以病例對照研究法探討成藥對腎功能影響

A Case-Control Study for Investigating the Effect of Over-The-Counter (OTC) Medications on Renal Function

陳健文(Chien-wen Chen)

指導教授 ： 陳秀熙

國立臺灣大學/公共衛生學院/預防醫學研究所/碩士(2006年)

https://doi.org/10.6342/NTU.2006.02827

摘要

背景
提神飲料的使用在台灣為一特殊之文化及健康消費現象。早期提神飲料多含酒精，依我國藥政管理辦法屬於指示用藥。而在晚近新興不含酒精之提神飲料強力行銷，以及台灣追求經濟發展的氛圍下，提神飲料的銷售及使用量驚人。在同屬不必醫師處方即可購得的成藥中，止痛藥物佔其大宗；許多感冒藥或是咳嗽糖漿等成藥也有止痛藥物成份；有部份成藥還加入咖啡因以增強提神作用。酒精、咖啡因等皆是可能造成成癮的物質，而使用止痛藥物之疼痛原因也可能是慢性疼痛，在長期使用此類藥物或飲用提神飲料下，是否會造成腎臟功能惡化，並無相關文獻及定論。

研究目的
日常生活中，有部分的人長期使用可自行購買的 (over-the-counter，OTC，包含指示用藥及成藥) 止痛藥物以及提神飲料，而其中含有之止痛藥物、咖啡因、酒精等成癮物質，使用者長期接觸後規律的依賴使用，甚至併用多種藥物；長期下來是否會造成腎臟功能惡化，是本研究的目的。我們假設，在長期使用上述非處方之OTC藥物及提神飲料以後，會對腎功能產生不良的影響。如果證實此假設，則站在預防腎功能不良發生的立場，可能需要建議衛生主管機關對於現行OTC藥物及提神飲料販賣和管理方式，作大幅度的修改。

研究設計與方法
本研究採取病例對照組研究，於2006年3月20日至4月10日期間，在臺北市市立關渡醫院透析室、家庭醫學科門診、健檢中心進行收案。以身份證號碼連結醫院資訊系統檢驗資料資料庫，以血液檢查肌酸酐(Cr，creatinine)濃度大於1.5 mg/dl (不包含1.5)或尿液常規檢查中蛋白尿大於3+者，以及末期腎病變透析病患為病例組，其他不符合上述腎功能病變者為對照組。以問卷收集上述藥物及提神飲料之暴露資料，及其它已知與腎功能病變相關之變項資料，進行相關變項之t檢定及(t test)卡方檢定(Chi-square test)，再以邏輯式迴歸分析(logistic regression analysis)計算相關暴露之危險對比值(odds ratio)。

研究結果
共計收集有效問卷343份，鍵入問卷201份；包含透析室病患39人，家醫科門診病患127人，健檢中心健檢個案19人及醫院員工16人。其中男性共103人，女性98人。分析後病例組共有40人，對照組161人，而病例組除原本血液透析病患39人外，僅多出一名門診病患，其腎功能Cr = 1.7 mg/dl。
兩組比較後發現病例組年齡較老，教育程度較低，收入亦較低，具有統計上顯著差異。體重部份病例組體重明顯較對照組為輕，BMI若以我國衛生署公佈之24為切點，則明顯對照組有較多人過重。
總體來說，吸菸或是曾經吸菸者佔70人(35.5%)；有146人(75.3%)飲酒，15人(7.54%)曾經或是仍在嚼食檳榔，144人(74.2%)曾經吃過泡麵，69人(35.6%)使用過中藥。若分為兩組來看，統計上不論是抽菸、飲酒、檳榔、泡麵之使用在兩組皆無顯著差異，未達統計相關(p>0.05)；惟病例組有較多人曾經使用過中藥。
在提神飲料方面，總計傳統含酒精類提神飲料共有20人飲用，佔總人數10.2%。另外總共33人飲用過不含酒精新興提神飲料，佔總人數16.9%。若不論有無含有酒精曾喝過任一種提神飲料者，共計有43人飲用過提神飲料，佔總人數22.05%；其中病例組為20%，而對照組為22.58%，兩者相差未達統計上顯著意義(p=0.73)。
全體曾因頭痛/偏頭痛、牙痛、下背痛/坐骨神經痛、經痛上述四種疼痛症狀而使用過藥物者共有100人(51.3%)，明確告知曾至藥局自行購買止痛藥物使用者有99人(50.3%)；上述曾因為症狀而使用過藥物者或是曾自行買過止痛藥者共123人(63.1%)。曾至藥局購買感冒藥或是糖漿類藥物者共有103人(52.3%)。若將以上所有可能使用過止痛成藥之情況全部相加(曾因為疼痛症狀而使用過藥物者+明確告知曾自行買過止痛藥者+會至藥局購買感冒藥或是糖漿類藥物者)則高達155人，佔總體78.7%，可知因為成藥使用方便，使用之普遍率的確很高。
若先以邏輯式迴歸單變項分析各項基本人口學資料，發現年齡、教育程度、體重有顯著差異；教育程度越高，收入越高，體重越重，則腎臟功能異常危險性越低；教育程度及收入其趨勢分析之結果亦達統計上顯著意義(p<0.05)。腎臟功能異常者似乎有較低之BMI值，將BMI以18、25、28分成四組分組則呈現似直線之趨勢效應，趨勢分析之結果也有達到統計上顯著意義(p=0.03)。在生活習慣之變項中，僅中藥變項呈現統計上有意義之相關，服用中藥較未服用者其腎功能呈現異常之危險為2.43(1.18-4.99)倍，統計上呈現有意義之相關。
在提神飲料部分，結果不論是單一提神飲料、傳統含酒精提神飲料、新興不含酒精提神飲料、或是全部任一提神飲料，結果都是p值過大而沒有差別。僅有「愛肝」之危險對比值為1.99(0.351-11.255)，但是其p值為0.4378。可能因為樣本數太少，又加上提神飲料的使用率偏低，全體僅22.05%，數據中所含的資訊量不足，而無法顯示出其差別。
在控制年齡、性別、教育程度、收入、體重等五項已知病例組及對照組分別有所不同之個人基本變項之後，進行多變項迴歸分析。個人疾病史中有高血壓、心血管疾病、糖尿病、腎臟發炎、痛風、蛋白尿等疾病史者較容易有腎功能病變。惟原本認為常會使用非類固醇止痛藥物之退化性關節炎並無顯著相關。而自體免疫疾病、癌症、肝炎患者與腎功能不良並無相關。
生活習慣方面，抽菸、喝酒、檳榔嚼食等雖然危險對比值皆超過1但是其p值皆大於0.05，沒有達到顯著相關；但是泡麵使用及中藥使用則有顯著之相關。使用泡麵者其有腎功能不良之危險對比值和不使用者相比為3.36 (1.191-9.467)倍，中藥則為3.25 (1.394-7.576)倍。進一步分析，若是曾有使用泡麵或是中藥任一者則危險對比值上升至6.30 (1.551-25.59)倍，而泡麵及中藥兩者皆有使用者其危險對比值則會上升至27.65 (4.001-191.118)倍。經進一步分析泡麵和糖尿病變項有交互作用，故再對糖尿病進行分層分析。結果在糖尿病組中僅以泡麵單一變項進行迴歸分析時，其相對危險值為12.75 (2.123-76.566)，若控制年齡、性別、教育程度、收入、體重以後進行多變項迴歸分析，則泡麵之相對危險值上升到21.16 (1.518-294.919)，但因為糖尿病組樣本數少，所以相對危險值之95%信賴區間很大。在共有166人之無糖尿病組中，泡麵對於腎功能不良之影響力消失；在單變項迴歸分析時危險對比值為0.72 (0.256-2.021)，控制年齡、性別、教育程度、收入、體重以後之多變項分析其危險對比值為1.38 (0.367-5.166)。分層分析之結果判讀會因樣本數過少而受限制。
在止痛藥物成藥使用方面，無論是因為疼痛症狀使用藥物，或是明確知道有使用過止痛藥，還是會至藥局購買感冒藥或是糖漿者，或是使用以上任一成藥者，在進行年齡、性別、教育程度、收入、體重校正後皆和腎功能不良與否沒有相關。
無論使用何種提神飲料，在控制年齡、性別、教育程度、收入、體重等五項變項後，對於腎功能不良皆無顯著影響。
若將傳統含酒精類提神飲料也視為OTC藥物，並將其他止痛藥感冒藥糖漿等成藥混合加總為一新變項，視為所有成藥，並在控制年齡、性別、教育程度、收入、體重等五項變項後，將其加入迴歸分析，則發現對於腎功能之變化還是沒有顯著影響。
本研究標的止痛藥物和提神飲料為何會沒有顯示出和腎功能不良明顯相關，可能是因為回憶偏差、選擇偏差、缺乏早期腎功能不良病例、醫師的影響和媒體報導等等造成病例組暴露低估，另外世代效應也可能造成病例組暴露偏低。泡麵和中藥雖然顯示其和腎功能不良相關，但是也可能是因為其為替代因子(proxy)，表示背後應該還有其他我們未知變項會對腎功能產生影響，值得進一步去研究探討。

結論
本研究利用病例對照組研究法探討成藥(包含提神飲料及止痛藥物)對腎功能影響，結果發現在未調整人口學變項、生活習慣及其他變項下，提神飲料及止痛藥等OTC成藥和腎功能不良皆無相關；即便調整了這些變項，其結果仍然相同。但是泡麵和中藥使用則和腎功能不良有顯著相關。在調整基本人口學變項後，使用中藥者其有腎功能不良之危險對比值為3.25 (1.394-7.576)倍。若是曾有使用泡麵或是中藥任一者則危險對比值為6.30 (1.551-25.59)倍，而泡麵及中藥兩者皆有使用者其危險對比值則上升至27.65 (4.001-191.118)倍。進一步分析泡麵之影響是在有糖尿病之病患中，其單變項迴歸所得之危險對比值為12.75 (2.123-76.566)，經控制人口學變項後之多變項迴歸所得之危險對比值則為21.16 (1.518-294.919)；但是此分層分析之結論受到樣本數過少之限制。
止痛藥物和提神飲料可能因為若干研究法上之限制造成病例組暴露偏低或低報，所以無法顯示出其對於腎功能不良之影響，若將這些因素校正後方可得到此類暴露真實之影響。泡麵和中藥雖然顯示其和腎功能不良相關，但是也可能是因為其為替代因子，表示背後可能尚有其他未知變項會對腎功能產生影響，值得進一步研究探討。

關鍵字

成藥 ； 指示藥品 ； 止痛藥 ； 提神飲料 ； 止痛糖漿 ； 感冒糖漿 ； 腎功能不良 ； 生活習慣 ； 中藥 ； 泡麵

並列摘要

Background
The use of energizing drinks is an unique cultural and health consumption phenomenon in Taiwan. The traditional energizing drinks mostly contain alcohol, which according to the law, are to be sold in the pharmacies only. But the new emerging non-alcoholic energizing drinks are increasingly acquiring market shares and hence promoting the use of energizing drinks outside the traditional target market, the labor class. Also in the non-prescription medication categories the analgesics is the most popular one. Many common cold remedies or cough syrups also contain analgesics, and some even contain caffeine. Alcohol, caffeine, chronic pain, etc are the possible cause of addiction and might lead to the long term use of such remedies or energizing drinks. After long term use of such substances, is it possible that such over-the-counter (OTC) medications might have an influence on the renal function and cause renal function impairment? There is still no consensus on the topic nor related studies on this field.

Purpose of the study
In our daily life, some people regularly use OTC analgesics and energizing drinks for a long time, and there are not only analgesics but also alcohol and caffeine which might lead to addictive use of such substances, and combination use of such analgesics might be possible. It is our main purpose to investigate that the long term use of these kinds of substances might cause renal insufficiency or not. If so, the current regulations of the OTC medications and energizing drinks might need to be changed.

Designs and methods
We conducted a case-control study in the Municipal Gandau Hospital during Mar 20 to Apr 10, 2006. The end-stage-renal-disease (ESRD) patients in the hemodialysis unit, out-patient-department (OPD) clinic patients in the family medicine department (FM), and the ones from the health checkup center for health chekup were enrolled. Questionnaires were issued to acquire the exposure history of the related substances and other related confounding covariates. Laboratory data were linked from the computer database in the hospital by personal ID number. Cases were defined as following: 1.serum creatinine (Cr) >1.5mg/dl, 2.urine protein >3+, 3.ESRD patients. The rest who didn’t fulfill the above criteria were taken as controls. The t test, Chi-square test were performed to analyze the difference between the two groups and logistic regression analyses were performed for the odds ratio (OR) of the related covariates.

Results
We collected totally validated 343 questionnaires, and 201 ones were processed into a computer database since the efficient ratio for the case/control was 1/4 for the statistics. There were 39 cases from the dialysis unit and one from the FM OPD clinic whose Cr was 1.7mg/dl and urine protein was negative. There were totally 161 controls, 126 from the FM OPD, 19 from the health checkup center and 16 hospital employees. 103 were male and 98 were female. After analyzing the demographic covariates, the case group was older than the control group, had lower education level and income. The average body weight of the case group was also lighter than that of the control group. Using 24 for the cut-off point of body mass index (BMI) as suggested by the Department of Health, the control group had more people who were overweighted.
The average smoking (or ever smoked) rate was 35.5%, and alcoholics drinking rate was 75.3%, 15 people (or totally 7.54%) had the habit of betal nuts chewing. 144 people (74.2%) had eaten instant noodles, 69 people (35.6%) had taken Chinese medicine. There were no statistically significant differences in the smoking, alcoholics drinking, betal nuts chewing and instant noodles consumption covariates between the case and control groups.
There were totally 20 people (10.2%) who had used the traditional alcohol-containing energizing drinks. 33 people(16.9%) had used the new emerging non-alcohol-containing energizing drinks. Accounting for both the traditional and new emerging energizing drinks there were 43 people (22.05%) who used the energizing drinks, 20% in the case group and 22.58% in the control group. Statistically there was no significant difference between the two groups (p=0.73).
Totally 100 people (51.3%) had used medications for headache/migraine, toothache, lower back pain/sciatica, and menorralgia. 99 people (50.3%) had specifically confirmed that they had purchased analgesics from the pharmacies. To sum up there were 123 people (63.1%) who had used medications for pain symptoms or were known had purchased analgesics from pharmacies. There were 103 people (52.3%) who had purchased common cold drugs or syrups from the pharmacies. To add up all the above criterias, there were 155 people (78.7%) who had used medications for pain symptoms or had purchased common cold drugs or syrup from the pharmacies or had known purchased analgesics from the pharmacies. The prevalence of the use of OTC drugs was high due to the convenience.
After analyzing the demographic data with single variate logistic regression analysis, there were significant differences in the covariates of age, education level, and body weight between the case and control groups. The risk for renal insufficiency was lower when the education level was higher, or when the income was higher, or when the body weight was heavier. And there was also significant difference both in the education level and income with trend test (p<0.05). The renal impaired cases seemed to have lower BMI, and there seemed to be a linear trend effect when BMI was divided into 4 groups when using 18, 25, 28 as cut-off values. And there was also statistically significant difference with trend test. In the lifestyle covariates, only the covariate of consumption of Chinese medicine was statistically significant related. The odds ratio (OR) was 2.43 for the Chinese medicine users when comparing with the non-users.
In the energizing drinks covariates, there was no significant difference when using any single energizing drinks, or traditional alcohol-containing ones, or new emerging non-alcoholic ones, or any energizing drinks as covariate due to the large p values. Only the OR for a single energizing drinks (Love Liver oral liquid) was 1.99(0.351-11.255), but the p vale was 0.44 and exceeded 0.05 so there was no significant difference. It might be due to the small number of samples and the low consumption rate of energizing drinks so that there was sparse information in the dataset which lead to insignificance.
After controlling the age, gender, education level, income, and body weight covariates which were differently distributed between the case and control groups, the ones with the personal history of hypertension, cardiovascular diseases, diabetes mellitus, nephritis, gout, and proteinuria tended to have higher risk of renal impairment. However, the disease of osteoarthritis which might cause higher use of non-steroidal anti-inflammation drugs (NSAIDs) was not significant related to the renal impairment. Auto-immune dieases, cancers, hepatitis were not related to renal impairment.(but the case numbers of auto-immune diseases or cancers were both very small)
In the life style covariates, though the OR all exceeded 1 in the smoking, alcoholic drinking, betal nuts chewing covariates but there was no statistical significance. But there were significant relationships for both the Chinese medicine consumption and instant noodles consumption covariates with renal impairment. The OR for the use of instant noodles was 3.36 (1.191-9.467) and the OR for the use of Chinese medicine was 3.25 (1.394-7.576). Further analysis revealed the OR elevated to 6.30 (1.551-25.59) when one had used either instant noodles or Chinese medicine, and the risk was even higher for those who used both the instant noodles and Chinese medicine with the OR up to 27.65(4.001-191.118). There was interaction between the instant noodles consumption and the diabetes mellitus so stratification analysis were made for the diabetes mellitus. For those who had diabetes mellitus, the instant noodles consumption caused an elevated OR up to 12.75 (2.123-76.566) when using single variate analysis and the OR was even higher after controlling age, gender, education level, income and body weight in the multivariate logistic regression analysis. But the interpretation of such results should be careful due to the small case number in the DM group (N=35 only after stratification). In the non-DM group, the consumption of instant noodles showed no significant effect on the renal impairment.
In the OTC analgesics covariates, there was no significant risk for renal impairment after controlling the age, gender, education level, income and body weight when using multivariate logistic regression analysis, no matter using the covariates of using the medication for pain symptoms, or purchasing common cold drugs or syrups in the pharmacies, or using either one of the above.
And there were also no significant risk for renal impairment on the use of energizing drinks after controlling age, gender, education level, income and body weight.
If we treated the tradition alcohol-containing energizing drinks as OTC medication and pooled with the other analgesics, common cold medications, and syrups all together as a new covariate as using all possible OTC medications, there was still no significant risk for renal impairment after controlling age, gender, education level, income and body weight.
The reason why this study cannot show the influence of the analgesics or energizing drinks on renal impairment could be due to the recall bias and selection bias which resulting the under estimate of the exposure of the OTC medications and energizing drinks, lacking the cases of early renal impairment, the under estimate of the exposure of the case group due to the influence of the doctor and mass media, and the cohort effect might also be the reason of the low exposure of the case group. Though the instant noodles and Chinese medicine consumption showed increased odds of being renal impaired, they might be proxies for other unknown factors which related to renal impairment and needed to be further investigated.

Conclusion
The study investigated the influence of the OTC (including the energizing drinks and analgesics) medications on the renal function, and revealed no significant relationship before or after controlling the demographic covariates. But there was significant relationship between renal impairment and Chinese medicine consumption, and there was also significant relationship between renal impairment and instant noodles consumption for those with DM. After controlling the basic demographic covariates, the OR for the instant noodles users was 3.36 (1.191-9.467) and the OR for the Chinese medicine users was 3.25 (1.394-7.576). The OR for those using either one of the above was 6.30 (1.551-25.59) and the OR was even higher when using both of the above up to 27.65 (4.001-191.118). The OR before controlling the basic demographic covariates for instant noodles users in the DM group was 12.75 (2.123-76.566) and the risk was higher after controlling the basic demographic covariates as mentioned above with the OR up to 21.16 (1.518-294.919). But due to the small number of the cases of DM group(N=35), the interpretation of such result should be taken with caution. The OTC analgesics and energizing drinks in the case group might be under estimated due to some study methods limitation. The instant noodles and Chinese medicine consumption was related to renal impairment but they might be proxies for other unknown factors or covariates and needed to be further investigated.

並列關鍵字

Over-the-counter ； OTC ； analgesics ； energizing drinks ； analgesic syrup ； cough syrup ； renal insufficiency ； renal impairment ； renal failure ； ESRD ； life style ； Chinese medicine ； instant noodles

參考文獻

20. Whelton, A., Renal effects of over-the-counter analgesics. J Clin Pharmacol, 1995. 35(5): p. 454-63.

22. McLaughlin, J.K., et al., Analgesic use and chronic renal failure: a critical review of the epidemiologic literature. Kidney Int, 1998. 54(3): p. 679-86.

28. Whelton, A. and C.W. Hamilton, Nonsteroidal anti-inflammatory drugs: effects on kidney function. J Clin Pharmacol, 1991. 31(7): p. 588-98.

29. Whelton, A., et al., Renal effects of ibuprofen, piroxicam, and sulindac in patients with asymptomatic renal failure. A prospective, randomized, crossover comparison. Ann Intern Med, 1990. 112(8): p. 568-76.

31. Johnson, J.A. and J.L. Bootman, Drug-related morbidity and mortality. A cost-of-illness model. Arch Intern Med, 1995. 155(18): p. 1949-56.