumer research perspective. Their resistance to preference drift across 12 years, in the face of sustained exposure opportunities and increasing age, suggests a level of pizza identity salience that may render them impervious to marketing. We regard this as, in some sense, admirable.

5. Conclusion

A genuine, intra-individual maturational shift toward deep-dish pizza preference occurs in the majority of adults between ages 18 and 30. This shift is mediated by the Meal Commitment Index and moderated by texture sensitivity. Three distinct preference trajectory subgroups exist. The Preference Maturation Hypothesis is supported by the data. Future work should extend the observation window beyond age 30 to determine whether the drift continues, plateaus, or — as one member of our advisory board suggested, with evident alarm — reverses.

Acknowledgments

The authors thank the 3,400 D2TC-LS participants for their remarkable loyalty to the study over 12 years and, in particular, for continuing to participate after Year 4 when the gift card value was reduced by $5 due to budget constraints. The PRI Data Management team maintained the study database for 14 years without a significant incident, which the authors consider remarkable. Funding was provided by the PRI Longitudinal Research Endowment (established 1997) and a 2014 supplemental grant from the National Pizza Council, whose unrestricted gift we disclose in the interest of transparency. The National Pizza Council had no role in study design, data collection, analysis, interpretation, or the writing of this report.

References

1. [1]

   American Pizza Association. (2021). State of the American pizza industry: Annual report. American Pizza Association.

2. [2]

   de la Salsa, R., & Umamida, H. (2013). Age-stratified pizza preference in four national markets: A cross-sectional analysis. Journal of Consumer Flavor Perception, 2(1), 12–31.

3. [3]

   Marinara, T., & Cheeseberg, M. (2022). Pepperoni placement and its statistically significant effect on flavor distribution. Journal of Applied Pizzology, 21(3), 188–209. https://doi.org/10.1883/jap.2022.034

4. [4]

   Napolitano, F., Romano, G., & Forno, L. (2010). Cross-sectional survey of pizza preference by age and region in the northeastern United States. Journal of Applied Pizzology, 9(2), 44–61.

5. [5]

   Napolitano, F., Romano, G., & Saucington, A. (2018). Development and validation of the Consumer Satisfaction Index (CSI). Annals of Culinary Chemistry, 12(3), 201–217.

6. [6]

   Romano, G., & Saucington, A. (2023). Structural integrity of pizza crust under variable topping load conditions. Journal of Applied Pizzology, 22(3), 114–131. https://doi.org/10.1883/jap.2023.017

7. [7]

   Saucington, A. (2021). Sourdough fermentation duration and its effect on pizza dough extensibility. Journal of Applied Pizzology, 20(1), 1–19. https://doi.org/10.1883/jap.2021.003