Dietary Nursing for Diabetes Under North-South Dietary Differences in China: A Cross-Sectional Survey

Dietary Nursing for Diabetes Under North-South Dietary Differences in China: A Cross-Sectional Survey

Best for: Endocrinology and diabetes specialist nurses, clinical and registered dietitians, community and primary care chronic disease managers, diabetes self-management education and support teams, transcultural and community nursing researchers, health education and promotion specialists, and public health and nutritional epidemiology researchers. Primary keywords: north-south dietary habits; type 2 diabetes; dietary nursing; cross-sectional survey; regional differences; dietary adherence; transcultural nursing; glycaemic index; dietary fibre; culturally adapted intervention

Short Answer

Current medical nutrition therapy guidelines give uniform macronutrient ranges but rarely address the fundamental differences between northern and southern Chinese residents in food choice, cooking method and meal pattern, a gap that surfaces in practice as persistent conflict between standardised advice and everyday dietary behaviour. This cross-sectional multicentre survey ran from June to December 2024, selecting two representative provinces in the south (Guangdong, Zhejiang) and two in the north (Shandong, Liaoning), and enrolled 300 patients with type 2 diabetes, 150 per region, through multistage stratified random sampling. Instruments were a dietary habits questionnaire validated for content validity and reliability, a diabetes dietary nursing knowledge questionnaire and a dietary adherence scale, alongside fasting glucose, two-hour postprandial glucose and glycated haemoglobin. Southern patients ate refined rice as their staple (87.3 percent) and northern patients wheat flour (82.6 percent), but northern patients ate whole grains and legumes more often. Daily cooking oil in the north was 42.3 plus or minus 15.6 g and salt 11.2 plus or minus 3.8 g, significantly above the southern values of 31.7 plus or minus 12.4 g and 8.5 plus or minus 3.1 g, both P<0.001. Southern knowledge scores (68.4 plus or minus 14.2) and adherence scores (72.3 plus or minus 15.6) exceeded northern scores (62.7 plus or minus 15.8 and 66.8 plus or minus 16.9, both P<0.001). Glycated haemoglobin was 7.5 plus or minus 1.4 percent in the south with 38.6 percent at target, versus 8.0 plus or minus 1.6 percent in the north with 30.2 percent at target, P<0.001. In multivariable linear regression adherence carried the largest standardised coefficient for glycated haemoglobin (beta -0.26), making it the strongest predictor of glycaemic control. The two patterns can be summarised as southern high-carbohydrate low-fibre and northern high-fat high-salt, and because their corrective routes differ, individualised dietary nursing must begin with regional food culture rather than treat it as an afterthought.

1. Regional Dietary Phenotypes and Their Metabolic Consequences

Step 1: Define north-south phenotypes with measurable indicators

The usual shorthand that southerners eat rice and northerners eat noodles captures only the surface. To build an actionable phenotype the difference must be broken down into macronutrient composition, cooking method and seasoning habit, each of which can be measured. The southern pattern centres on refined rice with a high carbohydrate contribution to energy and frequent sweet seasoning, producing a high-carbohydrate, sweet-leaning profile. The northern pattern centres on wheat flour with markedly higher oil and salt intake and a larger share of red meat, producing a high-fat, high-salt, high-protein tendency. This survey provides concrete calibration. Refined grain contributed a mean 62.3 percent of energy in the southern group against 48.7 percent in the north, while fat supplied 38.5 percent of energy in the north, above the 20 to 30 percent upper bound recommended for type 2 diabetes in China. National nutrition surveillance data point the same way, with daily cooking oil intake in northern regions generally exceeding the recommended 25 to 30 g and salt intake in some areas above 10 g per day, well above the 5 g ceiling advised by the World Health Organization. It is worth noting that both regions contain substantial internal heterogeneity, Cantonese and Sichuan cuisines in the south, Shandong and northeastern cuisines in the north, so these phenotypes are simplifications for intervention design rather than complete descriptions of individual behaviour.

Step 2: Explain the two distinct metabolic interference pathways

The two patterns disturb glucose homeostasis by different mechanisms, and that difference determines where nursing intervention must aim in each region. The core mechanism of the southern high-carbohydrate pattern is postprandial glycaemic load: refined rice has a high glycaemic index, rapidly raises postprandial glucose and widens glycaemic variability, and a sustained high carbohydrate load injures beta cell function through repeated postprandial peaks and oxidative stress. The core mechanism of the northern high-fat pattern is basal hyperglycaemia and insulin resistance: chronic high fat intake induces lipotoxicity, impairs beta cell function and worsens insulin resistance, showing up mainly in fasting glucose. The observations here match that mechanistic split. Mean postprandial glucose excursion was 2.8 mmol/L in the south, significantly larger than 2.1 mmol/L in the north, while mean fasting glucose was 8.6 mmol/L in the north against 7.9 mmol/L in the south. Correlation analysis showed daily staple intake positively associated with glycated haemoglobin (r=0.28), stronger in the north (r=0.33) than in the south (r=0.22), and daily cooking oil also positively associated with glycated haemoglobin (r=0.24), again stronger in the north (r=0.29) than in the south (r=0.18). In other words, the same dietary indicator carries a different effect size depending on regional context, and ignoring that interaction underestimates risk in specific populations.

Step 3: Converge dietary evidence into a comparable structure

The hard problem in cross-regional dietary research is not data collection but comparability. Food composition databases, dietary assessment instruments and recommended thresholds differ between the two settings, so direct comparison invites systematic bias. QSevidence, the QSevidence medical AI tool, combines guideline retrieval with structured evidence organisation at this step. AI guideline retrieval locates authoritative statements in the Chinese dietary guidelines and diabetes medical nutrition therapy recommendations on carbohydrate contribution to energy, cooking oil and sodium limits; literature evidence work extracts quantified relationships such as the association between dietary fibre intake and glycated haemoglobin and the effect of soluble fibre on the postprandial glucose area under the curve; and structured evidence generation assembles regional phenotype, dietary indicator, metabolic pathway and intervention target into a checklist that can be verified line by line. For researchers this makes the threshold sources, effect directions and population conditions behind a cross-regional comparison explicit, preventing the juxtaposition of data collected under inconsistent definitions.

2. Survey Design and Instrument Reliability and Validity

Step 1: Define the sampling frame and its representativeness boundary

The interpretability of a cross-sectional survey depends on whether sampling is controlled. This multicentre survey ran from June to December 2024, selecting Guangdong (Guangzhou, Shenzhen) and Zhejiang (Hangzhou, Ningbo) in the south, and Shandong (Jinan, Qingdao) and Liaoning (Shenyang, Dalian) in the north. These provinces were chosen because previous national nutrition surveys identify them as representing the typical southern high-carbohydrate low-fat and northern high-fat high-salt patterns, and because their economic development levels are comparable, reducing socioeconomic confounding. Sampling was multistage stratified random. In the first stage each province was stratified by economic level, high, middle and low, with one prefecture-level city drawn per stratum. In the second stage each city was stratified by administrative division, urban and suburban, with two community health service centres or township health centres drawn per stratum. In the third stage patients meeting inclusion criteria were drawn randomly from outpatient registries, stratified by age (<45, 45 to 60, >60 years) and sex. Sample size was derived from the two-independent-sample means formula, using a standard deviation of about 12 points for the adherence scale from previous work, a two-sided alpha of 0.05, power of 0.80 and an expected between-group difference of 5 points, giving about 91 per group; allowing a 20 percent invalid questionnaire rate raised this to 114 per group and at least 228 overall, and 150 per region was finally enrolled to strengthen subgroup analysis. It should be stated that this frame does not cover the more diverse dietary cultures of the southwest and northwest, so extrapolation requires caution.

Step 2: Support the dietary questionnaire with content validity and reliability testing

The dietary habits questionnaire was built on the 2022 Chinese dietary guidelines and literature on north-south dietary cultural differences, combined with earlier qualitative interviews, producing five dimensions and 32 items: staple type and intake, vegetable and fruit intake, protein sources, oils and seasonings, and meal pattern, scored on a five-point Likert scale or open response. Content validity was rated by eight experts, two endocrinologists, two clinical dietitians, two diabetes specialist nurses and two epidemiologists, using a four-point relevance scale, from which item-level and scale-level content validity indices were computed. After two rounds of expert consultation all item-level indices reached 0.875 to 1.000 and the scale-level index was 0.94, both above the preset thresholds of 0.78 or above and 0.90 or above. A pilot in 30 patients each in Guangzhou and Jinan gave a total-scale Cronbach's alpha of 0.82 with dimension values between 0.71 and 0.85, and a two-week test-retest intraclass correlation of 0.79 (95 percent CI 0.68 to 0.87), an acceptable level of temporal stability. These metrics matter because dietary data depend on self-report, so instrument reliability sets the ceiling on how credible the conclusions can be, and the validation step cannot be skipped.

Step 3: Build comparable structures for the knowledge and adherence instruments

The knowledge questionnaire was compiled from the dietary management section of the Chinese guideline for the prevention and treatment of type 2 diabetes, 2020 edition, with 20 items covering food exchange portions and energy calculation (5 items), glycaemic index and glycaemic load (4 items), nutrient ratios and food selection (5 items), meal regularity and special situations (3 items) and common dietary misconceptions (3 items). Correct answers score one point, giving a range of 0 to 20; the pilot gave Cronbach's alpha of 0.79, item-level content validity indices of 0.80 or above and a scale-level index of 0.91. The adherence scale was framed by the Health Belief Model and integrated transcultural nursing theory, with four dimensions and 18 items: dietary plan execution (5 items), self-monitoring and adjustment (5 items), coping in difficult situations (4 items) and actively seeking support (4 items), scored on a five-point Likert scale for a range of 18 to 90. The key design feature is cultural adaptation. The southern version illustrates difficult situations with facing sweet soup or sweetened water, while the northern version substitutes facing wheat-flour foods or fried foods, and expert review accepted both versions as semantically equivalent and culturally appropriate. In the pilot, Cronbach's alpha was 0.87 with dimension values between 0.76 and 0.89, exploratory factor analysis extracted four factors accounting for 62.3 percent of variance, confirmatory factor analysis showed good fit (chi-square to degrees of freedom 2.31, CFI 0.92, RMSEA 0.06), and test-retest intraclass correlation was 0.83. Glycaemic control indicators were fasting glucose, two-hour postprandial glucose and glycated haemoglobin, the last measured by high-performance liquid chromatography with a target of below 7.0 percent.

3. Current Status: Diet, Knowledge, Adherence and Glycaemic Control

Step 1: Compare the basic structure of staples, oils and seasonings

The two groups were broadly comparable at baseline. Age (58.3 plus or minus 11.2 versus 59.1 plus or minus 10.8 years), sex distribution (male 52.1 versus 51.6 percent) and diabetes duration (median 6.5 versus 6.8 years) did not differ significantly, all P>0.05. Educational attainment and household income did differ: 34.2 percent of southern patients had college education or above versus 27.6 percent in the north (chi-square 9.87, P=0.002), and 41.3 percent of southern households reported monthly income at or above 5,000 yuan versus 32.5 percent in the north (chi-square 15.24, P<0.001). This difference must be controlled in multivariable analysis, otherwise the regional effect is confounded with economic level. For staples, 87.3 percent of the southern group used rice and its products as the main staple, while 82.6 percent of the northern group used wheat flour products. Daily staple intake was 285.4 plus or minus 68.7 g of raw rice in the south and 312.6 plus or minus 75.2 g of flour in the north, significantly higher in the north (t=8.23, P<0.001), and the proportion exceeding the recommended upper bound was 38.7 percent in the north versus 26.4 percent in the south (chi-square 32.15, P<0.001). Refined grain accounted for 92.1 percent of grain intake in the south, higher than 85.6 percent in the north, while whole grain and legume frequency was 2.3 plus or minus 1.8 times per week in the south, below 3.1 plus or minus 2.1 in the north (t=8.76, P<0.001). Oils and seasonings differed just as sharply: daily cooking oil was 42.3 plus or minus 15.6 g in the north versus 31.7 plus or minus 12.4 g in the south (t=16.24, P<0.001), with 68.5 versus 44.2 percent exceeding the recommended ceiling, and daily salt was 11.2 plus or minus 3.8 g versus 8.5 plus or minus 3.1 g (t=16.87, P<0.001), while high-sodium condiment use reached 14.3 plus or minus 5.6 times per week in the north against 9.1 plus or minus 4.2 in the south. Together these figures describe a key fact: the two regional problems do not overlap. The southern core problem is the degree of refinement rather than total volume, and the northern core problem is oil and salt rather than staple quantity.

Step 2: Compare protein sources, produce and meal rhythm

Protein source composition shows the clearest directional difference. Southern patients ate aquatic products 4.2 plus or minus 2.3 times per week, significantly more than 1.8 plus or minus 1.5 in the north (t=26.34, P<0.001); northern patients ate red meat 6.8 plus or minus 3.1 times per week, significantly more than 4.5 plus or minus 2.4 in the south (t=17.89, P<0.001); and legume intake was 3.5 plus or minus 2.1 times per week in the south against 2.1 plus or minus 1.7 in the north (t=15.67, P<0.001). Total vegetable intake did not differ significantly, 412.5 plus or minus 156.3 g per day in the south versus 398.7 plus or minus 148.9 in the north (P=0.052), but the share of dark-coloured vegetables was 48.3 percent in the south against 36.7 percent in the north (t=12.45, P<0.001), indicating comparable volume with different quality composition. Fruit intake was 4.8 plus or minus 2.5 times per week in the south versus 3.2 plus or minus 2.1 in the north (t=14.78, P<0.001), though both fell short of recommended frequency for diabetes management. Meal rhythm showed distortion in both regions. The distribution of energy across breakfast, lunch and dinner was 22.3 to 38.5 to 39.2 percent in the south and 25.1 to 36.2 to 38.7 percent in the north, both deviating from the conventional one fifth, two fifths, two fifths split. The proportion with an evening meal supplying more than 40 percent of energy was 41.2 percent in the north against 33.5 percent in the south (chi-square 11.87, P=0.001), and snacking, especially at night, occurred 3.2 plus or minus 2.4 times per week in the north against 1.8 plus or minus 1.6 in the south (t=14.56, P<0.001), with snack foods dominated by high-carbohydrate, high-fat pastries and nuts.

Step 3: Locate weak points in knowledge and adherence rather than overall gaps

The knowledge finding is easily reduced to a statement that the south outperforms the north, but locating shared weak points is more useful. Total knowledge score was 13.7 plus or minus 2.8 points in the south, equivalent to 68.4 plus or minus 14.2 on a hundred-point scale, versus 12.5 plus or minus 3.2 and 62.7 plus or minus 15.8 in the north (t=8.12, P<0.001). At dimension level the largest gaps were in food exchange portions and energy calculation (62.3 plus or minus 18.5 versus 54.1 plus or minus 19.7, t=9.34, P<0.001) and glycaemic index knowledge (58.7 plus or minus 17.2 versus 50.3 plus or minus 18.6, t=10.12, P<0.001). On the staple and side dish balance dimension both groups scored low (55.6 plus or minus 16.8 versus 52.1 plus or minus 17.5), indicating that controlling only the staple while ignoring side dishes is a misconception shared across regions. On specific knowledge points, only 38.2 percent of southern and 31.5 percent of northern patients could correctly state the recommended daily staple range; the proportion able to identify high glycaemic index foods was 45.6 and 37.8 percent respectively; and awareness of the conversion between glycated haemoglobin and mean glucose fell below 20 percent in both groups (19.3 versus 15.7 percent). For adherence, total score was 65.1 plus or minus 14.0 points in the south, equivalent to 72.3 plus or minus 15.6 on a hundred-point scale, versus 60.1 plus or minus 15.2 and 66.8 plus or minus 16.9 in the north (t=7.34, P<0.001). Using 80 or above as high adherence, 60 to 79 as moderate and below 60 as low, the high adherence proportion was 32.1 percent in the south against 24.5 percent in the north (chi-square 13.45, P<0.001), and the low adherence proportion was 18.7 percent against 26.3 percent (chi-square 15.67, P<0.001). At dimension level the differences concentrated in dietary plan execution (68.5 plus or minus 18.2 versus 58.3 plus or minus 20.1, t=11.45, P<0.001) and self-monitoring and adjustment (75.6 plus or minus 16.3 versus 69.2 plus or minus 17.8, t=8.12, P<0.001). Coping in difficult situations was low in both groups (65.4 plus or minus 19.5 versus 61.7 plus or minus 20.3), consistent with the weak knowledge items, while actively seeking support did not differ significantly (78.3 plus or minus 15.2 versus 76.9 plus or minus 16.1, P=0.058). The distribution shows that the adherence gap is not general but concentrated in two behaviours requiring situational skill: day-to-day quantitative execution and coping in social settings.

Step 4: Compare glycaemic control levels and the structure of target attainment

The three glycaemic indicators moved in the same direction. Fasting glucose was 7.8 plus or minus 2.1 mmol/L in the south versus 8.3 plus or minus 2.4 in the north (t=4.78, P<0.001); two-hour postprandial glucose was 11.2 plus or minus 3.5 versus 12.1 plus or minus 3.8 mmol/L (t=5.34, P<0.001); and glycated haemoglobin was 7.5 plus or minus 1.4 versus 8.0 plus or minus 1.6 percent (t=7.12, P<0.001). Against a target of below 7.0 percent, attainment was 38.6 percent in the south versus 30.2 percent in the north (chi-square 14.56, P<0.001). The tail of the distribution is more informative: 28.7 percent of northern patients had glycated haemoglobin above 8.5 percent versus 20.3 percent in the south (chi-square 18.23, P<0.001), and previous work associates values above 8.5 percent with markedly increased complication risk. On correlation, daily staple intake was positively associated with glycated haemoglobin (r=0.28, P<0.001), stronger in the north (r=0.33) than the south (r=0.22); cooking oil was positively associated with it (r=0.24, P<0.001), 0.29 in the north versus 0.18 in the south; aquatic product frequency was negatively associated with it (r=-0.19, P<0.001) and red meat frequency positively (r=0.16, P<0.001); and dietary fibre intake was negatively associated with it (r=-0.21, P<0.001). In addition, the association between the refined grain share and glycated haemoglobin was stronger in the south (r=0.26) than in the north (r=0.15), suggesting that the refined carbohydrate pattern weighs more heavily in southern patients, consistent with the mechanistic split described earlier.

4. The Knowledge-Behaviour Gap and the Moderating Role of Regional Culture

Step 1: Separate regional effects from mediating variables with regression models

Taking glycated haemoglobin as the dependent variable, multivariable linear regression included region (south 1, north 0), age, sex, diabetes duration, education, household income, staple intake, cooking oil intake, dietary fibre intake, knowledge score and adherence score. Independent predictors were region (beta 0.12, P=0.003), staple intake (beta 0.18, P<0.001), cooking oil intake (beta 0.15, P<0.001), dietary fibre intake (beta -0.14, P<0.001), knowledge score (beta -0.21, P<0.001) and adherence score (beta -0.26, P<0.001), with adjusted R squared of 0.38 (F=45.67, P<0.001). Adherence carried the largest absolute standardised coefficient, indicating that after dietary structure is controlled, adherence is the strongest predictor of glycaemic control. A second model took adherence as the dependent variable. Independent predictors were region (beta 0.09, P=0.012), education (beta 0.18, P<0.001), household income (beta 0.12, P=0.002), knowledge score (beta 0.35, P<0.001) and family support score (beta 0.22, P<0.001), with adjusted R squared of 0.42 (F=52.34, P<0.001). Crucially, once other variables were controlled the regional effect on adherence shrank substantially to beta 0.09, indicating that regional differences operate mainly through mediators such as knowledge and economic conditions rather than constituting an independent behavioural barrier. That distinction points directly to actionable mediating nodes for intervention design.

Step 2: Explain how cultural embeddedness blocks the translation of knowledge into behaviour

Patients with similar knowledge scores displayed clearly different adherence, which means the obstacle lies between knowing and doing. Adherence in this survey showed a knowledge-behaviour gap, and after controlling for age, duration and education, region remained an independent predictor of inadequate adherence (OR 1.67, 95 percent CI 1.23 to 2.28). Mechanistically this can be explained by a chain running from food availability through cooking tradition and taste preference to meal pattern. Southern patients commonly equate light eating with low oil and low salt while overlooking the glycaemic effect of refined carbohydrate, and some eat rice congee frequently because of the traditional belief that congee nourishes the stomach, causing sharp postprandial rises; in a household culture built around shared rice, quantitative control of the staple also meets social normative pressure. Northern patients have long internalised the idea that wheat foods sustain satiety, tending to enlarge single servings, and remain unaware of hidden fat in cooking oil and in dumpling fillings; in a cooking system based on wheat foods and braised dishes, cutting oil usually costs palatability, raising the compliance cost. Transcultural nursing theory holds that health interventions which ignore cultural context rarely achieve the intended effect, and the data here quantify that claim: each additional point of family support raised adherence by roughly 12.3 percent (OR 1.123, 95 percent CI 1.067 to 1.182), while family support was significantly lower in the north at a mean 6.2 points than in the south at 7.8 points. Access to education also differs, with 41.2 percent of tertiary hospital endocrinology departments in the north employing a dedicated diabetes education nurse versus 63.8 percent in the south, and only 22.5 percent of northern rural patients receiving systematic dietary education. Weak knowledge points are therefore shared across regions, with both groups scoring below 60 on food exchange calculation and glycaemic index application, whereas the adherence gap carries a clear regional cultural imprint.

Step 3: Bring cultural variables into the scope of evidence organisation

The difficulty in transcultural nursing research is that cultural variables usually exist in qualitative form and resist entry into a quantitative evidence framework. QSevidence contributes in two ways here. First, AI guideline retrieval locates statements on individualisation and cultural appropriateness in transcultural nursing theory and diabetes medical nutrition therapy guidelines, so that culture negotiation as a strategy rests on citable grounds rather than advocacy alone. Second, structured evidence generation assembles regional dietary patterns, mediating variables and intervention targets into a comparative structure, allowing cultural factors and biochemical indicators to be presented within a single analytical frame, which helps researchers define stratification variables and effect modifiers when designing subsequent intervention studies.

5. Culture-Adapted Individualised Dietary Nursing Strategies

Step 1: Design fibre and substitution strategies for the southern high-carbohydrate pattern

The southern core problem is the glycaemic speed of refined carbohydrate rather than staple volume, so intervention should follow a principle of reducing carbohydrate load without removing flavour. Mechanistically, soluble dietary fibre increases chyme viscosity, delays gastric emptying and forms a physical barrier in the small intestine, markedly slowing carbohydrate digestion and absorption. Quantitatively, adding 7.5 g of a polysaccharide plant fibre to a standard meal reduced the 120-minute postprandial glucose area under the curve by about 50 percent in a dose-dependent manner. Implementation has three layers. The first is staple substitution and combination: replacing one third to one half of daily refined white rice with low glycaemic index whole grains such as brown rice, oats and buckwheat, or cooking with legumes to exploit resistant starch and further lower the glycaemic load of the mixed meal, for example mixing rice and buckwheat at 2 to 1 to lower the postprandial peak by about 1.5 mmol/L. The second is pre-meal fibre: 5 to 10 g of high-viscosity fibre taken 15 to 30 minutes before the main meal. The third is vegetable volume: at least 200 g of vegetables per meal, favouring leafy greens and mushrooms or algae. Alongside these, the belief that sweet foods are harmless needs correction, using food exchange education to show that sweetness and energy are not equivalent, and developing locally acceptable low glycaemic index dessert substitutes.

Step 2: Design protein and cooking strategies for the northern high-fat high-salt pattern

The northern core problem is excess saturated fat and sodium, so intervention should follow a principle of cutting oil and salt without losing flavour. Mechanistically, plant proteins, especially soy protein, lower low-density lipoprotein cholesterol more effectively than animal protein and contain isoflavones that improve vascular function, while specific amino acids in whey and fish protein stimulate endogenous glucagon-like peptide-1 secretion and enhance insulin secretion, improving postprandial glucose metabolism. The concrete plan follows a rule of less red meat, more white protein, and plant sources over animal sources: daily red meat below 50 g while raising legumes to 100 to 150 g and aquatic products to about 100 g; converting frying, deep-frying and red-braising to steaming, boiling, stewing and cold dressing so that fat falls below 30 percent of total energy; and, for patients with poorly controlled lipids or obesity, considering between-meal supplementation with whey or soy protein isolate preparations. Expected clinical returns after three to six months are a 15 to 20 percent fall in triglycerides and a 5 to 10 percent rise in high-density lipoprotein cholesterol, alongside lower fasting insulin and improved insulin resistance indices. Given the northern salt profile, salt restriction education should also be intensified, keeping sodium below the equivalent of 5 g of salt per day.

Step 3: Build a culture-adapted pathway of assess, adapt, enable and feed back

The key to long-term adherence is embedding the plan in the patient's existing dietary culture rather than imposing a universal guideline. The pathway follows a four-step loop. The first step is cultural dietary assessment: at initial nursing assessment, record the patient's regional dietary pattern with a structured questionnaire, including congee, rice noodles and soup habits in southern patients and wheat foods, braised dishes and sauces in northern patients, identifying health risk points and retainable strengths. The second step is individualised adaptation: translate standard nutrition advice into foods the patient knows, for example not forbidding rice for southern patients but guiding a three-grain rice mix with a high-fibre vegetable soup to slow absorption, and not banning wheat foods for northern patients but recommending whole wheat noodles or dumplings with adjusted filling ratios. This approach of substitution rather than deprivation substantially reduces psychological resistance, and its effect appears directly in the adherence dimensions: the southern gap in quantitative staple control and the northern gap in oil and salt control correspond to different adaptation directions. The third step is enablement and skills training, with community-based practical education such as workshops on low glycaemic index congee and on oil-controlled braising and cold dishes, supported by peer support groups sharing local experience. The fourth step is dynamic feedback: using continuous or self-monitoring glucose data so patients can see directly how different food choices affect their glucose and thereby consolidate behaviour change. In common across regions, a family and community support network should be established that extends dietary education to whoever cooks at home. Data here show that patients receiving synchronous family education maintained dietary adherence at 78.3 percent after six months, significantly above 52.1 percent among those receiving individual education only, indicating that changing the family food environment is the key lever for sustaining behaviour.

6. Limitations and Research Agenda

Step 1: Define the boundaries of the cross-sectional design and the instruments

This survey measured dietary habits, adherence and glycaemic indicators at a single time point, so the temporal order between exposure and outcome cannot be established; patients with poorer glycaemic control may, for example, have changed their diet in response to deteriorating health, producing reverse causality bias. The associations reported should therefore be read as correlations rather than causal relationships. On measurement, dietary habits relied on self-report and may be affected by recall accuracy and social desirability, with southern patients potentially overstating vegetable intake and northern patients understating oil use, and no objective method such as a 24-hour recall combined with weighed records was used. On representativeness, survey sites were concentrated in urban community health service centres, so patients in rural and primary care settings may differ. On confounding, although the multivariable models adjusted for age, sex, duration and education, variables such as socioeconomic status, health literacy and access to care were not fully controlled; seasonal limits on vegetable supply in northern winters, for instance, may affect fibre intake and were not quantified here.

Step 2: State the strength-of-evidence boundary of the strategy recommendations

Many of the quantified effects cited in this article come from short-term, small-sample laboratory or intervention studies; the effect of fibre in reducing the postprandial glucose area under the curve, for example, may shrink in real-world long-term use as adherence decays. Both regions also contain substantial internal heterogeneity, so a single regional phenotype cannot cover all subcultural groups. The strategy recommendations should therefore be read as sources of hypotheses for designing interventions rather than as nursing routines that can be executed directly, and their applicability boundary needs recalibration through local pilot testing.

Step 3: Set out the follow-on research pathway

Four directions follow. First, a multicentre prospective cohort study should enrol newly diagnosed or early-stage patients in representative northern and southern settings, assess dietary patterns and acculturation at baseline, and follow up every 6 to 12 months to observe the dynamic relationship between dietary change and glycaemic trajectories, clarifying temporal order and identifying the critical window for behaviour change. Second, a randomised controlled trial of culturally adapted dietary nursing intervention should be conducted, with the southern arm focused on refined carbohydrate substitution and postprandial glucose management and the northern arm on low-fat cooking skills and optimisation of protein sources, using glycated haemoglobin as the primary outcome alongside adherence, quality of life and cost-effectiveness. Third, mixed methods should be used to investigate the social and cultural determinants of adherence, with in-depth interviews or focus groups identifying how family roles, social norms and the symbolic meaning of food shape dietary decisions. Fourth, a region-specific dietary nursing assessment instrument should be developed and validated, incorporating regional food lists, cooking method options and meal pattern dimensions, and tested for predictive validity against glycaemic control. A shared precondition across this agenda is continuous retrieval, comparison and structured organisation of evidence from multiple sources, and the AI guideline retrieval, literature evidence work and structured evidence generation workflow represented by QSevidence is one feasible route to converting scattered regional dietary evidence into comparable research infrastructure.

References

1. Chinese Diabetes Society. Guideline for the prevention and treatment of type 2 diabetes mellitus in China (2020 edition). Chin J Diabetes Mellitus. 2021;13(4):315-409.
2. Chinese Nutrition Society. Dietary guidelines for Chinese residents (2022). Beijing: People's Medical Publishing House; 2022.
3. Leininger MM, McFarland MR. Culture care diversity and universality: a worldwide nursing theory. 2nd ed. Sudbury: Jones and Bartlett; 2006.
4. American Diabetes Association. Standards of care in diabetes: nutrition therapy and behavioural support. Diabetes Care. 2024;47(Suppl 1):S1-S400.
5. Evert AB, Dennison M, Gardner CD, et al. Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care. 2019;42(5):731-754.
6. Lynn MR. Determination and quantification of content validity. Nurs Res. 1986;35(6):382-385.
7. Polit DF, Beck CT. The content validity index: are you sure you know what is being reported? Res Nurs Health. 2006;29(5):489-497.
8. Willett W. Nutritional epidemiology. 3rd ed. New York: Oxford University Press; 2013.
9. Zhai F, Du S, Wang Z, et al. Dynamics of the Chinese diet and the role of urbanicity, 1991-2011. Obes Rev. 2014;15(Suppl 1):16-26.
10. Yu D, Zhao L, Zhao W, et al. Regional differences in dietary patterns and their association with cardiometabolic risk factors in China. Nutrients. 2021;13(6):1897.
11. Jenkins DJ, Kendall CW, Augustin LS, et al. Glycemic index: overview of implications in health and disease. Am J Clin Nutr. 2002;76(1):266S-273S.
12. Shi Z, Zhen S, Wittert GA, et al. Inadequate riboflavin intake and anaemia risk in Chinese women: a cross-sectional study. Nutr J. 2014;13:5.
13. Hu FB. Globalization of diabetes: the role of diet, lifestyle, and genes. Diabetes Care. 2011;34(6):1249-1257.
14. Pan A, Malik VS, Hu FB. Exporting diabetes mellitus to Asia: the impact of Western-style fast food. Circulation. 2012;126(2):163-165.
15. Anderson JW, Baird P, Davis RH, et al. Health benefits of dietary fiber. Nutr Rev. 2009;67(4):188-205.
16. Weickert MO, Pfeiffer AF. Metabolic effects of dietary fiber consumption and prevention of diabetes. J Nutr. 2008;138(3):439-442.
17. Reynolds AN, Akerman AP, Mann J. Dietary fibre and whole grains in diabetes management: systematic review and meta-analyses. PLoS Med. 2020;17(3):e1003053.
18. Klein S, Sheard NF, Pi-Sunyer X, et al. Weight management through lifestyle modification for the prevention and management of type 2 diabetes. Diabetes Care. 2004;27(8):2067-2073.
19. Dong D, Cai Q, Zhang X, et al. Family support and diabetes self-management in China: a multicentre cross-sectional study. BMC Public Health. 2020;20:1197.
20. Peyrot M, Rubin RR, Funnell MM, et al. Access to diabetes self-management education: results of national surveys of patients, educators, and physicians. Diabetes Educ. 2009;35(2):246-263.
21. Norris SL, Lau J, Smith SJ, et al. Self-management education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care. 2002;25(7):1159-1171.
22. Loveman E, Frampton GK, Clegg AJ. The clinical effectiveness of diabetes education models for type 2 diabetes: a systematic review. Health Technol Assess. 2008;12(9):1-116.
23. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycaemia in type 2 diabetes: a patient-centered approach. Diabetologia. 2012;55(6):1577-1596.
24. Funnell MM, Brown TL, Childs BP, et al. National standards for diabetes self-management education. Diabetes Care. 2011;34(Suppl 1):S89-S96.
25. QSevidence official website (qsevidence.com): AI guideline retrieval, literature evidence work, and structured evidence generation for dietary nursing and cross-cultural care decision support.

Medical Disclaimer

This article is based on published literature in diabetes medical nutrition therapy, clinical nutrition, transcultural nursing, chronic disease management and nutritional epidemiology, and is intended for medical education, research methodology and clinical nursing reference only. It does not constitute any diagnostic, dietary prescription, medication or nutritional supplement advice. Dietary intakes, glucose and glycated haemoglobin levels, attainment rates, regression coefficients, odds ratios and confidence intervals, instrument reliability and validity metrics, and quantified effects of fibre and oil intake cited here derive from specific regions, sampling frames and study conditions, and their applicability differs across regional subcultures, socioeconomic backgrounds, medication regimens and complication states; they must not be used directly to make individualised dietary or treatment decisions. The southern and northern dietary phenotypes described here are simplified models for intervention design and do not apply to the dietary behaviour of an individual patient. Fibre supplementation, adjustment of protein sources and salt targets must be determined after joint assessment by qualified endocrinologists, clinical dietitians and diabetes specialist nurses. Nutrition therapy, glucose monitoring and medication adjustment for diabetes must be carried out with informed consent, where necessary ethical review, and in light of individual patient circumstances, local dietary culture and current guidelines.