Spatial variation in milk microbiota and nutritional health quality across urbanizing environments: age-specific vulnerability in pediatric and geriatric populations

Variação espacial na microbiota do leite e qualidade nutricional em ambientes urbanizados: vulnerabilidade específica por idade em populações pediátricas e geriátricas

Abstract

Milk is a nutrient-dense matrix that can function as a micro-ecosystem in which environmental exposure, handling hygiene, and heat treatment jointly shape microbial load and nutritional health quality. In this study, we analyzed >100 milk samples (raw, boiled, and commercially pasteurized) collected from urban and peri-urban settings to characterize bacterial contamination alongside key physicochemical attributes. Raw milk showed compromised compositional integrity (fat 0.99%, solids-not-fat 5.76%) and substantial bacterial contamination, with total viable counts 5.5 × 10⁵ CFU mL⁻¹, coliforms 7.8 × 10⁴ CFU mL⁻¹, and fecal-origin coliforms 1.3 × 10⁵ CFU mL⁻¹. Across samples, microbial burden was strongly and inversely associated with solids-not-fat (r = −0.85, p < 0.001), indicating a quantifiable linkage between contamination pressure and deterioration in nutritional health quality. Heat treatment produced distinct ecological outcomes: boiling reduced overall counts yet permitted persistence of specific taxa, with Staphylococcus aureus remaining detectable (3.7 × 10⁵ CFU mL⁻¹), whereas pasteurization largely eliminated culturable bacteria but was accompanied by shifts in physicochemical attributes. Collectively, these findings show that urban–peri-urban supply conditions and processing intensity drive predictable changes in milk microbiota and nutritional health quality. From a public-health perspective, these patterns imply age-specific vulnerability, where microbial hazards and quality loss may disproportionately affect pediatric consumers and the geriatric population. Strengthening hygienic handling, cold-chain temperature control, and standardized processing is therefore essential to protect nutritional health quality and reduce milk-borne risk in urbanizing dairy chains, particularly for these high-susceptibility age groups.

Keywords:
microbial ecology; milk microbiota; physicochemical composition; environmental contamination; thermal processing

Resumo

O leite é uma matriz rica em nutrientes que pode funcionar como um microecossistema, no qual a exposição ambiental, as condições de higiene durante o manuseio e o tratamento térmico influenciam conjuntamente a carga microbiana e a qualidade nutricional. Neste estudo, foram analisadas mais de 100 amostras de leite (cru, fervido e pasteurizado comercialmente), coletadas em ambientes urbanos e periurbanos com o objetivo de caracterizar a contaminação bacteriana e avaliar importantes atributos físico-químicos. O leite cru apresentou integridade composicional comprometida (gordura 0,99%, sólidos não gordurosos 5,76%) e contaminação bacteriana substancial, com contagem total de bactérias viáveis ​​de 5,5 × 10⁵ UFC mL⁻¹, coliformes de 7,8 × 10⁴ UFC mL⁻¹ e coliformes de origem fecal de 1,3 × 10⁵ UFC mL⁻¹. Em todas as amostras, a carga microbiana apresentou forte correlação inversa com o teor de sólidos não gordurosos (r = −0,85, p < 0,001), indicando uma associação quantificável entre a pressão de contaminação e a deterioração da qualidade nutricional. O tratamento térmico produziu resultados ecológicos distintos: a fervura reduziu a contagem total, mas permitiu a persistência de táxons específicos, com Staphylococcus aureus permanecendo detectável (3,7 × 10⁵ UFC mL⁻¹), enquanto a pasteurização eliminou em grande parte as bactérias cultiváveis, mas foi acompanhada por alterações nos atributos físico-químicos. Coletivamente, esses resultados indicam que as condições de fornecimento urbano-periurbano e a intensidade do processamento impulsionam mudanças previsíveis na microbiota do leite e na qualidade nutricional. De uma perspectiva de saúde pública, esses padrões sugerem vulnerabilidade específica por idade, uma vez que os riscos microbiológicos e a perda de qualidade podem afetar desproporcionalmente os consumidores pediátricos e a população geriátrica. O fortalecimento das práticas de higiene no manuseio, do controle de temperatura na cadeia de frio e da padronização do processamento é, portanto, essencial para proteger a qualidade nutricional e reduzir o risco de doenças transmitidas pelo leite nas cadeias de produção e distribuição de laticínios em processo de urbanização, particularmente para esses grupos etários de maior suscetibilidade.

Palavras-chave:
ecologia microbiana; microbiota do leite; composição físico-química; contaminação ambiental; processamento térmico

1. Introduction

Milk is a nutrient-rich biological substrate whose composition and microbial ecology are shaped by environmental conditions, handling practices, and local distribution systems. In many regions, especially where informal dairy networks dominate, milk is exposed to a wide range of ecological pressures that influence its physicochemical traits and microbial assemblages. Reports from across Pakistan illustrate how milk quality varies widely due to environmental contamination, inadequate storage conditions, and unregulated distribution, resulting in substantial alterations to its microbial and chemical profiles (Independent News Pakistan, 2024; Asghar, 2025; Khan, 2024; News Desk, 2024; Ahmad, 2022). These observations reflect a broader national pattern in which environmental factors, handling routes, and on-farm management shape the biological state of milk as it moves through diverse ecological settings.

In low- and middle-income countries, the dynamics of food substrates, such as milk, are strongly influenced by gaps in environmental monitoring and the absence of comprehensive quality-control systems. Reports by international bodies emphasize that the biological condition of milk is tied to environmental exposures, source hygiene, and the broader ecology of livestock production systems (WHO, 2015; FAO, 2013). Fragmented monitoring, limited environmental surveillance, and variable distribution practices contribute to inconsistent quality profiles in dairy substrates across local ecosystems.

Milk, frequently regarded as a complete biological resource, contains proteins, lipids, carbohydrates, minerals, and vitamins that serve as growth media for diverse microbial taxa (WHO, 2015; FAO, 2013). Despite Pakistan being a major global milk producer with over 60 billion liters of annual output, the biological attributes of milk are strongly influenced by its predominantly informal distribution chain (Codex Alimentarius Commission, 2004). This system is characterized by minimal environmental regulation, heterogeneous handling practices, and susceptibility to microbial infiltration and chemical alteration (FAO, 2011).

Milk collected under hygienic conditions can still acquire microbial communities rapidly through environmental exposure during milking, storage, and transport (Trade Development Authority of Pakistan, 2022; Gallup Pakistan, 2024; Punjab Food Authority, 2018). The absence of effective cold-chain infrastructure allows rapid proliferation of organisms such as Escherichia coli, Listeria monocytogenes, Salmonella spp., Campylobacter jejuni, and Staphylococcus aureus, which are commonly associated with environmental contamination and livestock-associated ecological niches (Pakistan Today, 2022; Yunus et al., 2019; Chandio et al., 2007; Hussain et al., 2007; Lee et al., 2019; Haq et al., 2024). These taxa frequently expand under favorable temperature and nutrient conditions, shaping the overall microbial ecology of milk.

Thermal processing practices also influence milk’s microbial and physicochemical properties. Although pasteurization is globally recognized for its ability to reduce microbial communities, it is inconsistently applied in regions where traditional consumption patterns dominate (Ibrahim et al., 2023; Abbas et al., 2022; Hussain et al., 2007). Boiling—widely practiced in domestic settings—may reduce but does not eliminate heat-resilient microorganisms, and cannot reverse alterations resulting from prior contamination or adulteration (Kmiha et al., 2017). Such practices highlight how environmental factors and local processing behaviors jointly influence the biological condition of milk.

Environmental studies have documented widespread adulteration of milk with additives such as water, starch, detergents, formalin, urea, and synthetic preservatives (Nawaz et al., 2022; Ibrahim et al., 2023). These substances modify the physicochemical composition of milk, influence its biological stability, and may alter the nutritional quality of the product. Separately, multidrug-resistant Staphylococcus aureus has been reported in raw milk from cows with subclinical mastitis in Pakistan, reinforcing the food-safety importance of microbiological surveillance (Haq et al., 2024).

The informal distribution network, which supplies the majority of milk in both rural and urban settings, operates largely outside standardized environmental protocols. In major urban centers such as Islamabad and Rawalpindi, milk commonly originates from small-scale vendors and decentralized production systems without uniform hygienic practices, routine quality monitoring, or traceability mechanisms (Tirloni et al., 2022). These conditions shape the ecological landscape in which milk exists, affecting microbial colonization patterns and compositional variability as it moves through the supply chain.

This study provides an ecological and environmental assessment of raw, boiled, and pasteurized milk obtained from diverse locations across the twin urban–peri-urban ecosystems of Islamabad and Rawalpindi. The research quantifies microbial communities, physicochemical traits, and chemical alterations while also documenting environmental handling practices that influence milk quality. The findings contribute to a broader understanding of milk as a dynamic biological matrix affected by environmental contamination pathways, substrate–microbe interactions, and variable processing conditions.

2. Material and Methods

2.1. Sample collection and preparation

A total of 145 milk samples were collected from over 100 distinct regions within the twin cities of Islamabad and Rawalpindi, Pakistan, representing a broad socioeconomic spectrum. These included 71 raw, 37 pasteurized, and 37 boiled milk samples. The regions ranged from affluent urban areas to underprivileged neighborhoods, ensuring a diverse demographic representation of milk quality. Samples were collected in sterile glass containers, labeled appropriately, and transported to the laboratory at 4°C in insulated coolers to maintain sample integrity. Upon arrival, the samples were equilibrated to ambient temperature before analysis.

2.2. Organoleptic and physical analysis

2.2.1. Color and appearance

Each milk sample (10 mL) was transferred into a transparent glass beaker and visually assessed under standardized lighting against a white background. Discoloration, turbidity, and sedimentation were noted. Raw milk exhibited significant variability, with some samples displaying unusual coloration indicative of bacterial contamination or possible adulteration. Pasteurized milk was generally consistent, while boiled milk exhibited minor heat-induced changes, retaining acceptable organoleptic characteristics.

2.2.2. Odor analysis

Odor assessment was performed immediately upon opening the milk sample in a controlled, odor-free environment. Deviations from the fresh milk aroma, such as rancid, sour, or chemical-like odors, were recorded. Raw milk samples occasionally exhibited off-putting odors, while pasteurized and boiled milk samples showed stable, typical olfactory properties.

2.2.3. Organoleptic assessment

Sensory evaluation was performed by a trained panel, assessing the taste, texture, and overall acceptability of each sample following standardized procedures. Pasteurized milk aligned with quality benchmarks, while raw milk demonstrated notable inconsistencies.

2.3. Chemical analysis

2.3.1. Determination of milk fat (Gerber method)

Milk fat content was determined using the Gerber method. A 10 mL aliquot of sulfuric acid (specific gravity 1.820) was added to a Gerber butyrometer, followed by 11 mL of milk. After adding 1 mL of amyl alcohol, the butyrometer was sealed, mixed gently, and centrifuged at 1100 rpm for 5 minutes. The fat content was read directly from the graduated scale.

2.3.2. Total solids and Solids-Not-Fat (SNF) determination

Total solids content was determined by weighing an empty, clean china dish (W1). A 5 g aliquot of well-mixed milk was added (W2), and the dish was dried at 102°C until a constant weight (W3) was obtained. Total solids were calculated as Equation 1:

$$TotalSolids\left(\%\right)=\left(\left(W3-W1\right)/\left(W2-W1\right)\right)\times 100$$ (1)

SNF content was derived by subtracting the fat content from the total solids.

2.3.3. Adulterant detection

The presence of common milk adulterants was assessed using the following tests:

• Urea Detection (p-Dimethylaminobenzaldehyde Test): A 5 mL milk sample was mixed with 5 mL reagent. A yellow-to-brown color change indicated urea presence.

• Starch Detection (Iodine Test): A 1 mL aliquot of 1% iodine solution was added to 5 mL milk. A blue-black color confirmed starch presence.

• Sodium Carbonate (Soda) Detection: A 5 mL milk sample mixed with 2 mL dilute hydrochloric acid (HCl) produced effervescence in the presence of soda.

• Formalin Detection (Chromotropic Acid Test): A 2 mL milk sample was mixed with 2 mL chromotropic acid reagent and heated at 60°C for 10 minutes. The development of a violet color confirmed formalin presence.

• Detergent Detection (Bromocresol Purple Test): A 5 mL milk sample was mixed with 2 mL bromocresol purple dye. A violet or blue color indicated detergent contamination.

The qualitative adulterant assays were interpreted according to their specified color-change or effervescence endpoints. Category-specific findings for urea, starch, sodium carbonate (soda), formalin, and detergent are reported in the Results section.

2.4. Microbiological analysis

2.4.1. Total Viable Count (TVC)

Serial dilutions of the milk samples were plated on plate count agar and incubated at 37°C for 24-48 hours. Colony-forming units (CFUs) were counted and expressed as CFU/mL.

2.4.2. Coliform count and fecal contamination

1 mL of milk was inoculated onto MacConkey agar and incubated at 37°C for 24 hours. Pink lactose-fermenting colonies were counted. Fecal coliforms were assessed by streaking samples onto Eosin Methylene Blue (EMB) agar and incubating at 44°C for 24 hours, with dark-centered colonies indicating fecal contamination.

2.4.3. Pathogenic and spoilage microorganisms

• Staphylococcus aureus: A loopful of milk was streaked onto Mannitol Salt Agar (MSA) and incubated at 37°C for 24 hours. Yellow colonies confirmed the presence of Staphylococcus aureus.

• Salmonella Detection: Samples were streaked onto Xylose Lysine Deoxycholate (XLD) agar and incubated at 37°C for 24 hours. Black-centered colonies confirmed Salmonella spp. presence.

• Fungal Contamination: A 1 mL aliquot of milk was spread onto Sabouraud Dextrose Agar (SDA) and incubated at 25°C for 5-7 days. Fungal growth was observed and confirmed microscopically.

Microbiological findings were recorded separately by milk category and are reported in the Results section. Culture-based detection was interpreted under the incubation conditions specified above, and colony counts were expressed as CFU/mL where quantitative results were available.

2.5. Statistical analysis

Statistical analyses were performed using Python (version 3.9), utilizing libraries such as pandas for data cleaning, numpy for numerical computations, and statsmodels for regression analysis and predictive modeling. Seaborn and matplotlib were employed for data visualization, including correlation matrices and microbial risk assessments. Descriptive statistics, including mean, standard deviation, minimum, and maximum values, were computed for all physicochemical and microbiological parameters. Pearson correlation coefficients were used to assess relationships between microbial loads and chemical parameters, visualized through heatmaps.

Ordinary Least Squares (OLS) regression was used to explore the relationship of milk fat, SNF, and total solids with Total Viable Count (TVC). Model fit was described using R-squared, adjusted R-squared, the F statistic, corresponding p-values, and coefficient estimates. Because total solids were derived from fat and SNF in the analytical workflow, the multivariable model was treated as exploratory and interpreted with caution for collinearity. The 10-point trend graphic was used only as an illustrative scenario visualization with arbitrary time units; it was not treated as an observed longitudinal forecast.

2.6. Ethical considerations

All procedures were conducted in compliance with the Pakistan Pure Food Rules (PPFR), Pakistan Standards and Quality Control Authority (PSQCA), World Health Organization (WHO), and Food and Agriculture Organization (FAO) safety guidelines. Laboratory personnel followed stringent biosafety protocols to ensure sample integrity and minimize contamination risks.

2.7. Software and statistical validity

All statistical analyses were conducted in Python. Regression diagnostics, including the Durbin-Watson statistic and inspection of model conditioning, were used as supporting checks rather than as proof of model validity. Statistical significance was interpreted at p < 0.05, while non-significant model terms were reported without causal or predictive overinterpretation.

3. Results

Milk is an essential component of human nutrition, and its safety and quality are of paramount importance for public health. This study provides a comprehensive evaluation of raw, pasteurized, and boiled milk samples by analyzing their physical, chemical, and microbiological characteristics. The primary objective is to compare the quality of these milk types, assess their compliance with food safety standards, and determine potential health risks associated with their consumption.

4. Quality Analysis of Raw Milk Samples

4.1. Overview of raw milk sample quality

The raw milk samples were sourced from buffalo and cow. Physical, chemical, and microbiological parameters were analyzed to assess overall quality. Although the samples generally showed acceptable sensory characteristics, most raw-milk samples were classified as unsatisfactory when the combined microbiological and chemical criteria were applied.

4.2. Physical and chemical quality assessment

4.2.1. Physical and organoleptic parameters

Each analyzed milk sample demonstrated visually appealing characteristics, with an acceptable color, odor, and overall presentation that aligned with consumer expectations. Furthermore, the organoleptic evaluation verified that the taste and sensory attributes conformed to established dairy quality benchmarks, reinforcing the milk's suitability from a sensory perspective.

4.2.2. Chemical composition

Milk Fat ranged from 0.82% to 1.11%, slightly lower than standard requirements for buffalo milk (usually expected to be above 6%). Solids-Not-Fat varied from 5.13% to 6.24%, which is within standard ranges. Total Solids were found between 6.06% and 7.28%, which is slightly below expected values for buffalo milk. For adulterants, there was no detection of urea, starch, soda, formalin, or detergent, suggesting no intentional chemical adulteration.

4.3. Microbiological contamination and health risks

Microbiological contamination was widespread among the raw-milk samples, and most failed the applied safety classification. Elevated viable counts, coliform indicators, Staphylococcus aureus, and detection of Salmonella in multiple samples supported concern regarding hygiene during milking, storage, transport, and retail handling.

4.3.1. Microbial load analysis

Total Viable Count (TVC) ranged from 1.2 × 10⁵ to 7 × 10⁶ CFU/mL, exceeding permissible limits (10⁵ CFU/mL). High TVC suggests extensive bacterial contamination, indicating inadequate hygiene and potential post-milking contamination. Coliform and Fecal Coliform Count indicated that coliform bacteria were present in all samples (ranging from 2 × 10³ to 6 × 10⁵ CFU/mL), indicating possible fecal contamination and unsanitary handling. Fecal coliforms were found in numerous samples, signifying direct contamination from fecal matter, posing serious risks of gastrointestinal infections. Staphylococcus aureus was detected in all samples (1 × 10² to 5 × 10⁵ CFU/mL), representing a significant threat as it is associated with foodborne illnesses causing vomiting, diarrhea, and severe infections. Salmonella spp were present in multiple samples, indicating a high probability of foodborne disease outbreaks. Salmonella infections are linked to severe gastroenteritis, fever, and life-threatening complications. Fungal Contamination indicated that fungal spores were found in several samples, indicating mold growth, which can lead to spoilage and production of harmful mycotoxins.

4.3.2. Health hazards associated with contaminated raw milk

The presence of microbial contaminants in raw milk can lead to the following health concerns; such as gastrointestinal infections caused by coliform bacteria, Salmonella, and fecal contamination, leading to severe diarrhea, vomiting, dehydration, and stomach cramps. High counts of Staphylococcus aureus increase the risk of toxin-mediated illnesses indicative of food poisoning. Typhoid and Paratyphoid Fever due to potential transmission of Salmonella spp. Some fungi can produce harmful mycotoxins that may lead to long-term health effects, including liver damage and immunosuppression. Young children, elderly individuals, and immunocompromised patients are at heightened risk of severe illnesses due to contaminated milk consumption.

5. Quality Analysis of Boiled Milk Samples

Boiling is a widely practiced method of milk processing, aimed at eliminating microbial contaminants while preserving its nutritional integrity. Herein we examined the physicochemical, microbiological, and adulteration-related parameters of boiled milk samples, offering a comprehensive insight into its safety and quality.

5.1. Physical and organoleptic properties

The boiled milk samples generally retained acceptable color, odor, appearance, and organoleptic characteristics. A creamy-white appearance and smooth texture were common, with no consistent evidence of gross coagulation or phase separation. A mild caramelized aroma was noted in some samples after heating, and slight creaming could occur on cooling. These observations describe sensory effects of heating and do not indicate microbiological safety.

5.2. Chemical parameters

The milk fat content ranged significantly from as low as 0.98% to as high as 6.90%. The lowest fat levels were observed in samples FAL-504/507 and FAL-502/505, while the highest fat levels were detected in sample FAL-398/412. Solid-Not-Fat (SNF) values varied from 4.12% to 9.32%, with lower SNF levels detected in cow milk (e.g., FAL-320 with 4.12% SNF) compared to buffalo milk. Total Solids content showed variability, with some samples containing as low as 5.49% (FAL-320) and as high as 16.22% (FAL-398/412). Several samples were reported as spoiled (e.g., FAL-392/406, FAL-395/409, FAL-393/407, and FAL-399/413), indicating poor milk quality likely due to microbial contamination or inadequate handling and storage conditions.

5.3. Adulteration analysis

The tested adulterants—urea, starch, sodium carbonate (soda), formalin, and detergent—were not detected in the boiled-milk samples by the qualitative assays used. This finding addresses the tested chemical adulterants only; it does not establish microbiological acceptability. Boiling can reduce viable microorganisms, but residual contamination or post-boiling recontamination remains possible when handling and storage are inadequate.

5.4. Microbiological parameters

Boiled milk showed persistent culturable contamination rather than uniform sterility, indicating that heat treatment did not prevent residual or post-processing microbial burden in all samples. The Total Viable Count (TVC) ranged from 6x10⁴ to 1x10⁶, with sample FAL-463/489 presenting the highest microbial load (1x10⁶). The coliform count indicated that most samples showed substantial coliform contamination, with values reaching 8x10⁵ (FAL-425/445). High coliform counts indicate fecal contamination, improper handling, or unclean storage conditions. Fecal Coliform Count revealed that several samples (e.g., FAL-394/408, FAL-398/412, FAL-396/410) exhibited fecal coliform presence, signifying contamination linked to unhygienic conditions. Staphylococcus aureus was detected in nearly all samples, with counts ranging from 2x10⁴ to 9x10⁴, indicating potential risks of foodborne illness. Salmonella testing revealed that some samples (FAL-394/408, FAL-398/412, FAL-392/406, FAL-395/409, FAL-393/407, FAL-396/410, FAL-391/405) were tested positive for Salmonella, raising serious health concerns. Fungal contamination revealed that a few samples (FAL-394/408, FAL-398/412, FAL-392/406, FAL-395/409, FAL-393/407, FAL-396/410, FAL-391/405) exhibited fungal growth, further degrading the milk’s safety.

5.5. Overall quality assessment

Among the 36 boiled-milk samples assigned a binary safety classification in the source summary, 34 (94.4%) were unsatisfactory and 2 were satisfactory. Because 37 boiled samples were collected overall, one boiled sample was not represented in the binary satisfactory/unsatisfactory count. This discrepancy is retained transparently without altering the journal-proof table values.

6. Quality Analysis of Pasteurized and UHT-Treated Milk Samples

6.1. Physical and organoleptic parameters

All tested milk samples exhibited satisfactory results for color, odor, and appearance, ensuring acceptable consumer appeal. The organoleptic assessment confirmed that the taste and sensory characteristics were within acceptable standards, supporting the integrity of these dairy products.

6.2. Chemical parameters

The milk fat content varied across samples, with some brands maintaining compliance with regulatory standards including Nurpur 3.52%, Olpers 3.60%, Haleeb 5.51% and Tarang 6.51%. However, some brands, such as Dairy Omung (1.69%), exhibited low-fat content, which may indicate partial skimming. Solids-Not-Fat (SNF) levels ranged between 4.29% (Cup Shup) and 26.20% (Dairy Omung), suggesting variability in milk composition. Dairy Omung’s exceptionally high SNF content raises concerns about potential fortification or differences in milk processing. Total Solids levels were largely consistent, with values typically falling within regulatory thresholds. No samples tested positive for adulterants such as urea, starch, soda, formalin, or detergent, indicating adherence to safety standards.

6.3. Microbiological parameters

Most pasteurized milk samples showed no microbial contamination, with nil counts for Total Viable Count, Coliforms, Fecal Coliforms, Staphylococcus aureus, Salmonella, and Fungi. However, one sample (Cup Shup, FAL-366/374) reported a Total Viable Count of <1x10³, suggesting potential exposure to microbial contamination, though still within acceptable limits.

6.4. Compliance with regulatory standards

The majority of samples complied with Pakistan Pure Food Rules (PPFR), PSQCA, WHO, and FAO guidelines. Some samples were deemed “Unsatisfactory” based on chemical composition, particularly in cases where SNF or fat content did not meet expected standards. For example, Milk Fields Full Cream Milk and Asli Milk were flagged for chemical inconsistencies despite being microbiologically safe.

6.5. Overall quality assessment

The overall quality assessment categorized most samples as “Satisfactory,” ensuring their safety for consumption. The exceptions, such as Cup Shup and Tarang, which exhibited inconsistencies in fat/SNF content or microbial findings, necessitate further monitoring. The findings highlight the importance of stringent quality control measures in pasteurized milk production to maintain consumer confidence and regulatory compliance.

7. Comparative Nutritional Quality Assessment of Milk Categories and Health Compliances

The physical properties of all milk samples, including color, odor, and appearance, were found to be satisfactory, with no signs of visual deterioration. However, subtle variations were noted in the sensory characteristics, particularly in boiled milk, where a mild caramelized aroma was observed due to maillard reactions during heating. No phase separation or visible coagulation was detected in any sample category, ensuring their visual integrity. Raw milk, being unprocessed, posed the highest microbial risks, whereas pasteurized milk retained its original sensory attributes, maintaining consumer acceptability. Boiled milk exhibited minimal textural changes but showed slight creaming upon cooling, indicating some degree of alteration due to heat treatment.

The chemical analysis revealed notable variations in milk fat, solids-not-fat (SNF), and total solids content across the three milk types. Raw milk exhibited substantial variability, with some samples showing low fat and SNF levels, potentially indicating dilution or inadequate livestock nutrition. Pasteurized milk generally adhered to standard values, although a few brands displayed reduced fat content, likely due to partial skimming. Boiled milk preserved most of its nutritional properties, with only minimal reductions in solids and fat content due to heating. Notably, none of the analyzed samples contained adulterants such as urea, starch, soda, formalin, or detergent, confirming compliance with Pakistan Pure Food Rules (PPFR), PSQCA, WHO, and FAO safety regulations.

The nutritional profiles of the three milk types were compared based on the key chemical parameters as shown in Table 1.

Table 1
Chemical Composition Analysis of Selected Milk Samples (Mean ± Standard Deviation).

Parameter	Boiled Milk	Pasteurized Milk	Raw Milk
Milk Fat (%)	3.06 ± 1.25	3.83 ± 1.02	2.62 ± 1.53
Solid-Not-Fat (SNF) (%)	7.13 ± 1.10	9.12 ± 0.97	6.86 ± 1.22
Total Solids (%)	10.18 ± 2.41	12.94 ± 1.34	9.47 ± 2.02

Pasteurized Milk exhibited superior nutritional content, with the highest Milk Fat, SNF, and Total Solids. Boiled Milk showed moderate improvement over Raw Milk, but it did not reach the nutritional benchmarks of Pasteurized Milk. Raw Milk had lower Milk Fat and SNF values, indicating possible dilution or poor quality.

The microbiological assessment underscored the significant contamination risks associated with raw milk. Raw milk showed high viable and coliform burdens, whereas boiled milk retained substantial residual contamination in many samples. Nearly all pasteurized samples showed no detectable culturable bacteria; one Cup Shup sample had a low TVC within the stated acceptable threshold. The comparative microbiological results are summarized in Table 2.

Table 2
Microbiological Contamination and Selected Milk Samples (Mean ± Standard Deviation).

Parameter	Boiled Milk	Pasteurized Milk	Raw Milk
Total Viable Count (CFU/mL)	4.92 × 105 ± 3.14 × 105	0 (Safe)	5.75 × 105 ± 3.10 × 105
Coliform Count (CFU/mL)	3.20 × 105 ± 2.10 × 105	0 (Safe)	3.78 × 105 ± 4.10 × 105
Staphylococcus aureus (CFU/mL)	1.70 × 105 ± 1.64 × 105	0 (Safe)	1.78 × 105 ± 2.16 × 105

• Note: Values are mean ± standard deviation. CFU/mL, colony-forming units per milliliter. “0 (Safe)” denotes no detectable culturable growth under the stated assay conditions and should not be interpreted as proof of absolute sterility.

Raw milk demonstrated substantial microbial contamination, with elevated Total Viable Count (TVC), coliforms, and Staphylococcus aureus. Boiling reduced some microbial measures but did not eliminate residual contamination. Pasteurized milk had the most favorable microbiological profile: nearly all samples showed no detectable culturable contamination, with one low-TVC sample reported within the stated acceptable threshold.

The chemical analysis of raw milk indicated a fat content ranging between 0.82% and 1.11%, which is significantly lower than the expected standard for buffalo milk. The SNF content ranged from 5.13% to 6.24%, falling within standard limits, while total solids varied between 6.06% and 7.28%, slightly below expected values. The absence of chemical adulterants suggests no deliberate tampering; however, the compromised microbiological quality renders raw milk unsafe for human consumption. The high microbial load in raw milk poses severe health hazards. Consumption of contaminated raw milk can lead to gastrointestinal infections due to coliform and Salmonella contamination, resulting in symptoms such as diarrhea, vomiting, dehydration, and stomach cramps. The presence of Staphylococcus aureus significantly increases the risk of toxin-mediated food poisoning. Additionally, the detection of Salmonella spp. in multiple samples raises concerns regarding the potential transmission of typhoid and paratyphoid fever. Fungal contamination in raw milk may lead to mycotoxin production, which is associated with long-term health risks, including liver damage and immunosuppression. Infants, elderly individuals, and immunocompromised patients are particularly vulnerable to severe health consequences due to raw milk consumption. Pasteurized and UHT-treated milk samples demonstrated satisfactory physical and sensory attributes, maintaining acceptable color, odor, and appearance. The chemical composition varied among different brands, with milk fat content ranging from 1.69% (Dairy Omung) to 6.51% (Tarang), while solids-not-fat (SNF) values fluctuated between 4.29% (Cup Shup) and 26.20% (Dairy Omung), suggesting variability in milk formulation and potential fortification. No adulterants were detected in any sample, reaffirming compliance with safety regulations. Microbiologically, the majority of pasteurized milk samples were free from bacterial contamination, ensuring their safety for consumption. However, one sample (Cup Shup) exhibited a low TVC of <1×10³ CFU/mL, indicating potential microbial exposure, though still within permissible limits. Regulatory compliance assessments revealed that while most pasteurized milk brands met Pakistan Pure Food Rules (PPFR), PSQCA, WHO, and FAO standards, some samples displayed inconsistencies in chemical composition. Specifically, Milk Fields Full Cream Milk and Asli Milk showed deviations in fat and SNF content, affecting their classification as fully compliant despite being microbiologically safe.

7.1. Safety assessment of selected milk samples: satisfactory vs. unsatisfactory samples

The safety of the milk samples was evaluated using the combined microbial and chemical criteria applied in this study. The corresponding satisfactory and unsatisfactory counts are summarized in Table 3.

Table 3
Assessment of Safety During Consumption of Selected Milk Samples.

Milk Type	Satisfactory Samples	Unsatisfactory Samples
Boiled Milk	2	34
Pasteurized Milk	33	4
Raw Milk	8	63

• Note: Values are numbers of samples. Satisfactory/unsatisfactory classification was based on the combined chemical and microbiological criteria applied in this study. For boiled milk, the displayed counts total 36 (2 satisfactory + 34 unsatisfactory) although 37 boiled samples were collected; one boiled sample was not assigned a binary category in the source summary and is therefore not included in these counts.

Pasteurized Milk had the highest safety rating, with 89% of samples meeting safety standards. Boiled Milk was largely unsafe, with over 90% of samples failing due to residual contamination. Raw Milk was the most hazardous, with 88.7% of samples failing safety criteria.

7.2. Correlation analysis of selected milk samples: the link between milk quality and contamination

Statistical correlation analysis was conducted to explore the relationship between microbial contamination and nutritional quality. Table 4 presents the summary correlation coefficients reported for the selected-sample analysis. These summary coefficients are distinct from the category-specific matrices described later in the Results section.

Table 4
Correlation Analysis of Selected Milk Samples.

Parameter	Total Viable Count (TVC)	Coliform Count	Staphylococcus aureus
Milk Fat (%)	-0.92 (Strong Negative)	-0.78	-0.85
SNF (%)	-0.80 (Moderate Negative)	-0.65	-0.74
Total Solids (%)	-0.85 (Strong Negative)	-0.69	-0.79

7.3. Key findings

Higher Milk Fat, SNF, and Total Solids correlated with lower bacterial contamination. Raw Milk samples with lower solids content were more susceptible to microbial contamination. Elevated microbial loads were linked to inferior milk quality, reinforcing the importance of proper processing.

8. Descriptive Analysis of Selected Milk Samples

Descriptive summary statistics for the selected records used in the subsequent category-specific analyses are presented in Table 5. Because this table represents a selected analytical subset, its descriptive values should not be assumed to be identical to the broader category-level summaries reported earlier.

Table 5
Analysis of Milk Samples.

Parameters	Raw Milk (Mean ± SD)	Pasteurized Milk (Mean ± SD)	Boiled Milk (Mean ± SD)
Milk Fat (%)	0.99 ± 0.06 (Range: 0.91 - 1.05)	3.55 ± 0.06 (Range: 3.50 - 3.63)	2.46 ± 1.51 (Range: 0.98 - 3.99)
Solids-Not-Fat (SNF) (%)	5.76 ± 0.44 (Range: 5.13 - 6.02)	8.51 ± 0.49 (Range: 7.71 - 8.92)	7.39 ± 1.59 (Range: 5.89 - 9.17)
Total Solids (%)	6.75 ± 0.49 (Range: 6.09 - 7.06)	11.85 ± 0.51 (Range: 11.22 - 12.50)	9.86 ± 2.68 (Range: 6.87 - 13.16)
Total Viable Count (TVC) (CFU/mL)	552,000 ± 77,705 (Max: 700,000)	0 (Effective Pasteurization Confirmed)	412,000 ± 59,396 (Max: 600,000)
Coliform Count (CFU/mL)	77,600 ± 67,382 (Max: 280,000)	0 (Effective Pasteurization Confirmed)	166,000 ± 29,217 (Max: 240,000)
Fecal Coliform Count (CFU/mL)	130,025 ± 135,822 (Max: 500,000)	0 (Effective Pasteurization Confirmed)	Not Reported
Staphylococcus aureus (CFU/mL)	2,700 ± 836 (Max: 4,000)	0 (Effective Pasteurization Confirmed)	368,000 ± 56,789 (Max: 500,000)
Salmonella & Fungi	Not Reported	Absent (Effective Pasteurization Confirmed)	Not Reported

8.1. Raw milk samples

• Milk Fat (%): Mean = 0.99%, Range = 0.91% to 1.05%

• Solids-Not-Fat (SNF) (%): Mean = 5.76%, Range = 5.13% to 6.02%

• Total Solids (%): Mean = 6.75%, Range = 6.09% to 7.06%

• Total Viable Count (TVC) (CFU/mL): Mean = 552,000, Max = 700,000

• Coliform Count (CFU/mL): Mean = 77,600, Max = 280,000

• Fecal Coliform Count (CFU/mL): Mean = 130,025, Max = 500,000

• Staphylococcus aureus (CFU/mL): Mean = 2,700, Max = 4,000

8.2. Boiled milk samples

• Milk Fat (%): Mean = 2.46%, Range = 0.98% to 3.99%

• Solids-Not-Fat (SNF) (%): Mean = 7.39%, Range = 5.89% to 9.17%

• Total Solids (%): Mean = 9.86%, Range = 6.87% to 13.16%

• Total Viable Count (TVC) (CFU/mL): Mean = 412,000, Max = 600,000

• Coliform Count (CFU/mL): Mean = 166,000, Max = 240,000

• Staphylococcus aureus (CFU/mL): Mean = 368,000, Max = 500,000

8.3. Pasteurized milk samples

• Milk Fat (%): Mean = 3.55%, Range = 3.50% to 3.63%

• Solids-Not-Fat (SNF) (%): Mean = 8.51%, Range = 7.71% to 8.92%

• Total Solids (%): Mean = 11.85%, Range = 11.22% to 12.50%

• Microbial Parameters: In the selected pasteurized subset summarized in Table 5, all microbial counts (TVC, Coliforms, Fecal Coliforms, Staphylococcus, Salmonella, and Fungi) were zero, confirming effective pasteurization within that descriptive subset.

9. Raw Milk Correlation Matrix

As shown in Figure 1, the raw-milk correlation matrix demonstrated strong inverse associations of TVC with SNF (r = -0.86) and total solids (r = -0.85). These are descriptive associations within the analyzed raw-milk subset and do not establish causality.

Figure 1
Correlation Matrix of Raw Milk Samples.

Coliform and fecal-coliform counts were perfectly correlated in the raw-milk matrix (r = 1.00), consistent with the two indicators tracking the same contamination pattern in this subset. This association supports concern regarding hygienic handling but does not identify a unique contamination source.

Milk fat showed a strong inverse association with Staphylococcus aureus counts (r = -0.92), while TVC was positively associated with Staphylococcus aureus (r = 0.76). These relationships are observational and may reflect source, handling, storage, or other unmeasured factors.

Salmonella was not included as a quantitative variable in the raw-milk correlation matrix. Therefore, no Salmonella correlation coefficient can be inferred from Figure 1; Salmonella findings should be interpreted only from the culture results reported elsewhere in the manuscript.

Coliform count showed a moderate positive correlation with total solids (r = 0.40). This pattern shows that higher solid content did not guarantee microbiological safety and that compositional measures alone cannot substitute for direct microbiological testing.

Overall, Figure 1 shows strong microbial-compositional co-variation in the raw-milk subset, especially the inverse TVC-SNF and TVC-total-solids relationships and the close coupling of the two coliform indicators. The findings support improved hygiene, handling, and cold-chain control while remaining observational rather than causal.

10. Boiled Milk Correlation Matrix

Figure 2 shows that TVC was strongly inversely correlated with milk fat in boiled milk (r = -0.83). This relationship is descriptive because milk composition, heating conditions, and post-boiling handling were not independently randomized.

Figure 2
Correlation Matrix of Boiled Milk Samples.

Coliform count was inversely correlated with total solids in the boiled-milk subset (r = -0.60). The association indicates co-variation after boiling but does not establish that higher total solids improved microbial inactivation.

Staphylococcus aureus was positively correlated with TVC in boiled milk (r = 0.72), indicating that greater overall culturable burden tended to accompany greater S. aureus counts. This is compatible with persistent or post-processing contamination and reinforces the importance of hygienic handling after boiling.

The correlation between coliform count and Staphylococcus aureus was close to zero (r = 0.028), indicating little linear co-variation between these indicators in the boiled-milk subset. Overall, Figure 2 shows that boiling did not remove the need for hygienic post-treatment handling and safe storage.

11. Pasteurized and UHT Treated Milk Correlation Matrix

In the pasteurized correlation dataset, microbial variables were uniformly zero and therefore lacked the variance required for meaningful correlation coefficients. Figure 3 consequently provides interpretable correlations only among the physicochemical variables.

Figure 3
Correlation Matrix of Pasteurized and UHT-Treated Milk Samples.

Milk fat and SNF showed a weak positive correlation (r = 0.34) in the pasteurized subset, indicating modest compositional co-variation across the analyzed records.

Total solids were strongly correlated with milk fat (r = 0.99) and moderately correlated with SNF (r = 0.68), as shown in Figure 3. These relationships are expected because total solids are compositionally related to both fat and non-fat solids.

Overall, the pasteurized correlation subset showed no detectable culturable microbial burden, while the measurable variation was primarily physicochemical. This supports the microbiological effectiveness of the tested commercial processing within this subset, without implying absolute sterility of all pasteurized products.

Across the three category-specific matrices, raw milk showed TVC-SNF r = -0.86, TVC-total-solids r = -0.85, coliform-fecal-coliform r = 1.00, and milk-fat-Staphylococcus aureus r = -0.92. Boiled milk showed TVC-milk-fat r = -0.83, coliform-total-solids r = -0.60, and TVC-Staphylococcus aureus r = 0.72. In the pasteurized subset, microbial correlations were not estimable because microbial counts were uniformly zero; among physicochemical variables, milk fat-SNF was r = 0.34, SNF-total solids r = 0.68, and milk fat-total solids r = 0.99.

The correlation analysis showed several associations between compositional measures and microbial counts, but these associations should not be interpreted causally. An OLS model was fitted with TVC as the dependent variable and milk fat, SNF, and total solids as predictors. The model had R-squared = 0.735 and adjusted R-squared = 0.469; however, the overall F test was not statistically significant (p = 0.265), and none of the individual coefficients reached p < 0.05. Model-fit statistics are summarized in Table 6, and the coefficient estimates are presented in Table 7.

Table 6
Model Summary and Goodness of Fit.

Metric	Value	Explanation
R-squared	0.735	The model explains 73.5% of the variability in Total Viable Count (CFU/mL). This suggests a reasonably strong relationship between the independent variables and the dependent variable.
Adjusted R-squared	0.469	Adjusted R2 accounts for the number of predictors. A large difference between R2 and Adjusted R2 suggests that some predictors may not be significantly contributing to the model.
F-statistic	2.768	Tests whether the independent variables jointly explain a significant amount of variance in the dependent variable. A higher F-value generally indicates a better model fit.
Prob (F-statistic)	0.265	The p-value for the F-test is not statistically significant (p > 0.05), meaning the overall model may not provide a strong explanation of the variability in bacterial count.
Log-Likelihood	-61.347	Represents the likelihood of the observed data given the model parameters. A less negative value indicates a better model fit.
AIC (Akaike Information Criterion)	128.7	A lower AIC suggests a better model fit, but should be compared across models.
BIC (Bayesian Information Criterion)	127.5	Similar to AIC but penalizes model complexity more. Lower values indicate a preferable model.

Table 7
Ordinary Least Squares Regression.

Variable	Coefficient	Standard Error	t-value	P>|t|	95% Confidence Interval
Constant	2.042e+06	8.17e+05	2.498	0.130	(-1.48e+06, 5.56e+06)
Milk Fat (%)	1.059e+05	5.16e+05	0.205	0.856	(-2.11e+06, 2.33e+06)
SNF (%)	-1.846e+05	2.88e+05	-0.642	0.587	(-1.42e+06, 1.05e+06)
Total Solids (%)	-7.866e+04	2.4e+05	-0.328	0.774	(-1.11e+06, 9.52e+05)

The regression coefficients were negative for SNF (-1.846 x 10^5 CFU/mL per percentage point; p = 0.587) and total solids (-7.866 x 10^4; p = 0.774), and positive for milk fat (1.059 x 10^5; p = 0.856). Because all three p-values were non-significant, the coefficient directions are descriptive and should not be interpreted as statistically supported independent effects.

A major limitation of this specification is exact or near-exact multicollinearity. Total solids were derived from milk fat and SNF in the analytical workflow, so entering milk fat, SNF, and total solids simultaneously creates algebraic dependence among predictors. The very small reported eigenvalue is consistent with a singular or poorly conditioned design matrix, making individual coefficient estimates unstable.

Accordingly, the regression is best treated as an exploratory model summary rather than a validated predictive model. A confirmatory model should avoid algebraically dependent predictors and should be evaluated in an adequately powered independent dataset before use for prediction or causal interpretation.

Model Summary: R-squared: 0.735, Adjusted R-squared: 0.469, F-statistic: 2.768, Prob (F-statistic): 0.265, Log-Likelihood: -61.347, AIC: 128.7, BIC: 127.5, Standard Errors assume that the covariance matrix of the errors is correctly specified. The smallest eigenvalue is 1.47e-31. This might indicate strong multicollinearity or a singular design matrix.

The intercept and predictor coefficients are retained in Table 7 for transparency. The intercept is theoretical because a milk sample with 0% fat, 0% SNF, and 0% total solids is outside the meaningful data domain. Given the non-significant overall model, non-significant individual terms, and strong multicollinearity, none of the numerical coefficients should be interpreted as a reliable independent effect on bacterial burden.

11.1. Predictive trend analysis of microbial contamination in milk

Figure 4 presents the 10-point scenario visualization of microbial contamination using arbitrary time units rather than observed longitudinal follow-up. The declining trajectories illustrate an assumed reduction in TVC; they do not demonstrate that such reductions occurred over calendar time, cannot be attributed to specific interventions, and should not be interpreted as a validated forecast.

Figure 4
Scenario-Based Trend Analysis of Microbial Contamination.

12. Public Health Risk Assessment of Microbial Contamination in Milk

Milk contamination poses a public-health threat, particularly for children, older adults, pregnant women, and immunocompromised individuals. Table 8 summarizes the vulnerability categories used for these population groups, while Table 9 summarizes major microbial hazards relevant to contaminated milk. These are contextual public-health summaries; organisms or hazards not directly assayed in the present culture panel are not presented as detections from this study.

Table 8
High-Risk Populations and Their Vulnerability to Contaminated Milk.

Population Group	Vulnerability Level	Common Health Risks	Most Hazardous Milk Type
Infants (0–2 years)	Extreme Risk	Diarrhea, vomiting, dehydration, failure to thrive	Raw & Boiled Milk
Children (3–12 years)	High Risk	Gastrointestinal infections, weakened immunity, nutrient deficiencies	Raw & Boiled Milk
Elderly (>65 years)	High Risk	Food poisoning, sepsis, pneumonia, hospitalization risk	Raw & Boiled Milk
Pregnant Women	Moderate to High Risk	Listeriosis, miscarriage, intrauterine infections	Raw & Boiled Milk
Immunocompromised Individuals (e.g., Cancer Patients, HIV+)	Extreme Risk	Severe bacterial infections, prolonged illness, multi-organ failure	Raw & Boiled Milk
General Healthy Adults	Low to Moderate Risk	Mild digestive issues, foodborne illnesses in severe cases	Raw Milk

Table 9
Health Impacts of Major Contaminants Found in Raw & Boiled Milk.

Microbial Contaminant	Common Sources in Milk	Health Risks	Severity Level
Coliform Bacteria (E. coli, Enterobacter, Klebsiella, etc.)	Fecal contamination, unhygienic handling, animal infections	Severe diarrhea, dehydration, kidney failure (HUS), bloody stools	Severe
Staphylococcus aureus	Poor sanitation, raw milk, infected dairy workers	Food poisoning, nausea, vomiting, cramps, sepsis	Severe
Salmonella spp.	Contaminated dairy farms, inadequate pasteurization	Typhoid fever, intestinal infection, prolonged diarrhea, dehydration	Severe
Listeria monocytogenes	Unpasteurized milk, dirty processing equipment	Pregnancy complications, meningitis, septicemia, neonatal infections	Severe
Clostridium perfringens	Soil, fecal contamination, poor hygiene	Foodborne illness, gas gangrene, abdominal cramping	Moderate
Fungal Contaminants (Molds, Yeasts, Mycotoxins)	Improper storage, long shelf life, dairy spoilage	Toxicity, immune suppression, allergic reactions, cancer risk	Moderate to Severe

A comparative risk framework for raw, boiled, and pasteurized milk is shown in Table 10. In this dataset, raw milk carried the greatest microbiological burden, boiling reduced but did not consistently eliminate contamination, and pasteurized milk showed the most favorable microbiological profile.

Table 10
Comparative Risk Model: Raw Milk vs. Pasteurized Milk.

Factor	Raw Milk (High Risk)	Boiled Milk (Moderate Risk)	Pasteurized Milk (Low Risk)
Bacterial Load	Very High (TVC: 105 – 106 CFU/mL)	Moderate (TVC: 103 – 105 CFU/mL)	Zero (TVC: 0 CFU/mL)
Coliform Presence	High (≥105 CFU/mL, unsafe)	Reduced but not eliminated (103 – 105 CFU/mL)	Zero (Safe)
Staphylococcus aureus	Frequent & Dangerous (105 CFU/mL)	Present (103 – 105 CFU/mL, can cause food poisoning)	Absent (Safe)
Fungal & Mycotoxin Risk	Moderate (Storage & Handling Issues)	Low (Boiling reduces fungal growth)	Minimal (Strict storage controls)
Digestive Safety for Children	Unsafe (Diarrhea, cramps, infections)	Moderate Risk	Safe
Pregnancy Safety	High risk of Listeriosis, miscarriage, and neonatal infections	Moderate risk	Safe
Risk for Immunocompromised Individuals	Extreme (Sepsis, severe illness, hospitalization)	Moderate (Not fully safe)	Safe

The broader regional comparison is summarized in Table 11. It is included only to contextualize the Pakistani findings within differing milk-distribution and regulatory environments and should not be interpreted as primary regional data generated from the 145 samples in this study.

Table 11
Regional Analysis: Identifying High-Risk Zones for Contaminated Milk. Certain regions are more vulnerable to milkborne disease outbreaks due to lax regulations, informal dairy markets, and limited access to pasteurized milk. The following risk assessment categorizes different regions based on their level of milk safety enforcement.

Region	Raw Milk Consumption Trend	Regulatory Strength	Public Health Risk Level
South Asia (Pakistan, India, Bangladesh, Nepal)	Extensive (up to 80% in rural areas)	Weak enforcement	High Risk
Sub-Saharan Africa	Common, especially in informal markets	Limited regulations	High Risk
Southeast Asia (Vietnam, Indonesia, Philippines, Thailand)	Moderate use, increasing pasteurization	Improving regulations	Moderate Risk
Middle East & North Africa (Egypt, Iran, Turkey, Morocco)	Common, but urban areas prefer pasteurized milk	Strong in some countries	Moderate Risk
Europe (France, Germany, Italy, UK, Scandinavia)	Very low raw milk consumption	Strict pasteurization laws	Low Risk
North America (USA, Canada, Mexico)	Illegal in most areas but still sold in raw milk cooperatives	Strict in US & Canada, weak in Mexico	Moderate Risk
Latin America (Brazil, Argentina, Colombia, Peru)	Raw milk still used, but pasteurization increasing	Varied enforcement	Moderate Risk
China & East Asia (China, Japan, South Korea)	Strict regulations, rare raw milk consumption	High enforcement	Low Risk

12.1. Overall critical findings

Overall, 63 of 71 raw-milk samples (88.7%) were classified as unsatisfactory. Among the 36 boiled-milk samples assigned a binary classification, 34 (94.4%) were unsatisfactory; one of the 37 collected boiled samples was not represented in that binary summary. Coliforms and Staphylococcus aureus contributed substantially to the microbiological concern.

12.2. Urgent public health recommendations

To mitigate the significant public health risks associated with raw milk consumption, a multi-faceted approach is urgently needed. Firstly, the sale of raw milk for direct human consumption must be banned through the implementation of mandatory pasteurization laws aimed at preventing outbreaks of milk-borne infections. Enforcing strict hygiene protocols for dairy farms and processing units is also essential to ensure safe milk production. Secondly, public awareness must be raised about the dangers of raw and improperly boiled milk. This can be achieved by launching nationwide educational campaigns focusing on the health risks of raw milk consumption, with particular emphasis on distributing informative materials in schools, hospitals, and rural communities. Additionally, introducing rigorous quality control measures and ensuring regular testing of dairy products are critical steps toward compliance with safety standards. Retailers and distributors should be held accountable for selling unsafe dairy products. Moreover, promoting the accessibility of safe, pasteurized milk is crucial. This effort can be supported by subsidizing pasteurized milk production to make it affordable for all socio-economic groups and strengthening supply chain regulations to prevent contamination from the farm to the consumer.

Raw milk consumption remains a substantial public-health hazard, particularly for vulnerable groups such as infants, pregnant women, and immunocompromised individuals. Boiling reduced some microbial measures but did not eliminate risk consistently in this dataset, whereas pasteurized milk provided the most favorable microbiological profile among the tested categories. Without improved control, milk-borne risk is likely to persist; stronger hygiene, regulation, public education, and access to appropriately processed milk can reduce that burden.

13. Discussion

The recurring deaths of children across Pakistan from consuming contaminated milk have transformed an everyday dietary staple into a symbol of public-health neglect. From Haripur to Rawalpindi, Khairpur, Faqirwali, and Khanpur, the same heartbreaking pattern echoes, young lives lost within hours of drinking milk tainted by invisible microbial toxins. These tragedies are not isolated accidents; they represent the biological footprints of a fractured dairy ecosystem where milk, once synonymous with nourishment, has become a vector of infection. At the heart of this crisis lies a pathogen-rich pour, raw and unpasteurized milk that circulates freely through informal supply chains, unregulated, unrefrigerated, and unmonitored, carrying coliforms, Staphylococcus aureus, and antibiotic-resistant enteric bacteria. Such contamination thrives amid the absence of cold-chain infrastructure, hygienic milking practices, and vendor awareness, thereby amplifying the risk of zoonotic transmission and antimicrobial resistance. The deaths of these children expose more than microbial negligence, they reveal a deeper systemic flaw: the persistent disconnect between food microbiology, regulatory enforcement, and public-health education.

In this context, the present study carries profound educational and translational significance. Within the realm of medical and allied-health education, understanding food safety, especially concerning widely consumed staples like milk, is essential for future practitioners. By elucidating the microbiological and nutritional disparities across raw, boiled, and pasteurized milk in Pakistan, this work provides evidence-based insights that can strengthen curriculum development in nutrition, microbiology, and public health courses (Nawaz et al., 2022; Ibrahim et al., 2023; Abbas et al., 2022). The findings not only underscore the urgent need for microbiology-driven interventions in national dairy regulation but also serve as a foundation for preventive strategies aimed at mitigating milk-borne diseases and safeguarding community health. Ultimately, this study strives to restore milk to its rightful place, a source of life, not loss.

This comprehensive investigation of 145 milk samples, classified into boiled, pasteurized, and raw, revealed marked differences in chemical integrity and microbial safety, underscoring the public health risks associated with substandard milk handling. Pasteurized milk consistently demonstrated superior nutritional metrics, including the highest mean levels of Milk Fat (3.83 ± 1.02%), Solid-Not-Fat (SNF) (9.12 ± 0.97%), and Total Solids (12.94 ± 1.34%), aligning well with national and international benchmarks such as PPFR, PSQCA, WHO, and FAO standards. Conversely, raw milk showed substantial deficiencies, with notably lower fat (2.62 ± 1.53%) and SNF (6.86 ± 1.22%) contents, suggesting issues in livestock feeding, adulteration, or dilution practices.

Critically, raw milk samples exhibited high microbial loads, including elevated Total Viable Count (TVC), coliforms, and Staphylococcus aureus, together with detection of Salmonella and fungal growth in multiple samples. Pasteurized milk showed marked microbial suppression in nearly all samples, whereas boiled milk retained substantial culturable contamination in many samples. These findings indicate that heat treatment must be accompanied by hygienic handling and storage to sustain microbiological quality.

Statistical analysis identified several strong correlations between compositional measures and microbial loads, particularly in the raw- and boiled-milk subsets. Because these were observational associations and the multivariable OLS model was non-significant and strongly collinear, the results should be interpreted as descriptive signals rather than evidence that changing fat, SNF, or total solids directly alters microbial burden.

The scenario-based visualization in Figure 4 illustrates how microbial burden could decline under an assumed progressive reduction in contamination, but it is not an empirical five-year forecast and does not support specific percentage reductions. Quantitative forecasting would require prospectively collected longitudinal data and external validation.

14. Conclusion

From a clinical and public-health standpoint, raw milk represented the highest-risk category in this study, particularly for populations vulnerable to food-borne infection. Most raw-milk samples failed the applied safety classification, and the dataset included substantial viable counts, coliform contamination, Staphylococcus aureus, and detection of Salmonella in multiple samples. Boiling reduced some microbial measures but did not consistently ensure acceptability, while pasteurized milk showed the most favorable microbiological profile. Future research should prioritize standardized sampling, validated microbial identification, transparent denominator reporting, and prospective evaluation of handling and processing interventions.

Concurrently, policy reforms must prioritize the implementation of stringent safety regulations and industry accountability measures. To mitigate the significant health risks uncovered in this study, a series of targeted strategic interventions are strongly recommended. Foremost among these is the immediate prohibition of raw milk sales for direct human consumption, supported by the nationwide enforcement of pasteurization as a regulatory standard. Implementing mandatory microbial screening at all critical control points within the milk supply chain is vital to detect and prevent contamination. Institutional training programs should be established for dairy workers and public health professionals to enhance hygienic handling practices and strengthen overall food safety compliance. Public education efforts must focus on vulnerable populations, particularly in schools, hospitals, and rural communities, ensuring widespread awareness of the dangers linked to unsafe milk. Finally, expanding the availability of pasteurized milk through government subsidies and the optimization of distribution channels will help bridge accessibility gaps and promote safer consumption across diverse socioeconomic groups. Without these critical interventions, milk-borne illnesses will likely escalate, jeopardizing population health. However, through evidence-driven policy, educational reform, and universal pasteurization, Pakistan can ensure a safer, more reliable milk supply, an outcome vital not only for consumer health but for strengthening the broader framework of medical education and disease prevention.

Acknowledgement:

We would like to acknowledge the support of Dr Naila Azam from Foundation University Medical College. We would like to acknowledge the research grant support by Foundation University Islamabad for this study in collaboration with Fauji Foods and National Institutes of Health Islamabad Pakistan. We would like to acknowledge the research support of International Center of Medical Sciences Research (ICMSR), Islamabad Pakistan.

Data Availability Statement

The research data supporting the findings of this study are available from the corresponding author upon reasonable request.

References

• ABBAS, S.Z., IMRAN, M., AHMED, A., HUSSAIN, M. and KHAN, S., 2022. Microbiological assessment of raw milk available in the metropolitan city of Sindh (Karachi), Pakistan. Pakistan Journal of Medical & Health Sciences, vol. 16, no. 3, pp. 897-902.
• AHMAD, M., 2022 [viewed 13 October 2025]. Two children die in Pakistan after consuming toxic milk[online]. ProPakistani. Available from: http://propakistani.pk/2022/05/13/two-children-die-in-pakistan-after-consuming-toxic-milk
  » http://propakistani.pk/2022/05/13/two-children-die-in-pakistan-after-consuming-toxic-milk
• ASGHAR, M., 2025 [viewed 13 October 2025]. Three children die after taking poisonous milk in Pindi[online]. Dawn News. Available from: https://www.dawn.com/news/1919947
  » https://www.dawn.com/news/1919947
• CHANDIO, T.H., SHEIKH, A.A., RIND, R., SOOMRO, A.H. and BHUTTO, M.B., 2007. Occurrence of Listeria monocytogenes in bovine milk in Hyderabad, Pakistan. Annals of Microbiology, vol. 57, no. 3, pp. 289-292. https://doi.org/10.1007/BF03175070
  » https://doi.org/10.1007/BF03175070
• CODEX ALIMENTARIUS COMMISSION, 2004. Code of hygienic practice for milk and milk products (CAC/RCP 57-2004) Rome: FAO/WHO.
• FOOD AND AGRICULTURE ORGANIZATION – FAO, 2011. Dairy development in Pakistan Rome: FAO.
• FOOD AND AGRICULTURE ORGANIZATION – FAO, 2013. Milk and dairy products in human nutrition Rome: FAO.
• GALLUP PAKISTAN, 2024. Fresh milk dominates market with 84% share compared to packaged milk 10% Islamabad: Gallup Pakistan.
• HAQ, I.U., KAMAL, M., SWELUM, A.A., KHAN, S., DE LOS RÍOS-ESCALANTE, P.R. and USMAN, T., 2024. Alarming multidrug resistance in Staphylococcus aureus isolated from raw milk of cows with subclinical mastitis: antibiotic resistance patterns and occurrence of selected resistance genes. PLoS One, vol. 19, no. 5, pp. e0301200. https://doi.org/10.1371/journal.pone.0301200
  » https://doi.org/10.1371/journal.pone.0301200
• HUSSAIN, I., JAVED, S., MUHAMMAD, K. and ALTAF, I., 2007. Prevalence of Campylobacter species in meat, milk and other food commodities in Pakistan. Food Control, vol. 18, no. 6, pp. 628-631. PMid:17188200.
• IBRAHIM, T., ASLAM, A., SHAFIQ, M., ZAFAR, S. and HUSSAIN, S., 2023. Supply of raw milk and prevalence of adulterated milk in Pakistan. Pak J Health Sci., vol. 4, no. 1, pp. 41-47.
• INDEPENDENT NEWS PAKISTAN, 2024 [viewed 13 October 2025]. Contaminated milk kills two children in Haripur[online]. Independent News Pakistan. Available from: https://www.inp.net.pk/news-detail/pakistan/contaminated-milk-kills-two-children-in-haripur
  » https://www.inp.net.pk/news-detail/pakistan/contaminated-milk-kills-two-children-in-haripur
• KHAN, N., 2024 [viewed 13 October 2025]. Four children among six killed in southern Pakistan after drinking ‘poisoned milk’[online]. Arab News. Available from: https://arab.news/z4rv4
  » https://arab.news/z4rv4
• KMIHA, S., AOUADHI, C. and BAKHROUF, A., 2017. Seasonal and regional occurrence of heat-resistant spore-forming bacteria in milk. Journal of Dairy Science, vol. 100, no. 1, pp. 50-58. PMid:28571988.
• LEE, S.H.I., CAPPATO, L.P., CORASSIN, C.H., CRUZ, A.G. and DE OLIVEIRA, C.A.F., 2019. Listeria monocytogenes in milk: occurrence and recent advances in methods for its detection. Beverages, vol. 5, no. 1, pp. 14. https://doi.org/10.3390/beverages5010014
  » https://doi.org/10.3390/beverages5010014
• NAWAZ, T., REHMAN, Z.U., ULLAH, R., AHMED, N. and SAYED, S.M., 2022. Physicochemical and adulteration study of fresh milk collected from different locations in Pakistan. Saudi Journal of Biological Sciences, vol. 29, no. 12, pp. 103449. https://doi.org/10.1016/j.sjbs.2022.103449
  » https://doi.org/10.1016/j.sjbs.2022.103449
• NEWS DESK, 2024 [viewed 13 October 2025]. Two infants die after consuming toxic milk[online]. Available from: https://24newshd.tv/23-Jul-2024/two-infants-die-after-consuming-toxic-milk
  » https://24newshd.tv/23-Jul-2024/two-infants-die-after-consuming-toxic-milk
• PAKISTAN TODAY, 2022 [viewed 13 October 2025]. Punjab Food Authority plans gradual ban on sale of loose milk in Punjab[online]. Available from: https://www.pakistantoday.com.pk/2023/01/03/pfa-plans-gradual-ban-on-sale-of-loose-milk-in-punjab
  » https://www.pakistantoday.com.pk/2023/01/03/pfa-plans-gradual-ban-on-sale-of-loose-milk-in-punjab
• PUNJAB FOOD AUTHORITY, 2018. Punjab Pure Food Regulations 2018 Lahore: Government of Punjab.
• TIRLONI, E., COSCIANI-CUNICO, E., STELLA, S. and BERNARDI, C., 2022. Bacillus cereus in dairy products and production plants: occurrence, epidemiology, and control. Dairy., vol. 3, no. 3, pp. 476-497.
• TRADE DEVELOPMENT AUTHORITY OF PAKISTAN, 2022. Dairy industry Pakistan Karachi: TDAP.
• WORLD HEALTH ORGANIZATION – WHO, 2015. WHO estimates of the global burden of foodborne diseases Geneva: WHO.
• YUNUS, A.W., RAZZAZI-FAZELI, E. and BOHM, J., 2019. Aflatoxin contamination of milk marketed in Pakistan: a longitudinal study. Toxins, vol. 11, no. 2, pp. 110. https://doi.org/10.3390/toxins11020110 PMid:30781806.
  » https://doi.org/10.3390/toxins11020110

*

e-mail: [e-mail]

Editor:

Marcelo A.M. Esquisatto

Publication Dates

• Publication in this collection
  19 Oct 2026
• Date of issue
  2026

History

• Received
  13 Nov 2025
• Accepted
  26 Feb 2026