Blended Wines: A Review of Chemical, Sensory, and Biological Perspectives

Abstract

Blending has long been an integral part of winemaking practice across many wine-producing regions. Through different blending strategies, monovarietal (base) wines are combined to produce a final product, commonly referred to as a blended wine (coupage, assemblage, cuvée, wine blend, blend) with targeted sensory and quality attributes. Depending on the stage at which it is applied, either before or after fermentation, blending can be used to adjust complexity, balance, and consistency, while also fulfilling economic and branding objectives within winery practice. Beyond its practical applications, blended wines have increasingly drawn scientific interest. This review surveys existing literature on the still wine blends produced from Vitis vinifera grapes, with a primary focus on chemical and sensory aspects, and a more limited discussion of their reported biological activities.

1. Introduction—Wine Blending: Definition and Purpose

Historically, wine production in many regions relied primarily on blending, with wines assembled from different grape varieties, vineyard parcels, or vintages in order to achieve balance and consistency [1,2]. Within regulatory and technical frameworks, the International Organisation of Vine and Wine [3] defines blending as the combination of different wines to obtain a final product with specific, desired characteristics. In contrast to single-varietal wines, which are composed of at least 75% [4] to 85% [5,6] of a single grape variety, blended wines (also referred to as coupage, assemblage, cuvée, wine blends, or blends) are produced by combining monovarietal wines (base wines) originating from different grape varieties, terroirs, or vintages. The term blend may also refer to a wine obtained by combining several wines of the same grape variety, from the same or different vintages, resulting in a single-varietal wine produced through a blending operation. Therefore, according to the OIV, blended wine refers to the final product obtained by combining different base wines. In practice, however, this definition encompasses a range of technological routes that differ substantially in timing and mechanism. Such products may result from different blending strategies, including post-fermentation blending of finished wines, co-fermentation of different grape varieties or lots, and related practices such as co-maceration or grape juice blending. While these approaches differ technologically, they all yield wines that can be classified as blended wines based on their composite origin [7,8].

Some of the most recognized traditional blended wines include the Bordeaux blend, typically based on Cabernet Sauvignon, Merlot, and Cabernet Franc, with possible inclusion of Petit Verdot [9]; the Rioja blend, typically based on Tempranillo, with Garnacha, Graciano, and Mazuelo as traditional blending components defined within the Rioja DOCa specification [10]; and the Cape blend, commonly based on Pinotage, and often combined with international varieties such as Cabernet Sauvignon, Merlot, or Shiraz [11,12]. These classic blends illustrate how regional practices and grape selection shape characteristic sensory profiles and wine typicity.

Throughout this review, the term blended wine is used to describe the final product, while blending strategies refer to the technological approaches by which it is produced. Blending is employed to achieve several important goals. Firstly, it enhances the quality of wine by optimizing color, aroma, and taste by integrating complementary varietal characteristics. Secondly, it ensures consistency, allowing for the maintenance of uniform sensory profiles across various production batches and vintages. Additionally, blending contributes to complexity by adding depth to flavor and aroma through combining different wine components. Finally, it promotes balance by adjusting the tannin structure, acidity, and sweetness, resulting in a well-rounded final product [13,14,15]. In addition, climate change is increasingly recognized as a factor influencing grape composition and, consequently, wine typicity, particularly in classical wine regions such as Bordeaux, where temperature-driven shifts affect key compositional and sensory attributes [16]. These changes implicate the importance of blending as a strategy to mitigate compositional variability and maintain typicity under evolving climatic conditions. More recently, blending strategies have also been explored in response to increasing consumer interest in wines with reduced alcohol content. Beyond traditional varietal blending, such approaches may involve combining wines, grape juices or must obtained from grapes harvested at different maturity levels in order to modulate alcohol content while maintaining sensory balance [8].

Not all wines are suitable as single-varietal expressions (e.g., wines with excessive or insufficient ethanol content, excessive acidity, low aromatic intensity, etc.) and may require strategic quality enhancement; thus, blending serves various economic purposes. These include the development of winery-specific wines to enhance brand identity, recognition, and typicity [9,14] or optimization of grape variety utilization and production costs, such as blending of wines of lower quality/price with higher-quality wines [15,17,18]. Earlier projections suggested that red wine blends would become increasingly popular among younger generations who prefer higher-priced wines with more interesting and character-rich sensory profiles [19]. Furthermore, blending is strategically employed to ensure the final product adheres to legal and regulatory standards, particularly regarding physicochemical limits (e.g., alcohol content, volatile acidity, SO₂) and regional compositional requirements [3,4,5,20].

Beyond the strong interest of wine professionals in blended wines, these wines have also garnered attention from the scientific community. As evidenced in the following review, research on blends typically encompasses physicochemical and chemical characterization, sensory evaluation, and, to a lesser extent, investigations into their biological activity. Research in this field seeks to elucidate the impact of blending on critical wine attributes, primarily its organoleptic properties and chemical composition, their interrelationships, and, to a lesser extent, its potential health-related effects. This multidisciplinary approach is crucial for a comprehensive understanding of both the scientific and practical aspects of wine blending, bridging the gap between traditional winemaking practices and modern analytical advancements.

This review surveys existing literature on the chemical composition of blended wines, emphasizes the role of sensory perception in evaluating blends, and briefly addresses computer-based sensory analysis and biological activity. The focus is placed on still wine blends produced from Vitis vinifera grapes, although it is acknowledged that certain sparkling wines, such as Champagne, are also obtained through blending of different base wines. Accordingly, the review aims not only to synthesize existing knowledge on wine blending, but also to highlight key chemical, sensory, and methodological principles that can inform targeted blend design.

2. Chemical Determinants Underlying the Sensory Characteristics of Blended Wines

2.1. Polyphenols in Blended Wines

Polyphenols represent a major class of secondary metabolites in grapes and wine and include both flavonoid compounds, such as flavan-3-ols, tannins, flavonols, and anthocyanins, and non-flavonoid constituents, including phenolic acids and stilbenes. During winemaking, polyphenols undergo extensive chemical and structural transformations and are therefore frequently used as indicators of wine quality and authenticity. Polyphenolic composition has been linked to specific sensory attributes, most notably colour and mouthfeel; however, direct causal relationships remain complex and only partially understood [21]. Representative examples of total polyphenol, flavonoid, anthocyanin, and tannin contents in selected wine blends are presented in

Table 1. Representative ranges of total polyphenol (TPC), flavonoid (TFC), anthocyanin (TAC), and tannin (TTC) contents in example wine blends.

Number and Type of Blends (Blending Strategies), * Base Wines (Blend Composition)	Origin, (Vintage)	Total Polyphenols, Flavonoids, Anthocyanins and Tannin Levels	References
6 binary (co-blended) Cabernet Sauvignon Carménère Merlot Cabernet Franc	Maipo Valley, Chile, (2010)	TPC: 857.9–924.5 mg GAE/L ** TTC: 1701–1992 mg CE/L ** TAC: 426.7–533.3 mg ME/L **	[22]
14 ternary (co-blended) Sauvignon Blanc Chenin Blanc Thompson Seedless Italia Shiraz Ruby Red Bangalore Blue	Andhra Pradesh, India, (2008, 2009)	TPC: 221.6–607.0 mg/L TTC: 60–400 mg/L TFC: 95.34–280.1 mg/L	[23]
24 binary (co-fermented) Muscat Bailey A (various manufacturing methods) Campbell Early Gerbong	Yongdong, Korea, (2008)	TPC: 703.0–1417 mg/L	[24]
4 binary (co-blended) Cabernet Sauvignon Merlot	Ćemovsko polje, Montenegro, (2015, 2016)	TPC: 1809–1926 mg GAE/L TTC: 863–1151 mg CE/L TFC: 52.87–55.87 mg QE/L ** TAC: 23.75–30.72 mg ME/L	[25]

* The number of blends indicates the total number of experimental blend samples, while the terms binary and ternary refer to blends composed of two or three base wines, respectively; the blending strategy (e.g., co-blending or co-fermentation) is indicated in parentheses. ** CE: (+)-catechin equivalent; GAE: gallic acid equivalents; ME: malvidin equivalent; QE—quercetin equivalent.

, with studies selected based on the availability of detailed and comparable analytical data rather than geographic representativeness. Accordingly, the table is not intended to reflect the distribution of wine production across regions, but to illustrate compositional variability reported in the literature [22,23,24,25]. It should be noted that the values reported in Table 1 are not intended for direct comparison with corresponding monovarietal wines, as such comparative data were not consistently available across studies. Accordingly, the table reflects representative concentration ranges rather than direct evidence of compositional changes attributable to blending.

Several studies have examined the polyphenolic composition of blended wines in greater analytical detail. For example, Cáceres-Mella et al. [22] applied HPLC-DAD to quantify low-molecular-weight phenolic compounds and anthocyanins in blended wines, reporting that flavonoids constituted the dominant fraction (66.1–83.6%), whereas non-flavonoids accounted for 16.4–33.9%. Flavanols were the most abundant flavonoids subclass across all wine samples, while monoglucoside forms accounted for the majority of total anthocyanins (81.2–84.3%). Generally, only blends with polyphenol-rich modifiers (e.g., Cabernet Franc, Carménère) showed additive effects, while varieties like Cabernet Sauvignon with a comparatively lower-content of polyphenolic compounds commonly reduced overall phenolic content and masked compositional differences. Li et al. [26] analyzed blends of base wines Cabernet Franc and Cabernet Sauvignon with modifier wines Marselan and Petit Verdot for their phenolic composition. They detected 21 non-anthocyanin phenolic compounds, 15 grape-derived anthocyanins, and 33 anthocyanin derivatives. Adding modifier wines improved the colour characteristics of base wines across different oxygen exposure levels, suggesting that the formation of anthocyanin derivatives and the resulting chromatic properties are modulated by both the phenolic composition of the blending components and the oxidative environment during wine aging. HPLC-UV/VIS analysis of selected phenolic acids, flavonoids, stilbenes, and anthocyanins demonstrated that the predominant polyphenols in Cabernet Sauvignon and Merlot blends were gallic acid and catechin. Among anthocyanins, malvidin-3-O-glucoside was the most abundant. Interestingly, resveratrol and delphinidin-3-O-glucoside were not detected in the commercially available blend, despite their prior identification in the corresponding base wines [25]. This absence may reflect matrix effects, differences in stability, or detection limits rather than true compositional loss. More focused analyses were also conducted, including the quantification of resveratrol in Barbera and Croatina blends [27], as well as intensive research of anthocyanins in Tempranillo and Graciano blends [17,28,29,30].

Overall, available evidence indicates that the dominant polyphenols present in blended wines largely mirror those found in monovarietal wines, with anthocyanins receiving the greatest research attention due to their central role in colour expression and consumer appeal. Despite extensive compositional characterization, current evidence does not support the existence of a unique phenolic signature specific to blended wines; rather, blending alters the relative contribution and dominance of existing phenolic drivers, resulting in context-dependent and often non-linear outcomes. From a practical blending perspective, phenolic composition offers a useful framework for selecting base and modifier wines, particularly in relation to colour stability, mouthfeel, and ageing potential. Differences in anthocyanin structure, tannin composition, and polymerisation behaviour underpin many of the non-linear effects observed in blends, highlighting that blending outcomes cannot be predicted simply by averaging compositional parameters but instead require consideration of dominant phenolic drivers and matrix effects.

2.2. Volatile Compounds in Blended Wines

Aroma perception in blended wines is shaped by complex interactions among volatile compounds, which has motivated numerous studies aimed at understanding how blending modifies sensory expression. Although the concentrations of volatile compounds in blends often change in a roughly proportional manner, aroma perception is far less predictable due to non-linear sensory interactions among volatiles [31]. Changes in aromatic profiles, most frequently investigated using gas chromatography-mass spectrometry (GC-MS) coupled with headspace solid-phase microextraction (HS-SPME), have been shown to arise not only from the blending process itself but also from subsequent wine maturation. For example, Kovačević Ganić et al. [15] reported that chromatographic profiles of Malvasia Istriana blends closely resembled the base wine during the initial three months of aging, after which gradual divergence became evident. The observed divergence in volatile profiles over time suggests that blending not only affects initial aroma composition but also modulates its temporal evolution during aging. In the study by Ling et al. [31], 91 volatile compounds were identified, and it was demonstrated that blending not only altered their concentrations but also influenced the evolution of volatiles during aging, resulting in divergent volatile compound profiles across varying blend proportions, and following six months of wine aging. Compared to wine blending, the co-fermentation of Merlot and Cabernet Sauvignon grapes generally resulted in greater aromatic complexity and elevated concentrations of volatiles, including ethyl acetate, isoamyl acetate, isopentyl octanoate, phenethyl acetate, ethyl laurate, isoamyl alcohol, 2,3-butanediol, and phenylethyl alcohol [32]. Recent study [33] has shown that multi-varietal co-fermentation (Cabernet Sauvignon with Marselan, Merlot, or Cabernet Gernischt) generates distinct microbial communities depending on the grape varieties interaction and fermentation environment, leading to differences in volatile compound production and aroma profiles. For example, enrichment of lactic acid bacteria during co-fermentation of Cabernet Sauvignon and Merlot has been linked to increased levels of esters (e.g., isoamyl acetate, ethyl hexanoate) and terpenes (e.g., linalool, geraniol), enhancing aromatic complexity, while dominance of other microbial groups such as acetic acid bacteria in Cabernet Sauvignon-Marselan co-fermentation may favor the formation of higher alcohols (e.g., isoamyl and phenethyl alcohol), emphasizing the role of microbial interactions in modulating wine sensory properties.

On the other hand, lower-alcohol wines are frequently associated with reduced aroma intensity and the persistence of herbaceous notes, which can negatively influence consumer perception [8,34,35]. To address these sensory challenges, Longo et al. [34] evaluated monovarietal blends combining wines from less ripe and riper grapes of Verdelho and Petit Verdot. The most abundant aroma compounds were higher alcohol acetates. At low concentrations, these compounds enhance the aromatic complexity of wine, but elevated levels may obscure distinct varietal characteristics. Schelezki et al. [8] also examined changes in the volatile profile while investigating strategies for producing lower-alcohol Cabernet Sauvignon wines. Additionally, the volatile profile may serve as a basis for differentiating blends from other categories, such as rosé or claret wines [36].

In addition to the widely used GC–MS-based approaches for volatile profiling, a range of advanced analytical techniques has been applied to characterize, differentiate, and authenticate wine blends. These include nuclear magnetic resonance (NMR) spectroscopy [37], spectrofluorometric analysis [38], hybrid electronic tongues based on arrays of electrochemical microsensors [39], molecular techniques such as high-resolution melting (HRM) analysis of single-nucleotide polymorphisms (SNPs) [40], and synthetic receptors, including peptide-based sensing arrays [41].

Overall, blending often enhances aromatic complexity; however, sensory perception remains inherently unpredictable due to interactions among volatile compounds and the wine matrix. The volatile composition of blended wines does not represent a distinct chemical category but rather reflects dynamic interactions among aroma compounds and the surrounding wine matrix [42]. Thus, advancements in analytical methods are expected to offer deeper insight into these complex sensory interactions.

From an applied standpoint, these analytical techniques support more informed selection of blending components by linking compositional fingerprints with sensory outcomes. For example, GC–MS and HS-SPME [14,15,31,32] facilitate identification of aroma drivers and potential masking effects, while spectroscopic and sensor-based approaches offer rapid screening tools for blend differentiation and authenticity [38,39,41]. When combined with sensory data, these methods support more informed and rational blending decisions, moving beyond purely empirical trial-and-error approaches, particularly when combined with mixture design or multivariate modelling approaches [18,24,43,44,45].

3. Sensory Profile as a Key Characteristic of Blends

3.1. Sensory Evaluation of Blended Wines Produced Using Different Blending Strategies

Since blending primarily aims to enhance the sensory characteristics of base wines, it is unsurprising that much of the scientific research focuses on the sensory evaluation of both base and final wines, as summarized in Table 2, Table 3 and Table 4. It is important to emphasize that, given the inherent variability of blended wines and the diversity of sensory evaluation approaches, the conclusions presented in this review should be interpreted as general tendencies rather than definitive or broadly generalizable rules, and represent a synthesis of available findings across the literature.

The predominant approach in scientific studies involves blending monovarietal wines to assess sensory improvements or comparative differences between the base wines and the resulting blends, ever since the pioneering studies conducted by Singleton and Ouch [46] in the early 1960s. In this study, sensory analysis of blends from 34 pairs of commercially available, single-varietal wines in a 1:1 ratio, was done. Blends received higher quality ratings than the individual wines, surpassing even the better-rated wine in 7 cases. This fundamental methodology has continued to be applied in the last 25 years, with notable advancements in complexity. For example, to enhance the sensory quality of the Croatian wine Malvasia Istriana, enrich its aroma, and extend its shelf life while maintaining its varietal distinctiveness, researchers blended it with high-quality wines such as Chardonnay, Sauvignon Blanc, Pinot Blanc, Muscat, and Prosecco [15]. Blending Malvasia Istriana improved its sensory quality while preserving its varietal characteristics. Malvasia Istriana (85%), Sauvignon Blanc (7.5%), and Pinot Blanc (7.5%) blend was the most optimal, while Muscat was unsuitable for blending due to its dominant aroma overpowering the base wine. Hopfer et al. [14] examined the sensory and chemical characteristics of eleven binary and ternary blends of Cabernet Sauvignon, Merlot, and Cabernet Franc. The blending of wines induces modifications in both sensory and chemical attributes, often leading to more intricate profiles that surpass simple averages. While blends from “smooth” wines generally resemble their monovarietal counterparts, combining “uneven” wines, such as Cabernet Sauvignon, tends to produce more distinct blends with varying sensory qualities. The interaction of contrasting attributes within these blends can suppress or enhance specific characteristics in the final wine.

Table 2. Summary of sensory evaluations for different wine blends (sensory effects of wine blending assessed by trained panels and experts).

Base Wines (Blend Composition)	Number and Type of Blends (Blending Strategies) *	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
A. Co-blended wines
68 base wines	34 binary (co-blending)	California, USA (1960)	Trained panel (n = 10): 20-point scale	Basic composition	Blends were rated higher than the average of component wines; in several cases, blends exceeded the highest-rated base wine.	[46]
Malvasia Istriana Chardonnay Sauvignon Blanc Pinot Blanc Muscat Prosecco	3 binary (co-blending) 5 ternary (co-blending)	Poreč, Croatia, (2000)	Wine judges (n = 6): Buxbaum model of positive rating (colour, clearness, odour and taste)	Volatiles (HS-SPME followed by GC-FID/GC-MS)	Sensory quality improved while preserving varietal character; blends containing muscat showed dominance effects that reduced balance.	[15]
Cabernet Sauvignon Merlot Cabernet Franc	11 binary (co-blending) 4 ternary (co-blending)	California, USA (2009)	Trained panel (n = 14): descriptive analysis (aroma, taste and mouthfeel)	Basic composition Volatiles (HS-SPME followed by GC-MS) Total phenolics, anthocyanins (spectrophotometry)	Blending produced sensory profiles that deviated from simple averages; “smooth” wines tend to produce blends like their monovarietal base; blends of “uneven” base wines (e.g., Cabernet Sauvignon) showed enhanced differentiation and masking/amplification of specific attributes	[14]
Cabernet Sauvignon Carménère Merlot Cabernet Franc	6 binary (co-blending)	Maipo Valley, Chile, (2010)	Trained panel (n = 12): descriptive analysis (astringency and bitterness)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-DAD)	Blending did not significantly alter perceived astringency or bitterness; mouthfeel-related descriptors varied depending on modifier wine, indicating non-additive sensory interactions.	[22]
Cencibel Rojal Tortosí Moravia Dulce	4 binary (co-fermenting)	La Mancha, Spain	Trained panel (n = 15): descriptive analysis (odor, aroma)	Basic composition	Co-fermentation with minority varieties enhanced aromatic complexity and key odor attributes relative to monovarietal wines	[47]
Cabernet Sauvignon Merlot Cabernet franc Petit Verdot Fer Servadou Duras Manseng noir Vinhão Arinarnoa	11 quinary (co-blending)	Saint-Laurent Médoc, France (2018–2022)	Wine professionals (n = 37): typicity assessment (olfactive and gustative typicity) highly experienced professionals in wine blending (n = 20): typicity ranking		Perceived typicity showed some overall change; reductions were observed for blends containing higher proportions of certain non-traditional varieties while remaining within a narrow sensory variation range.	[9]
B. Diverse production technologies/blending strategies
Muscat Bailey A (various manufacturing methods) Campbell Early Gerbong	14 ternary (co-blending)	Yongdong, Korea, (2008)	Expert panellists (n = 30): 9-point hedonic scale (color, aroma, taste, balance, overall acceptability)	Total phenolics (spectrophotometry)	Mixture design enabled identification of an optimal blend with improved aroma, taste balance, and overall acceptability; sensory preference was aligned with higher total polyphenol content	[24]
Cabernet Sauvignon Merlot Marselan Syrah Pinot Noir	4 binary (co-blending) 4 binary (co-fermenting)	Xinjiang, China, (2021)	Trained panel (n = 30): 20-point scale (bitterness, astringency, viscosity, mouthfeel, fruity, floral, and herbal notes)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-UV/VIS) Volatiles (SPME-GC-MS) Antioxidant activity (spectrophotometry)	Co-fermentation enhanced fruity aroma and mouthfeel smoothness compared with post-fermentation blends	[48]
Cabernet Sauvignon Merlot	3 binary (co-fermenting) 3 binary (co-blending)	Xinjiang, China, (2024)	Panel (n = 30): 25-point scale (bitterness, astringency, viscosity, mouthfeel, fruity, floral, and herbal notes)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-UV/VIS) Volatiles (SPME-GC-MS) Antioxidant activity (spectrophotometry)	Co-fermentation, particularly at 80:20 ratio, intensified aroma, mouthfeel, and fruit perception	[32]
Malbec Merlot	1 binary (co-fermenting) 1 binary (postalcoholic fermentation) 1 binary (postmalolactic fermentation)	Paso Robles, California, USA, (2018)	Trained panel (n = 10): descriptive analysis	Basic composition and color Anthocyanins, total polymeric pigments, tannins (spectrophotometry) Monomeric anthocyanins, anthocyanin- derived pigments, flavonols (HPLC-DAD)	Blending immediately after alcoholic fermentation better preserves the distinct sensory characteristics of monovarietal wines; co-fermentation and post-malolactic blending tend to homogenize flavors, though both methods enhance overall wine complexity.	[7]
Sauvignon Blanc Chenin Blanc Thompson Seedless Italia Shiraz Ruby Red Bangalore Blue	24 binary (co-fermenting)	Andhra Pradesh, India (2008, 2009)	Panel (n = 10): 20 point scale (colour, aroma, flavour, taste, astringency, overall acceptability)	Basic composition Total phenolics, tannins, flavonoids (spectrophotometry)	Blending resulted in variable sensory quality depending on grape combination; blends with white base wines and red modifiers showed improved colour and overall acceptability, while some combinations led to diminished balance.	[23]

* The number of blends indicates the total number of experimental blend samples, while the terms binary, ternary and quinary refer to blends composed of two, three or five base wines, respectively; the blending strategy (e.g., co-blending or co-fermentation) is indicated in parentheses.

Table 2. Summary of sensory evaluations for different wine blends (sensory effects of wine blending assessed by trained panels and experts).

Base Wines (Blend Composition)	Number and Type of Blends (Blending Strategies) *	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
A. Co-blended wines
68 base wines	34 binary (co-blending)	California, USA (1960)	Trained panel (n = 10): 20-point scale	Basic composition	Blends were rated higher than the average of component wines; in several cases, blends exceeded the highest-rated base wine.	[46]
Malvasia Istriana Chardonnay Sauvignon Blanc Pinot Blanc Muscat Prosecco	3 binary (co-blending) 5 ternary (co-blending)	Poreč, Croatia, (2000)	Wine judges (n = 6): Buxbaum model of positive rating (colour, clearness, odour and taste)	Volatiles (HS-SPME followed by GC-FID/GC-MS)	Sensory quality improved while preserving varietal character; blends containing muscat showed dominance effects that reduced balance.	[15]
Cabernet Sauvignon Merlot Cabernet Franc	11 binary (co-blending) 4 ternary (co-blending)	California, USA (2009)	Trained panel (n = 14): descriptive analysis (aroma, taste and mouthfeel)	Basic composition Volatiles (HS-SPME followed by GC-MS) Total phenolics, anthocyanins (spectrophotometry)	Blending produced sensory profiles that deviated from simple averages; “smooth” wines tend to produce blends like their monovarietal base; blends of “uneven” base wines (e.g., Cabernet Sauvignon) showed enhanced differentiation and masking/amplification of specific attributes	[14]
Cabernet Sauvignon Carménère Merlot Cabernet Franc	6 binary (co-blending)	Maipo Valley, Chile, (2010)	Trained panel (n = 12): descriptive analysis (astringency and bitterness)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-DAD)	Blending did not significantly alter perceived astringency or bitterness; mouthfeel-related descriptors varied depending on modifier wine, indicating non-additive sensory interactions.	[22]
Cencibel Rojal Tortosí Moravia Dulce	4 binary (co-fermenting)	La Mancha, Spain	Trained panel (n = 15): descriptive analysis (odor, aroma)	Basic composition	Co-fermentation with minority varieties enhanced aromatic complexity and key odor attributes relative to monovarietal wines	[47]
Cabernet Sauvignon Merlot Cabernet franc Petit Verdot Fer Servadou Duras Manseng noir Vinhão Arinarnoa	11 quinary (co-blending)	Saint-Laurent Médoc, France (2018–2022)	Wine professionals (n = 37): typicity assessment (olfactive and gustative typicity) highly experienced professionals in wine blending (n = 20): typicity ranking		Perceived typicity showed some overall change; reductions were observed for blends containing higher proportions of certain non-traditional varieties while remaining within a narrow sensory variation range.	[9]
B. Diverse production technologies/blending strategies
Muscat Bailey A (various manufacturing methods) Campbell Early Gerbong	14 ternary (co-blending)	Yongdong, Korea, (2008)	Expert panellists (n = 30): 9-point hedonic scale (color, aroma, taste, balance, overall acceptability)	Total phenolics (spectrophotometry)	Mixture design enabled identification of an optimal blend with improved aroma, taste balance, and overall acceptability; sensory preference was aligned with higher total polyphenol content	[24]
Cabernet Sauvignon Merlot Marselan Syrah Pinot Noir	4 binary (co-blending) 4 binary (co-fermenting)	Xinjiang, China, (2021)	Trained panel (n = 30): 20-point scale (bitterness, astringency, viscosity, mouthfeel, fruity, floral, and herbal notes)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-UV/VIS) Volatiles (SPME-GC-MS) Antioxidant activity (spectrophotometry)	Co-fermentation enhanced fruity aroma and mouthfeel smoothness compared with post-fermentation blends	[48]
Cabernet Sauvignon Merlot	3 binary (co-fermenting) 3 binary (co-blending)	Xinjiang, China, (2024)	Panel (n = 30): 25-point scale (bitterness, astringency, viscosity, mouthfeel, fruity, floral, and herbal notes)	Basic composition and color Total phenolics, flavonoids, anthocyanins (spectrophotometry) Phenolics (HPLC-UV/VIS) Volatiles (SPME-GC-MS) Antioxidant activity (spectrophotometry)	Co-fermentation, particularly at 80:20 ratio, intensified aroma, mouthfeel, and fruit perception	[32]
Malbec Merlot	1 binary (co-fermenting) 1 binary (postalcoholic fermentation) 1 binary (postmalolactic fermentation)	Paso Robles, California, USA, (2018)	Trained panel (n = 10): descriptive analysis	Basic composition and color Anthocyanins, total polymeric pigments, tannins (spectrophotometry) Monomeric anthocyanins, anthocyanin- derived pigments, flavonols (HPLC-DAD)	Blending immediately after alcoholic fermentation better preserves the distinct sensory characteristics of monovarietal wines; co-fermentation and post-malolactic blending tend to homogenize flavors, though both methods enhance overall wine complexity.	[7]
Sauvignon Blanc Chenin Blanc Thompson Seedless Italia Shiraz Ruby Red Bangalore Blue	24 binary (co-fermenting)	Andhra Pradesh, India (2008, 2009)	Panel (n = 10): 20 point scale (colour, aroma, flavour, taste, astringency, overall acceptability)	Basic composition Total phenolics, tannins, flavonoids (spectrophotometry)	Blending resulted in variable sensory quality depending on grape combination; blends with white base wines and red modifiers showed improved colour and overall acceptability, while some combinations led to diminished balance.	[23]

* The number of blends indicates the total number of experimental blend samples, while the terms binary, ternary and quinary refer to blends composed of two, three or five base wines, respectively; the blending strategy (e.g., co-blending or co-fermentation) is indicated in parentheses.

To evaluate potential changes in mouthfeel characteristics, particularly astringency and bitterness, blends of Cabernet Sauvignon with Carménère, Merlot, and Cabernet Franc at different ratios were investigated [22]. No significant differences in these attributes were observed between the varietal wines and their corresponding blends. A representative example of blended wine production using a co-fermentation-based blending strategy is provided in the study by Sánchez-Palomo et al. [47]. They reported that co-fermenting minority grape varieties Tortosí, Rojal, and Moravia Dulce with Cencibel in a 1:1 ratio enhanced the key aromatic attributes of the monovarietal wines; the co-fermented wines exhibited a more complex aroma profile. One of the most recent and notable studies examined the introduction of new grape varieties into the classical, famous blend Bordeaux (Cabernet Sauvignon, Merlot, Cabernet Franc, and Petit Verdot). The impact of incorporating five red grape varieties (Fer Servadou, Duras, Manseng Noir, Vinhão, or Arinarnoa) at 10% and 30% was evaluated. According to assessments by trained professionals, changes in perceived typicity were minimal overall, except when Vinhão was included. However, highly experienced professionals in wine blending observed a reduction in typicity when Fer Servadou (10% or 30%), Manseng noir (30%), or Vinhão (30%) were incorporated into the Bordeaux reference blend [9].

A more complex approach involves the application of diverse blending strategies and production methodologies. In the study of Korean wines [24], one of the base wines (Muscat Bailey A) was produced using different manufacturing methods: fermentation with and without bacteria at various temperatures, followed by aging with American or French medium and heavy toast oak. Its quality was subsequently optimized through blending with Gerbong and Campbell Early wines. Using a mixture design combined with graphical and numerical optimization techniques, the optimal blend was determined to consist of 40% Muscat Bailey A, 48.9% Campbell, and 11.1% Gerbong wines, with a desirability score of 90.4%, based on total polyphenol content, aroma and taste preference. The effects of various production methodologies on the physicochemical properties, polyphenolic composition, volatile profiles, and sensory attributes of the resulting wines were also investigated. In one study [48], Cabernet Sauvignon was modulated through blending with Merlot, Marselan, Syrah, and Pinot Noir, while another research by the same group [32] focused on enhancing Cabernet Sauvignon wine quality by incorporating Merlot, examining both co-fermentation of grapes and post-fermentation blending approaches. Co-fermentation of grapes resulted in stronger fruity aromas and general aroma complexity than base wine blending, while Cabernet Sauvignon-Merlot and Cabernet Sauvignon-Pinot Noir wines obtained after grapes co-fermentation had the smoothest taste among all samples [48]. At the same time, co-fermentation significantly enhanced wine aroma, astringency, mouthfeel, and fruit flavor, particularly in the wine prepared from Cabernet Sauvignon and Merlot co-fermented grapes in ratio 80:20 [32]. A similar methodology was applied in producing Malbec–Merlot blends, prepared by co-fermentation, post-alcoholic and post-malolactic fermentation blending, at a 50:50 ratio [7]. Wines produced by co-fermentation and post-malolactic blending exhibited greater complexity, whereas post-alcoholic blending preserved the distinctive characteristics of each monovarietal wine. In addition, some studies considered biochemical and sensory properties of blends obtained by mixing juices of white and red grape varieties, such as Sauvignon Blanc, Chenin Blanc, Thompson Seedless, and Italia with Shiraz, Ruby Red, and Bangalore Blue, respectively. Blended wines exhibited varying quality, with Chenin Blanc (66.6%)—Shiraz (33.3%) achieving the highest score (17.23) and Italia (66.6%)—Ruby Red (33.3%) the lowest (11.96); five blends were rated as good, while most others were classified as fair [23].

It should be noted that the sensory outcomes summarized in Table 2, but Table 3 and Table 4 are also strongly influenced by methodological factors, including panel composition and size, assessor training level, scaling systems, and experimental design [2,49]. Considerable heterogeneity exists among studies, ranging from small trained panels using descriptive analysis to consumer-based hedonic evaluations, which limits direct quantitative comparison across datasets. In addition, differences in blending stage (co-fermentation vs. post-fermentation), blend ratios, and wine matrix composition further modulate sensory perception and may confound interpretation of blending effects. Importantly, many of these non-linear outcomes arise from perceptual sensory interactions, including masking, suppression, and enhancement among aroma, taste, and mouthfeel attributes, which explain why sensory responses of blends often deviate from simple additive behavior. Despite these limitations, Table 2, Table 3 and Table 4 reveals consistent qualitative trends, notably that blending can modulate aroma expression and mouthfeel attributes through sensory interactions rather than simple additive effects. These observations highlight the need for more standardized sensory protocols and integrative experimental designs to enable more robust cross-study comparisons.

3.2. Sensory Evaluation in Wine Blending: Methodologies, Challenges, and Standards

To better interpret these sensory outcomes, it is important to consider the methodologies used in sensory evaluation. A large share of wines produced worldwide are blended in order to refine attributes such as colour, alcohol content, mouthfeel (i.e., perceived weight and viscosity), and aroma. While adjustments to physicochemical and chemical parameters are primarily guided by analytical methods, the optimization of sensory attributes (visual, olfactory, tactile, and taste) relies exclusively on sensory evaluation. This process is inherently complex and requires both methodological rigor and experienced assessors [43]. Interpretation of sensory data can vary considerably among consumers, wine producers, and researchers, reflecting differences in experience, training, and expectations [2,50]. Despite the availability of international guidelines (ISO) [51], the degree of standardization in sensory analysis and assessor training remains variable across studies and institutions [50,52,53]. Fundamental tasting procedures assess visual, aromatic, taste, and mouthfeel parameters sequentially, with further distinctions applied to each parameter as necessary, contributing to flavour perception as a multisensory experience [2,53].

In research settings, sensory evaluations are most often conducted by trained wine tasters whose skills are continuously refined through calibration and repeated practice. Blending studies commonly employ a combination of discrimination tests (e.g., triangle and duo–trio tests), descriptive analysis, and hedonic evaluation, depending on the study objective [2,49]. Standardized discrimination methods, including the triangle test (ISO 4120:2021), duo–trio test (ISO 10399:2017), and ranking test (ISO 8587:2006/Amd 1:2013), are frequently used to determine whether blends differ perceptibly from their base wines [49,50,52].

From a practical perspective, different sensory methodologies serve distinct purposes within blending studies. Discrimination tests are commonly applied to assess whether blending results in perceptible differences, descriptive analysis is used to characterize modulation of specific sensory attributes (e.g., astringency or aroma balance), while hedonic testing is typically employed when consumer acceptance is the primary focus [2,49]. Taken together, these approaches highlight that sensory methods function as complementary decision-making tools within a blending workflow rather than as isolated or interchangeable evaluations.

3.3. Blending Strategies for Lower-Alcohol and Consumer-Preferred Wines: Sensory and Market-Driven Perspectives

The growing interest among quality- and lifestyle-oriented consumers in wines with reduced alcohol content has also influenced blending practices and highlighted the need for optimization in production [8]. As presented in Table 3, Longo et al. [34] investigated the blends composed of monovarietal wines produced from Verdelho and Petit Verdot less ripe and riper grapes, to achieve lower alcohol levels while enhancing ripe fruit flavor expression. Blending wines obtained from grapes harvested at different maturity stages can result in moderate reductions in ethanol content. In the case of Verdelho, blends prepared from first- and second-harvest base wines showed an alcohol content of 8.8 ± 0.1% (v/v), compared to 7.2 ± 0.1% and 10.3 ± 0.1% (v/v) for the respective base wines. Similarly, for Petit Verdot, the resulting blends reached 11.0 ± 0.2% (v/v), while the corresponding base wines had alcohol levels of 9.3 ± 0.1% and 12.6 ± 0.1% (v/v). These blends, composed of equal proportions of wines produced from less ripe and fully mature grapes, exhibited sensory profiles comparable to those of wines made exclusively from riper fruit, with unripe sensory attributes such as ‘herbaceous’ (Verdelho), ‘tomato leaf,’ and ‘green pepper’ (Petit Verdot) being less pronounced. Schelezki et al. [8] partially substituted juice from grapes, harvested at the commercial maturity date, with either water or juice from earlier-harvested grapes, before fermentation. Alcohol content in wines produced from earlier-harvested grapes to the last (overripe) harvest stage (control) ranged from 4.5% to 18.1% (v/v), whereas blending treatments yielded wines with alcohol levels of approximately 14.5–17% (v/v). Substitution with water preserved favorable aroma and flavor characteristics; however, overripe sensory attributes such as ‘hotness’ and ‘port wine’ remained, whereas these attributes were absent in wines produced from earlier-harvested grapes or green harvest wine.

Table 3. Summary of sensory evaluations for different wine blends with reduced alcohol.

Base Wines (Blend Composition)	Number and Type of Blends (Blending Strategies) *	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
Verdelho Petit Verdot	2 binary (co-fermenting monovarietal wines produced from grapes collected during two distinct harvests)	Mudgee region, New South Wales, Australia (2015; two harvests)	Trained panel (n = 12): descriptive analyses	Basic composition Volatiles (GC-MS)	Equal-proportion blends reduced unripe sensory notes while maintaining aroma and flavour profiles comparable to wines produced from fully ripe grapes, supporting blending as an effective strategy for alcohol reduction.	[34]
Cabernet Sauvignon	3 binary (co-fermenting monovarietal wines produced from grapes collected during two distinct harvests)	McLaren Vale, South Australia, (2015; five harvests)	Wine science researchers (n = 10), Expert panellist (n = 1): descriptive analysis	Volatiles (HS-SPME followed by GC-MS)	Substitution with water preserved favourable aroma and flavour attributes but retained some overripe sensory notes, whereas blending with earlier-harvest grape juice reduced “hotness” and “port-like” attributes, highlighting the impact of blending strategy on sensory outcomes	[8]

* The number of blends indicates the total number of experimental blend samples, while the term binary refers to blends composed of two wines; the blending strategy (e.g., co-fermentation) is indicated in parentheses.

Table 3. Summary of sensory evaluations for different wine blends with reduced alcohol.

Base Wines (Blend Composition)	Number and Type of Blends (Blending Strategies) *	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
Verdelho Petit Verdot	2 binary (co-fermenting monovarietal wines produced from grapes collected during two distinct harvests)	Mudgee region, New South Wales, Australia (2015; two harvests)	Trained panel (n = 12): descriptive analyses	Basic composition Volatiles (GC-MS)	Equal-proportion blends reduced unripe sensory notes while maintaining aroma and flavour profiles comparable to wines produced from fully ripe grapes, supporting blending as an effective strategy for alcohol reduction.	[34]
Cabernet Sauvignon	3 binary (co-fermenting monovarietal wines produced from grapes collected during two distinct harvests)	McLaren Vale, South Australia, (2015; five harvests)	Wine science researchers (n = 10), Expert panellist (n = 1): descriptive analysis	Volatiles (HS-SPME followed by GC-MS)	Substitution with water preserved favourable aroma and flavour attributes but retained some overripe sensory notes, whereas blending with earlier-harvest grape juice reduced “hotness” and “port-like” attributes, highlighting the impact of blending strategy on sensory outcomes	[8]

* The number of blends indicates the total number of experimental blend samples, while the term binary refers to blends composed of two wines; the blending strategy (e.g., co-fermentation) is indicated in parentheses.

Understanding consumer preferences is crucial for tailoring wine characteristics to align with market preferences and drive product success. As presented in Table 4, Dooley et al. [13,20], besides evaluating compositional changes during blending, conducted a study to evaluate consumer preferences for Californian base and blended wines, obtained by augmented simplex-centroid mixture design. They developed optimized blends based on consumer feedback and assessed consumer acceptance of the optimized blends in comparison to the original base wines (Cabernet Sauvignon, Merlot, Zinfandel). Blending reduces wine polarization, with segment-specific optimization (68–26–6% and 27–2–71% Cabernet Sauvignon-Merlot-Zinfandel, respectively), offering greater consumer satisfaction than a general blend (44% Cabernet Sauvignon, 21% Merlot, 35% Zinfandel). Another study focused on consumer perceptions of blends, aiming to assess how social drinkers and experts evaluate the complexity of U.S. wines, particularly by identifying the attributes most associated with perceived complexity [54]. Wine complexity was shown to be an intricate concept shaped not only by the wine’s sensory attributes (quality, intensity, and liking) but also by individual factors (age and gender) and contextual influences (such as the order of tasting). Expert tasters perceived the wines as more intense, recognizable, and of superior quality, and demonstrated a greater willingness to pay for them compared to novice consumers. In addition to confirming that complexity perception is influenced by multiple factors, a noteworthy finding was that tasters across all expertise levels were generally unable to reliably distinguish blends from monovarietal wines, while less familiar wines were more frequently misidentified as blends.

Table 4. Summary of sensory evaluations considering consumer perception and preference of blended wines.

Base Wines (Blend Composition)	Number and Type of Blends * (Blending Strategies)	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
Cabernet Sauvignon Merlot Zinfandel	3 binary (co-blending) 4 ternary (co-blending)	California, USA, (2009)	Trained panel (n = 9): structured 10-point scale (appearance, aroma, flavor by mouth) Consumers (n = 108): 9-point hedonic scale (overall liking, appearance, aroma, sweetness, tartness, persistency, mouthfeel and body)	Basic composition and color	Two major consumer segments were identified, primarily based on their responses to Zinfandel, and it was found that segment-specific blends received higher overall liking scores compared to the blend optimized across all consumers.	[13]
Merlot Cabernet Franc Cabernet Sauvignon	3 binary (co-blending)	Finger Lakes, New York, USA (2015)	Trained tasters (n = 6): list of flavour descriptors (WSET SAT) Expertise groups (three expertise levels: novice (n = 41), intermediate (n = 30), expert (n = 16): 9-point scale (liking, familiarity, complexity, and flavour intensity); flavour descriptors; 20-point scale (quality); willingness to pay		Perceived wine complexity depended on both sensory attributes and taster expertise; across expertise levels, blends were not reliably distinguished from monovarietal wines, while experts rated wines as more intense, recognizable, and of higher quality	[54]

* The number of blends indicates the total number of experimental blend samples, while the terms binary or ternary refer to blends composed of two or three base wines, respectively; the blending strategy co-blending) is indicated in parentheses.

Table 4. Summary of sensory evaluations considering consumer perception and preference of blended wines.

Base Wines (Blend Composition)	Number and Type of Blends * (Blending Strategies)	Origin, (Vintage)	Sensory Analyses	Other Analyses	Main Conclusions	References
Cabernet Sauvignon Merlot Zinfandel	3 binary (co-blending) 4 ternary (co-blending)	California, USA, (2009)	Trained panel (n = 9): structured 10-point scale (appearance, aroma, flavor by mouth) Consumers (n = 108): 9-point hedonic scale (overall liking, appearance, aroma, sweetness, tartness, persistency, mouthfeel and body)	Basic composition and color	Two major consumer segments were identified, primarily based on their responses to Zinfandel, and it was found that segment-specific blends received higher overall liking scores compared to the blend optimized across all consumers.	[13]
Merlot Cabernet Franc Cabernet Sauvignon	3 binary (co-blending)	Finger Lakes, New York, USA (2015)	Trained tasters (n = 6): list of flavour descriptors (WSET SAT) Expertise groups (three expertise levels: novice (n = 41), intermediate (n = 30), expert (n = 16): 9-point scale (liking, familiarity, complexity, and flavour intensity); flavour descriptors; 20-point scale (quality); willingness to pay		Perceived wine complexity depended on both sensory attributes and taster expertise; across expertise levels, blends were not reliably distinguished from monovarietal wines, while experts rated wines as more intense, recognizable, and of higher quality	[54]

* The number of blends indicates the total number of experimental blend samples, while the terms binary or ternary refer to blends composed of two or three base wines, respectively; the blending strategy co-blending) is indicated in parentheses.

In general, studies focusing on reduced-alcohol wines indicate that blending can partially mitigate the sensory drawbacks commonly associated with lower ethanol levels, particularly when wines from different harvest maturities are combined rather than diluted with water. However, the effectiveness of such strategies appears highly dependent on the blending approach, grape variety, and proportion used, and is not uniformly successful across sensory attributes. Consumer-oriented studies further demonstrate that perceived quality, complexity, and preference for blended wines are strongly modulated by taster expertise, familiarity, and contextual factors, rather than by blending per se. Some studies report limited ability of consumers to reliably distinguish blended from monovarietal wines, suggesting that blending outcomes may be perceptually subtle. Collectively, these findings highlight that while blending offers valuable tools for sensory optimization and market alignment, its success in reduced-alcohol and consumer-driven contexts is contingent upon carefully targeted formulation and validation.

4. Computational Approaches to Wine Blending

Computer-based analysis of sensory attributes integrates data derived from sensory evaluations, whether obtained through human tasting or instrumental analysis, with mathematical and statistical methods. The primary objective of this approach is to formulate blends that exhibit optimal sensory quality.

In one of the pioneering studies in this field [43], varietal (primary blending wine), white stock, and target blends were analyzed using HS-GC, under the assumption that volatile compounds are the main contributors to aroma. Computer-assisted blending experiments employed simplex optimization and GC-derived data to determine the blending proportions of varietal and stock wines that best replicated the target wine’s aroma profile, as measured by the similarity coefficient. Sensory testing of blends, conducted using triangle tests, confirmed that the computer-optimized blend could not be distinguished from the target wine by the sensory panel.

Subsequent studies focused on neural networks. An optimization method using artificial neural networks was developed to model the nonlinear sensory response of blends based on data from base wines and a limited number of trial blends [18]. This algorithm, verified through a study with 24 wines made from three base wines and sensory data from a trained panel (descriptive analysis), successfully predicted optimal blends that closely matched the target attributes. Ren and Li [45] developed an analytical system involving several steps. Firstly, developing a database management system to store and manage physicochemical data from user-provided liquor samples, ensuring data integrity across approximately 120 floating-point variables. Next, the system integrates the database with MATLAB 7.x, where neural network algorithms are applied to analyze and extract, thereby supporting more accurate and informed blending decisions.

Constraint and mathematical programming techniques were used to optimize the blending process by simultaneously creating multiple target wines with specified aroma concentrations, ensuring that the final volume of each target wine matches the intended production volume [44]. While an oenologist created a target wine by blending several base wines (selected volumes were recorded, and the final blend’s aromatic profile was analyzed), the tool generated a similar blend with the same aroma concentrations but different base wine proportions. A blind test revealed that the oenologist could not distinguish between the human-made and computer-generated wines, suggesting the tool’s potential relevance despite the need for further validation.

A mathematically optimized formulation, based on the mixture designs and triangular surfaces module applied using the STATISTICA v.10 (Tibco, Santa Clara, CA, USA) package, for the Zvezda Kubani blended wine was established based on the results of 31 experiments, effectively capturing the relationship between sensory assessment and possible mixture combinations. The optimized formulation for Zvezda Kubani was established: 48% Merlot, 35% Cabernet Sauvignon, and 17% Pinot Noir, which also achieved the highest sensory ratings in expert evaluations [55].

Generally, computational approaches to wine blending complement traditional oenological expertise, as well as chemical analysis, offering precise, data-driven strategies that enhance blending accuracy, consistency, and sensory quality. Such integrative approaches represent a shift from empirical blending toward rational, data-driven blend design, particularly in complex blending scenarios involving multiple objectives and constraints.

5. Biological Activities of Blended Wines

5.1. In Vitro Studies of Blended Wines’ Biological Activity

Wine polyphenols have been widely investigated in relation to potential biological activities, including antioxidant, anti-inflammatory, and cardioprotective effects, as reported in experimental and epidemiological studies [56,57,58,59]. However, it must be emphasized that any discussion of biological activity of wine components must be interpreted within the broader context of alcohol consumption, for which potential health risks are well documented. The studies summarized herein focus on in vitro and in vivo experimental models and do not imply health benefits of wine consumption per se, i.e., these findings should not be interpreted as a recommendation for wine consumption. Accordingly, this section aims to describe observed bioactivities of wine-derived compounds in blended matrices rather than to advocate alcohol intake.

Wine health benefits are primarily attributed to the bioactivity of polyphenols found in wine [59,60,61]. Given the established role of oxidative stress in the pathogenesis of these conditions, it is not surprising that numerous in vitro assays, such as oxygen radical absorbance capacity (ORAC), ferric reducing antioxidant power (FRAP), the cupric reducing antioxidant capacity (CUPRAC), neutralisation of 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), 1,1-diphenyl-2-picrylhydrazyl; DPPH), peroxynitrite (ONOO−), hydroxyl (OH) and superoxide anion radical (SOA), etc., have been widely employed to assess the antioxidant capacity of wines, particularly red varieties [62,63,64,65,66]. These simple spectrophotometric assays can provide a preliminary estimation of antioxidant potential, but their relevance is limited to in vitro conditions and does not fully reflect the in vivo activity of wine. This is primarily because wine polyphenols undergo absorption and metabolism, with phenolic metabolites as the actual bioactive compounds, exerting systemic effects in vivo [60]. In most reports assessing the antioxidant activity of wine blends, the samples are typically referred to only as ‘blends,’ ‘red blends,’ or ‘white blends,’ without specification of the constituent varietals [67,68,69,70,71]. Nonetheless,

Table 5. Examples of wine blends in vitro antioxidant activity.

Blends	Origin, (Vintage)	Assay/Results	Reference
Cabernet Sauvignon (70%), Syrah (30%)	Argentina, (2007)	DPPH: 46.13% ORAC: 20,124 μmol TE/L *	[67]
Malvazija, Sauvignon Blanc, Pinot Gris	Istria, Croatia, (2008)	DPPH: >90%	[72]
Cabernet Sauvignon, Grenache rouge	Crete, Greece, (2007)	DPPH: 807 ± 22, 913 ± 29 mg GAE/L ** ABTS: 350 ± 13, 444 ± 19 GAE/L CUPRAC: 1252 ± 40, 1451 ± 32 GAE/L FRAP: 546 ± 27, 628 ± 30 GAE/L	[73]
Syrah, Mantilaria	Crete, Greece, (2007)	DPPH: 940 ± 21 GAE/L ABTS: 464 ± 28 GAE/L CUPRAC: 1637 ± 50 GAE/L FRAP: 701 ± 33 GAE/L
Malvasia, Chardonnay	Crete, Greece, (2009)	DPPH: 60.4 ± 3.3 GAE/L ABTS: 46.1 ± 3.4 GAE/L CUPRAC: 105 ± 7 GAE/L FRAP: 51.3 ± 4.3 GAE/L
Sauvignon Blanc, Assyrtico	North Greece, (2009)	DPPH: 106 ± 6 GAE/L ABTS: 118 ± 6 GAE/L CUPRAC: 182 ± 12 GAE/L FRAP: 83.3 ± 7 GAE/L
“Buttafuoco” (Croatina, Barbera, Uva Rara) extract	Oltrepò Pavese, Italy	DPPH: 53.3% OH: 8.2% SOA: 39.6%	[74]

* TE/L: Trolox equivalents/L. ** GAE/L: gallic acid equivalents/L.

presents selected examples of in vitro antioxidant activities of blended wines (with specified base varieties) [72,73,74,75].

Several studies stand out for their interesting approaches to assessing the bioactivity of blends. Wang et al. [48] conducted study in which differences between the antioxidant activity of base wine Cabernet Sauvignon and its blends produced by the addition of 20% Merlot, Marselan, Syrah, or Pinot Noir, using both co-fermentation and post-fermentation blending approaches were observed. Among the monovarietal wines, Cabernet Sauvignon exhibited the highest antioxidant capacity. Among the blends (co-wines), the Cabernet Sauvignon–Merlot combination showed the strongest antioxidant activity, whereas the Cabernet Sauvignon–Pinot Noir blend exhibited the lowest. A similar trend was noted in co-fermentation trials. Overall, co-fermented blends demonstrated greater antioxidant potential compared to post-fermentation blended wines. Comparable conclusions were driven when different ratios of Cabernet Sauvignon and Merlot were used (9:1, 8:2 or 7:3, respectively) [32]. It was confirmed that co-fermented wines had higher antioxidant capacity than monovarietal or co-wines. Although the overall trend in antioxidant capacity for post-fermentation blends indicated a decline, an exception was observed with Merlot as the blending component, where increasing its proportion gradually enhanced antioxidant activity. In addition to the antioxidant activity of the Gaglioppo–Magliocco blend (1:1), D’Onofrio et al. [75] investigated its cytoprotective effects on endothelial cells. The blend did not compromise cell viability and demonstrated potential in protecting endothelial cells from high-glucose and palmitic acid-induced cytotoxicity. However, its efficacy was lower compared to the antioxidant and cytoprotective effects observed for Magliocco alone. Majkić et al. [25] investigated the bioactivities of Cabernet Sauvignon and Merlot monovarietal wines, their blends in ratios of 1:1, 3:1, and 1:3, as well as a commercial blend. Evaluated parameters included inhibition of digestive enzymes (α-amylase, α-glucosidase, lipase), tyrosinase inhibition, and antioxidant and anti-inflammatory effects in a cell-based model. All samples demonstrated significant bioactivity with varying potencies. However, the incorporation of Merlot exerted a more pronounced effect on modulating the bioactivity of Cabernet Sauvignon than the reverse, highlighting the complexity and distinctive nature of each blend.

Generally, in vitro bioassays represent a fundamental initial step in evaluating varietal wines, blends or any other foods, but they primarily serve as a preliminary screening tool, and it is essential to remain mindful of their inherent limitations when interpreting the results.

5.2. In Vivo Studies of Blended Wines Biological Activity

To indicate the concept of a link between moderate consumption of blends and health benefits, in vivo studies represent the most relevant approach. To the best of our knowledge, only a few studies have explored this direction. Demrow et al. [76] evaluated the effects of red wine (1987 Châteauneuf-du-Pape, a Grenache-based red blend), white wine (1990 Château Villotte Bordeaux, a blend of Sémillon, Sauvignon Blanc, and Muscadelle), and natural, purple grape juice in a canine model of coronary artery stenosis with thrombus-induced cyclic flow reductions (CFR). Intravenous (IV) and intragastric (IG) administration of red wine (1.62 ± 1.12 mL/kg IV; 4.0 mL/kg IG) and grape juice (2.04 ± 1.42 mL/kg IV; 10 mL/kg IG) effectively abolished CFRs, whereas white wine (2.0 mL/kg IV; 4.0 mL/kg IG) exhibited only minimal efficacy at the tested doses.

Macedo et al. [77] investigated the correlation between in vitro and in vivo antioxidant activity of red wines (Syrah, Cabernet Sauvignon, and a blend), using rats on a high-fat diet supplemented with wines classified by their antioxidant capacity (3 levels). The study revealed a correlation between the in vitro and in vivo antioxidant activity in plasma. In the liver, only the wine with the highest in vitro antioxidant activity reduced malondialdehyde concentrations, while wines with lower antioxidant activity increased antioxidant enzyme activity. This suggests that the relationship between in vitro and in vivo antioxidant effects varies depending on the specific biomarker used to assess oxidative stress. In a study by Qian et al. [78], hamsters fed an atherogenic diet were administered different red wines: Grenache, Syrah, or a blend of Grenache (70%), Syrah (20%), and Carignan (10%), produced using flash release, tannin-enriched, or traditional winemaking techniques. In general, all wines demonstrated protective effects against diet-induced atherosclerosis, with varying efficacy in reducing total, HDL-, and LDL-cholesterol; plasma uric acid levels, aortic fatty streak lesion area, superoxide anion production, and NAD(P)H oxidase expression. The study by Blackhurst and Marais [79] aimed to compare the lipid peroxidation status of oil rich in polyunsaturated fatty acid, consumed by volunteers, with that of the chylomicrons formed, both with and without the consumption of a red wine blend. The findings showed minimal impact of red wine on lipid peroxidation in plasma, despite elevated catechin levels, and highlighted limitations due to the small sample size and use of a single wine and oil. The authors stress the need for further research on polyunsaturated fatty acid metabolism and individual differences in oxidative stress susceptibility to inform more targeted future studies [79].

Due to the limited data on the in vivo bioactivity of wine blends, studies involving human volunteers are particularly valuable, as they offer critical insights into the potential health impacts of wine consumption under real-life conditions, while also enabling exploration of consumer preferences in relation to lifestyle factors and health status. However, such findings must be interpreted with caution, given the well-established risks associated with alcohol consumption.

6. Conclusions

In an increasingly competitive wine market, producers are under pressure to develop distinctive products with recognizable flavour profiles, while maintaining consistent quality, reasonable costs, and economic sustainability. Within this context, blending has emerged as a practical tool to refine sensory attributes while also supporting production efficiency. Despite its practical importance, blending remains relatively underrepresented in the scientific literature compared to other areas of oenological research. Most existing studies place strong emphasis on sensory perception, reflecting both consumer interest and the central role of sensory quality in blending decisions. Furthermore, there is considerable variation in the analytical and oenological methodologies employed, as well as in the types of blends investigated. This heterogeneity extends to the grape varieties considered, the blending techniques applied (e.g., co-maceration, co-fermentation, or post-fermentation blending), and the experimental conditions under which blends are evaluated. A common feature across most studies is the use of grapes or wines originating from the same geographic region, although harvest times and vintages may vary.

Among the topics addressed in this review, computational and data-driven approaches to wine blending appear particularly promising for future research and practical application. These methods have the potential to support more rational blend design while reducing reliance on extensive empirical trial-and-error experimentation. Future research is thus expected to increasingly rely on computational assistance to complement traditional oenological expertise.

However, with increasing attention to the potential biological effects of wine constituents, forthcoming research may shift toward investigating the bioactivity of blends. In this context, bio-guided design, where specific biological responses guide ingredient selection and formulation, may offer a useful complementary perspective for future blending studies. Nevertheless, such approaches should be interpreted with caution and within the broader context of the well-established health risks associated with alcohol consumption.