楼主问题提的太宽泛了，没法回答呀。在下倒是知道如果年龄作为协变量，需要将其均数作为基线参照进行校正，比如你的样本年龄在50~70岁之间，那么最好进行x-60=新变量的校正，然后用新变量替代原来的年龄变量，不然基线参照会是1岁，就不合适了。性别校正还真没见过，按说性别非1即0，还用校正吗？

研究目的：说明一个血清物质和其他血清物质之间的相关性，资料：为连续变量

看了很多文献，都说做多元线性回归，做了之后要剔除年龄、性别的影响，其中一篇文章说法如下：Statistical analysis was performed using
SPSS (version 13.0; SPSS, Chicago, IL).
Data are presented as means  SD and as
a number (in percentages) for categorical
measures. Data that were not normally
distributed were logarithmically transformed
before analysis. For continuous
variables, the differences between groups
were compared using either an unpaired
Student’s t test or one-way ANOVA. The
2 test was used to compare categorical
variables between groups. Correlations of
A-FABP with various metabolic parame-ters were analyzed using Pearson correlation
and multiple regression analysis after
adjustments for age and sex. Logistic regression
analysis was performed to assess
the odds ratio (OR) of the metabolic parameters
for the presence of overt NAFLD
after adjustments for age and sex. A-FABP
levels were grouped into tertiles in a sexspecific
manner. Multiple logistic regression
analysis was used to assess the OR for
the presence of overt NAFLD in subjects
with the higher A-FABP tertiles compared
with those with the lowest tertile. Twosided
values of P  0.05 were considered
significant.
另一篇的说法是Data are presented as mean±standard deviation (SD). Variables
such as fasting insulin, HOMA-IR, hs-CRP, triglyceride, estradiol,
and visfatin levels were logarithmically transformed prior to statistical
analyses to approximate a normal distribution. Baseline
characteristics in subjects with and without metabolic syndrome
were compared using t-tests for continuous variables and a chisquare
(2) test or Fisher’s exact test for categorical variables in
cells with an expected count <five. Circulating visfatin levels were
compared with or without each component of metabolic syndrome
using a t-test. Visfatin levels according to the number of components
(0, 1–2, ≥3) of metabolic syndrome were compared using
a one-way ANOVA. A Tukey’s post hoc test was performed to test
the significance of the pairwise difference when significant differences
were found. Pearson correlation coefficients were calculated
to evaluate the relationship between visfatin and components of
metabolic syndrome or cardiometabolic variables. In order to confirm
that the association between visfatin and metabolic syndrome
is independent of insulin resistance, which is known to be the
key factor in the development of metabolic syndrome, we evaluated
interactions between visfatin, age, BMI, and HOMA-IR using a
multiple logistic regression analysis. A multiple logistic regression
analysis was also used to evaluate interactions between visfatin,
age, BMI, HOMA-IR, and inflammatory indices such as WBC. Significance
was defined at the 0.05 level of confidence. All calculations
were performed using the Statistical Package for Social Sciences
software, version 15.0 (SPSS, Chicago, IL, USA).
是真的不明白两者的差异哈

万分感谢有高手帮忙！

好像明白些了。第一篇的Correlations of A-FABP with various metabolic parame-ters were analyzed using Pearson correlation and multiple regression analysis after adjustments for age and sex.说的是各种代谢参数与A-FABP的相关性用Pearson相关并在调整了年龄性别后用多变量回归分析。Pearson相关就是通常所说相关，其实多余，因为后面的多变量回归会更好的说明，校正年龄性别后的多变量回归，其实就是指将变量年龄和性别保留在模型内的条件下的回归，在这种情况下，年龄性别会对主要分析因素有影响，校正这种影响一般是分层分析。不过年龄最好还是取均数为“0”最妥当。好像对非正态数据还进行了对数转换。将可能对主要分析变量有影响的变量作为协变量包括在模型内而不论其有无意义应该是回归分析的常见做法。
在下说的全面吗？欢迎大侠拍砖。

意思是年龄，性别在多元回归时纳入，本来就算校正了？不需要额外的处理？

自己UPUPUPUPUP！

11# 橘子和小猫
是这个意思，年龄性别如果在模型中，主要分析因素已经“自动”校正了，这是程序设计的功劳。如果不确定，可以拟合单纯包括主要分析因素的模型，然后再拟合包括主因、性别和年龄的模型，一对比就知道了，回归系数会有变化。